diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,97639 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 69713,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5,
+      "learning_rate": 2.8686173264486516e-08,
+      "loss": 0.859,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.434308663224326e-07,
+      "loss": 1.2508,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.868617326448652e-07,
+      "loss": 1.1806,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.3029259896729773e-07,
+      "loss": 1.1581,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.737234652897304e-07,
+      "loss": 1.0476,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.17154331612163e-07,
+      "loss": 1.126,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.605851979345955e-07,
+      "loss": 1.1032,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.0040160642570282e-06,
+      "loss": 1.1629,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.1474469305794607e-06,
+      "loss": 1.2222,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.2908777969018933e-06,
+      "loss": 1.307,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.434308663224326e-06,
+      "loss": 1.0781,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.5777395295467586e-06,
+      "loss": 1.2634,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.721170395869191e-06,
+      "loss": 1.0647,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.8646012621916239e-06,
+      "loss": 1.1197,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.451171875,
+      "learning_rate": 2.0080321285140564e-06,
+      "loss": 1.0787,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.151462994836489e-06,
+      "loss": 1.0495,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.2948938611589215e-06,
+      "loss": 1.1097,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.438324727481354e-06,
+      "loss": 1.0675,
+      "step": 85
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.5817555938037866e-06,
+      "loss": 1.2212,
+      "step": 90
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.725186460126219e-06,
+      "loss": 1.0741,
+      "step": 95
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.868617326448652e-06,
+      "loss": 1.2163,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.0120481927710846e-06,
+      "loss": 1.1092,
+      "step": 105
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.155479059093517e-06,
+      "loss": 1.1085,
+      "step": 110
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.2989099254159493e-06,
+      "loss": 0.9502,
+      "step": 115
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.4921875,
+      "learning_rate": 3.442340791738382e-06,
+      "loss": 1.1783,
+      "step": 120
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.466796875,
+      "learning_rate": 3.585771658060815e-06,
+      "loss": 1.1416,
+      "step": 125
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.7292025243832477e-06,
+      "loss": 1.0175,
+      "step": 130
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.87263339070568e-06,
+      "loss": 1.1436,
+      "step": 135
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.016064257028113e-06,
+      "loss": 1.0344,
+      "step": 140
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.159495123350545e-06,
+      "loss": 1.1485,
+      "step": 145
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.302925989672978e-06,
+      "loss": 1.2148,
+      "step": 150
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.4463568559954104e-06,
+      "loss": 0.9588,
+      "step": 155
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.589787722317843e-06,
+      "loss": 1.1672,
+      "step": 160
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.7332185886402755e-06,
+      "loss": 1.0145,
+      "step": 165
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.876649454962708e-06,
+      "loss": 1.1636,
+      "step": 170
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.490234375,
+      "learning_rate": 5.020080321285141e-06,
+      "loss": 1.1657,
+      "step": 175
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.4609375,
+      "learning_rate": 5.163511187607573e-06,
+      "loss": 1.053,
+      "step": 180
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.306942053930006e-06,
+      "loss": 1.0516,
+      "step": 185
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.450372920252438e-06,
+      "loss": 1.0715,
+      "step": 190
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.44921875,
+      "learning_rate": 5.593803786574872e-06,
+      "loss": 0.9784,
+      "step": 195
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.737234652897304e-06,
+      "loss": 1.0953,
+      "step": 200
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.466796875,
+      "learning_rate": 5.880665519219737e-06,
+      "loss": 0.9477,
+      "step": 205
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.024096385542169e-06,
+      "loss": 1.1993,
+      "step": 210
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.167527251864602e-06,
+      "loss": 1.1235,
+      "step": 215
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.310958118187034e-06,
+      "loss": 1.0081,
+      "step": 220
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.478515625,
+      "learning_rate": 6.454388984509467e-06,
+      "loss": 1.0265,
+      "step": 225
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.44921875,
+      "learning_rate": 6.5978198508318986e-06,
+      "loss": 0.9588,
+      "step": 230
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.498046875,
+      "learning_rate": 6.741250717154332e-06,
+      "loss": 1.0807,
+      "step": 235
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.884681583476764e-06,
+      "loss": 1.214,
+      "step": 240
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.494140625,
+      "learning_rate": 7.028112449799197e-06,
+      "loss": 1.1833,
+      "step": 245
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.17154331612163e-06,
+      "loss": 1.0432,
+      "step": 250
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.314974182444062e-06,
+      "loss": 1.1392,
+      "step": 255
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.4584050487664955e-06,
+      "loss": 1.08,
+      "step": 260
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.601835915088927e-06,
+      "loss": 0.9535,
+      "step": 265
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.74526678141136e-06,
+      "loss": 1.1902,
+      "step": 270
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.888697647733792e-06,
+      "loss": 1.0534,
+      "step": 275
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.46484375,
+      "learning_rate": 8.032128514056226e-06,
+      "loss": 0.9991,
+      "step": 280
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.175559380378659e-06,
+      "loss": 1.03,
+      "step": 285
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.498046875,
+      "learning_rate": 8.31899024670109e-06,
+      "loss": 1.0939,
+      "step": 290
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.462421113023524e-06,
+      "loss": 1.1392,
+      "step": 295
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.49609375,
+      "learning_rate": 8.605851979345956e-06,
+      "loss": 0.9538,
+      "step": 300
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.490234375,
+      "learning_rate": 8.74928284566839e-06,
+      "loss": 0.9449,
+      "step": 305
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.892713711990821e-06,
+      "loss": 1.1433,
+      "step": 310
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.439453125,
+      "learning_rate": 9.036144578313253e-06,
+      "loss": 0.9169,
+      "step": 315
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.179575444635686e-06,
+      "loss": 0.9726,
+      "step": 320
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.323006310958118e-06,
+      "loss": 1.2255,
+      "step": 325
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.466437177280551e-06,
+      "loss": 1.2127,
+      "step": 330
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.609868043602983e-06,
+      "loss": 1.0853,
+      "step": 335
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.753298909925416e-06,
+      "loss": 0.9329,
+      "step": 340
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.4609375,
+      "learning_rate": 9.896729776247848e-06,
+      "loss": 0.8877,
+      "step": 345
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.0040160642570281e-05,
+      "loss": 1.2255,
+      "step": 350
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.0183591508892715e-05,
+      "loss": 0.9594,
+      "step": 355
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.0327022375215146e-05,
+      "loss": 1.2497,
+      "step": 360
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.047045324153758e-05,
+      "loss": 0.9763,
+      "step": 365
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.0613884107860011e-05,
+      "loss": 0.9794,
+      "step": 370
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.0757314974182445e-05,
+      "loss": 0.9919,
+      "step": 375
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.0900745840504876e-05,
+      "loss": 0.9107,
+      "step": 380
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.104417670682731e-05,
+      "loss": 1.1007,
+      "step": 385
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.1187607573149743e-05,
+      "loss": 1.1553,
+      "step": 390
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.1331038439472175e-05,
+      "loss": 0.9138,
+      "step": 395
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.1474469305794608e-05,
+      "loss": 0.9199,
+      "step": 400
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.161790017211704e-05,
+      "loss": 0.8803,
+      "step": 405
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.1761331038439473e-05,
+      "loss": 0.9837,
+      "step": 410
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.1904761904761905e-05,
+      "loss": 1.0048,
+      "step": 415
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.2048192771084338e-05,
+      "loss": 0.9014,
+      "step": 420
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.2191623637406772e-05,
+      "loss": 0.9297,
+      "step": 425
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.2335054503729204e-05,
+      "loss": 1.0564,
+      "step": 430
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.2478485370051635e-05,
+      "loss": 0.9937,
+      "step": 435
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.2621916236374069e-05,
+      "loss": 0.9515,
+      "step": 440
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.27653471026965e-05,
+      "loss": 1.1943,
+      "step": 445
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.2908777969018934e-05,
+      "loss": 1.0089,
+      "step": 450
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.3052208835341367e-05,
+      "loss": 0.9401,
+      "step": 455
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.3195639701663797e-05,
+      "loss": 1.0221,
+      "step": 460
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.333907056798623e-05,
+      "loss": 0.9598,
+      "step": 465
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.3482501434308664e-05,
+      "loss": 1.02,
+      "step": 470
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.3625932300631097e-05,
+      "loss": 1.0364,
+      "step": 475
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.3769363166953527e-05,
+      "loss": 1.068,
+      "step": 480
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.391279403327596e-05,
+      "loss": 1.1302,
+      "step": 485
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.4056224899598394e-05,
+      "loss": 0.9552,
+      "step": 490
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.4199655765920827e-05,
+      "loss": 1.1418,
+      "step": 495
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.434308663224326e-05,
+      "loss": 0.9391,
+      "step": 500
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.448651749856569e-05,
+      "loss": 1.0627,
+      "step": 505
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.4629948364888124e-05,
+      "loss": 1.0318,
+      "step": 510
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.4773379231210558e-05,
+      "loss": 0.9085,
+      "step": 515
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5,
+      "learning_rate": 1.4916810097532991e-05,
+      "loss": 0.9348,
+      "step": 520
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.5060240963855424e-05,
+      "loss": 1.0198,
+      "step": 525
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.5203671830177854e-05,
+      "loss": 0.9923,
+      "step": 530
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.5347102696500288e-05,
+      "loss": 1.1061,
+      "step": 535
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.549053356282272e-05,
+      "loss": 0.9106,
+      "step": 540
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.5633964429145155e-05,
+      "loss": 0.9399,
+      "step": 545
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.5777395295467585e-05,
+      "loss": 0.9962,
+      "step": 550
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.5920826161790018e-05,
+      "loss": 1.0109,
+      "step": 555
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5,
+      "learning_rate": 1.606425702811245e-05,
+      "loss": 0.9186,
+      "step": 560
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.6207687894434885e-05,
+      "loss": 1.1396,
+      "step": 565
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.6351118760757318e-05,
+      "loss": 1.006,
+      "step": 570
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.6494549627079748e-05,
+      "loss": 0.9358,
+      "step": 575
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.663798049340218e-05,
+      "loss": 1.0365,
+      "step": 580
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.6781411359724615e-05,
+      "loss": 0.976,
+      "step": 585
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.6924842226047048e-05,
+      "loss": 0.9621,
+      "step": 590
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.706827309236948e-05,
+      "loss": 1.0107,
+      "step": 595
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.721170395869191e-05,
+      "loss": 0.972,
+      "step": 600
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.7355134825014345e-05,
+      "loss": 1.0638,
+      "step": 605
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.749856569133678e-05,
+      "loss": 0.985,
+      "step": 610
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.764199655765921e-05,
+      "loss": 0.9469,
+      "step": 615
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.7785427423981642e-05,
+      "loss": 1.1185,
+      "step": 620
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.7928858290304075e-05,
+      "loss": 1.0834,
+      "step": 625
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.8072289156626505e-05,
+      "loss": 0.9151,
+      "step": 630
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.821572002294894e-05,
+      "loss": 0.8918,
+      "step": 635
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.8359150889271372e-05,
+      "loss": 1.0947,
+      "step": 640
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.8502581755593802e-05,
+      "loss": 0.9554,
+      "step": 645
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.8646012621916235e-05,
+      "loss": 1.0884,
+      "step": 650
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.878944348823867e-05,
+      "loss": 0.9896,
+      "step": 655
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.8932874354561102e-05,
+      "loss": 1.026,
+      "step": 660
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.9076305220883535e-05,
+      "loss": 1.047,
+      "step": 665
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.9219736087205965e-05,
+      "loss": 0.9945,
+      "step": 670
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.93631669535284e-05,
+      "loss": 1.0214,
+      "step": 675
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.9506597819850832e-05,
+      "loss": 0.9594,
+      "step": 680
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.9650028686173266e-05,
+      "loss": 0.9142,
+      "step": 685
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.9793459552495696e-05,
+      "loss": 1.0362,
+      "step": 690
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.993689041881813e-05,
+      "loss": 0.9183,
+      "step": 695
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.0080321285140562e-05,
+      "loss": 1.0075,
+      "step": 700
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.0223752151462996e-05,
+      "loss": 0.9477,
+      "step": 705
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.036718301778543e-05,
+      "loss": 1.0359,
+      "step": 710
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.051061388410786e-05,
+      "loss": 1.0742,
+      "step": 715
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.0654044750430293e-05,
+      "loss": 0.8609,
+      "step": 720
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4921875,
+      "learning_rate": 2.0797475616752726e-05,
+      "loss": 1.1525,
+      "step": 725
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.094090648307516e-05,
+      "loss": 0.9171,
+      "step": 730
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.1084337349397593e-05,
+      "loss": 1.0555,
+      "step": 735
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.1227768215720023e-05,
+      "loss": 1.0368,
+      "step": 740
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.1371199082042456e-05,
+      "loss": 0.9625,
+      "step": 745
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.151462994836489e-05,
+      "loss": 1.0213,
+      "step": 750
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.1658060814687323e-05,
+      "loss": 1.1338,
+      "step": 755
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.1801491681009753e-05,
+      "loss": 0.9872,
+      "step": 760
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.1944922547332186e-05,
+      "loss": 1.0546,
+      "step": 765
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.208835341365462e-05,
+      "loss": 1.1086,
+      "step": 770
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.2231784279977053e-05,
+      "loss": 0.936,
+      "step": 775
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.2375215146299486e-05,
+      "loss": 1.058,
+      "step": 780
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.2518646012621916e-05,
+      "loss": 0.88,
+      "step": 785
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.266207687894435e-05,
+      "loss": 1.0216,
+      "step": 790
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.2805507745266783e-05,
+      "loss": 0.8847,
+      "step": 795
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5,
+      "learning_rate": 2.2948938611589217e-05,
+      "loss": 0.9893,
+      "step": 800
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.309236947791165e-05,
+      "loss": 0.9958,
+      "step": 805
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.323580034423408e-05,
+      "loss": 0.9778,
+      "step": 810
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.3379231210556513e-05,
+      "loss": 1.105,
+      "step": 815
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.3522662076878947e-05,
+      "loss": 1.0109,
+      "step": 820
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.366609294320138e-05,
+      "loss": 1.1851,
+      "step": 825
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.9586,
+      "step": 830
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.3952954675846244e-05,
+      "loss": 1.0967,
+      "step": 835
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.4096385542168677e-05,
+      "loss": 1.1236,
+      "step": 840
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.423981640849111e-05,
+      "loss": 0.8882,
+      "step": 845
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.4383247274813544e-05,
+      "loss": 1.0636,
+      "step": 850
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.4526678141135974e-05,
+      "loss": 0.9651,
+      "step": 855
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.4670109007458407e-05,
+      "loss": 1.023,
+      "step": 860
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.481353987378084e-05,
+      "loss": 0.9432,
+      "step": 865
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.495697074010327e-05,
+      "loss": 1.2359,
+      "step": 870
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.5100401606425704e-05,
+      "loss": 1.031,
+      "step": 875
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.5243832472748137e-05,
+      "loss": 1.0033,
+      "step": 880
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.538726333907057e-05,
+      "loss": 0.9732,
+      "step": 885
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.5530694205393e-05,
+      "loss": 1.0595,
+      "step": 890
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.5674125071715434e-05,
+      "loss": 0.9085,
+      "step": 895
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.5817555938037867e-05,
+      "loss": 1.0979,
+      "step": 900
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.5960986804360297e-05,
+      "loss": 1.088,
+      "step": 905
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5,
+      "learning_rate": 2.6104417670682734e-05,
+      "loss": 1.0669,
+      "step": 910
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.6247848537005164e-05,
+      "loss": 0.9043,
+      "step": 915
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.6391279403327594e-05,
+      "loss": 0.9435,
+      "step": 920
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.46875,
+      "learning_rate": 2.653471026965003e-05,
+      "loss": 0.9912,
+      "step": 925
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.667814113597246e-05,
+      "loss": 0.9251,
+      "step": 930
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4921875,
+      "learning_rate": 2.682157200229489e-05,
+      "loss": 0.9888,
+      "step": 935
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.6965002868617328e-05,
+      "loss": 0.9168,
+      "step": 940
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.7108433734939758e-05,
+      "loss": 0.9026,
+      "step": 945
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.7251864601262195e-05,
+      "loss": 1.0875,
+      "step": 950
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.7395295467584625e-05,
+      "loss": 0.8695,
+      "step": 955
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4921875,
+      "learning_rate": 2.7538726333907055e-05,
+      "loss": 0.922,
+      "step": 960
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.768215720022949e-05,
+      "loss": 1.0311,
+      "step": 965
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.782558806655192e-05,
+      "loss": 1.0643,
+      "step": 970
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.7969018932874358e-05,
+      "loss": 1.128,
+      "step": 975
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.8112449799196788e-05,
+      "loss": 1.0122,
+      "step": 980
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.466796875,
+      "learning_rate": 2.8255880665519218e-05,
+      "loss": 1.1364,
+      "step": 985
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.8399311531841655e-05,
+      "loss": 1.0668,
+      "step": 990
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.8542742398164085e-05,
+      "loss": 0.9467,
+      "step": 995
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5,
+      "learning_rate": 2.868617326448652e-05,
+      "loss": 1.0225,
+      "step": 1000
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.882960413080895e-05,
+      "loss": 0.9799,
+      "step": 1005
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.897303499713138e-05,
+      "loss": 1.1023,
+      "step": 1010
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.911646586345382e-05,
+      "loss": 0.9273,
+      "step": 1015
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.925989672977625e-05,
+      "loss": 1.0199,
+      "step": 1020
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.48828125,
+      "learning_rate": 2.9403327596098685e-05,
+      "loss": 1.0596,
+      "step": 1025
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.9546758462421115e-05,
+      "loss": 0.8337,
+      "step": 1030
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.9690189328743545e-05,
+      "loss": 0.9971,
+      "step": 1035
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.9833620195065982e-05,
+      "loss": 0.9867,
+      "step": 1040
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5,
+      "learning_rate": 2.9977051061388412e-05,
+      "loss": 0.8743,
+      "step": 1045
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.012048192771085e-05,
+      "loss": 1.0844,
+      "step": 1050
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.490234375,
+      "learning_rate": 3.026391279403328e-05,
+      "loss": 0.9911,
+      "step": 1055
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.040734366035571e-05,
+      "loss": 0.965,
+      "step": 1060
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.055077452667814e-05,
+      "loss": 0.9537,
+      "step": 1065
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.0694205393000576e-05,
+      "loss": 0.9508,
+      "step": 1070
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.0837636259323e-05,
+      "loss": 1.0471,
+      "step": 1075
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.098106712564544e-05,
+      "loss": 1.1665,
+      "step": 1080
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.112449799196787e-05,
+      "loss": 0.8957,
+      "step": 1085
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.126792885829031e-05,
+      "loss": 1.0339,
+      "step": 1090
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.1411359724612736e-05,
+      "loss": 0.9343,
+      "step": 1095
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.155479059093517e-05,
+      "loss": 1.0931,
+      "step": 1100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.16982214572576e-05,
+      "loss": 0.932,
+      "step": 1105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5,
+      "learning_rate": 3.1841652323580036e-05,
+      "loss": 0.97,
+      "step": 1110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.198508318990247e-05,
+      "loss": 1.0856,
+      "step": 1115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.21285140562249e-05,
+      "loss": 0.8973,
+      "step": 1120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.227194492254733e-05,
+      "loss": 1.018,
+      "step": 1125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.241537578886977e-05,
+      "loss": 1.0977,
+      "step": 1130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.2558806655192196e-05,
+      "loss": 0.8744,
+      "step": 1135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.2702237521514636e-05,
+      "loss": 0.8979,
+      "step": 1140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.284566838783706e-05,
+      "loss": 0.9762,
+      "step": 1145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.2989099254159496e-05,
+      "loss": 0.9018,
+      "step": 1150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.313253012048193e-05,
+      "loss": 1.0398,
+      "step": 1155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.327596098680436e-05,
+      "loss": 1.0457,
+      "step": 1160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.3419391853126796e-05,
+      "loss": 0.984,
+      "step": 1165
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.356282271944923e-05,
+      "loss": 0.8453,
+      "step": 1170
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4921875,
+      "learning_rate": 3.3706253585771656e-05,
+      "loss": 1.1258,
+      "step": 1175
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.3849684452094096e-05,
+      "loss": 0.9574,
+      "step": 1180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.399311531841652e-05,
+      "loss": 0.9937,
+      "step": 1185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.413654618473896e-05,
+      "loss": 1.0114,
+      "step": 1190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4609375,
+      "learning_rate": 3.427997705106139e-05,
+      "loss": 0.9564,
+      "step": 1195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.442340791738382e-05,
+      "loss": 1.1259,
+      "step": 1200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.466796875,
+      "learning_rate": 3.456683878370626e-05,
+      "loss": 0.9919,
+      "step": 1205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.471026965002869e-05,
+      "loss": 1.0561,
+      "step": 1210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.474609375,
+      "learning_rate": 3.485370051635112e-05,
+      "loss": 0.9138,
+      "step": 1215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.499713138267356e-05,
+      "loss": 0.9386,
+      "step": 1220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.5140562248995983e-05,
+      "loss": 0.9422,
+      "step": 1225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.482421875,
+      "learning_rate": 3.528399311531842e-05,
+      "loss": 0.9768,
+      "step": 1230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.542742398164085e-05,
+      "loss": 1.1266,
+      "step": 1235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.5570854847963284e-05,
+      "loss": 1.0733,
+      "step": 1240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 1.001,
+      "step": 1245
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.585771658060815e-05,
+      "loss": 0.8879,
+      "step": 1250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.6001147446930584e-05,
+      "loss": 1.0183,
+      "step": 1255
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.614457831325301e-05,
+      "loss": 0.9491,
+      "step": 1260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.6288009179575444e-05,
+      "loss": 0.8943,
+      "step": 1265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.643144004589788e-05,
+      "loss": 1.1007,
+      "step": 1270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.657487091222031e-05,
+      "loss": 0.9368,
+      "step": 1275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.6718301778542744e-05,
+      "loss": 0.9788,
+      "step": 1280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.686173264486518e-05,
+      "loss": 0.9342,
+      "step": 1285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5,
+      "learning_rate": 3.7005163511187604e-05,
+      "loss": 1.0328,
+      "step": 1290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.380859375,
+      "learning_rate": 3.7148594377510044e-05,
+      "loss": 0.8514,
+      "step": 1295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.729202524383247e-05,
+      "loss": 1.0245,
+      "step": 1300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.743545611015491e-05,
+      "loss": 0.9722,
+      "step": 1305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.757888697647734e-05,
+      "loss": 1.0261,
+      "step": 1310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.772231784279977e-05,
+      "loss": 0.9276,
+      "step": 1315
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5,
+      "learning_rate": 3.7865748709122204e-05,
+      "loss": 1.0778,
+      "step": 1320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.482421875,
+      "learning_rate": 3.800917957544464e-05,
+      "loss": 1.0856,
+      "step": 1325
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4921875,
+      "learning_rate": 3.815261044176707e-05,
+      "loss": 0.928,
+      "step": 1330
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.8296041308089504e-05,
+      "loss": 0.9267,
+      "step": 1335
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.453125,
+      "learning_rate": 3.843947217441193e-05,
+      "loss": 0.9876,
+      "step": 1340
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.858290304073437e-05,
+      "loss": 1.0872,
+      "step": 1345
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4921875,
+      "learning_rate": 3.87263339070568e-05,
+      "loss": 0.8564,
+      "step": 1350
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.886976477337924e-05,
+      "loss": 1.1186,
+      "step": 1355
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.498046875,
+      "learning_rate": 3.9013195639701665e-05,
+      "loss": 0.9905,
+      "step": 1360
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.91566265060241e-05,
+      "loss": 1.0624,
+      "step": 1365
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5,
+      "learning_rate": 3.930005737234653e-05,
+      "loss": 0.9626,
+      "step": 1370
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.9443488238668965e-05,
+      "loss": 0.9534,
+      "step": 1375
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.958691910499139e-05,
+      "loss": 1.0367,
+      "step": 1380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.973034997131383e-05,
+      "loss": 0.9886,
+      "step": 1385
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4765625,
+      "learning_rate": 3.987378083763626e-05,
+      "loss": 1.1053,
+      "step": 1390
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5,
+      "learning_rate": 4.00172117039587e-05,
+      "loss": 0.9696,
+      "step": 1395
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.0160642570281125e-05,
+      "loss": 1.128,
+      "step": 1400
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.030407343660356e-05,
+      "loss": 0.9621,
+      "step": 1405
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.044750430292599e-05,
+      "loss": 0.9351,
+      "step": 1410
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.0590935169248425e-05,
+      "loss": 0.9294,
+      "step": 1415
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.073436603557086e-05,
+      "loss": 1.0204,
+      "step": 1420
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.087779690189329e-05,
+      "loss": 1.0157,
+      "step": 1425
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5,
+      "learning_rate": 4.102122776821572e-05,
+      "loss": 0.9373,
+      "step": 1430
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.116465863453816e-05,
+      "loss": 1.138,
+      "step": 1435
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5,
+      "learning_rate": 4.1308089500860585e-05,
+      "loss": 0.8953,
+      "step": 1440
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.1451520367183025e-05,
+      "loss": 1.0174,
+      "step": 1445
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.159495123350545e-05,
+      "loss": 0.9657,
+      "step": 1450
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.1738382099827885e-05,
+      "loss": 1.029,
+      "step": 1455
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.188181296615032e-05,
+      "loss": 1.0178,
+      "step": 1460
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.202524383247275e-05,
+      "loss": 0.9021,
+      "step": 1465
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.2168674698795186e-05,
+      "loss": 0.8547,
+      "step": 1470
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.231210556511762e-05,
+      "loss": 0.9427,
+      "step": 1475
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.2455536431440046e-05,
+      "loss": 0.9815,
+      "step": 1480
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.259896729776248e-05,
+      "loss": 0.9458,
+      "step": 1485
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.274239816408491e-05,
+      "loss": 1.0863,
+      "step": 1490
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.2885829030407346e-05,
+      "loss": 0.9778,
+      "step": 1495
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.302925989672978e-05,
+      "loss": 1.0188,
+      "step": 1500
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.317269076305221e-05,
+      "loss": 0.8671,
+      "step": 1505
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.3316121629374646e-05,
+      "loss": 0.9179,
+      "step": 1510
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.345955249569707e-05,
+      "loss": 1.0784,
+      "step": 1515
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.3602983362019506e-05,
+      "loss": 0.9223,
+      "step": 1520
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.374641422834194e-05,
+      "loss": 1.0345,
+      "step": 1525
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.388984509466437e-05,
+      "loss": 1.042,
+      "step": 1530
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.4033275960986806e-05,
+      "loss": 0.9997,
+      "step": 1535
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.417670682730924e-05,
+      "loss": 1.0368,
+      "step": 1540
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.4320137693631666e-05,
+      "loss": 0.9121,
+      "step": 1545
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.4463568559954106e-05,
+      "loss": 0.9296,
+      "step": 1550
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.460699942627653e-05,
+      "loss": 0.9765,
+      "step": 1555
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.475043029259897e-05,
+      "loss": 1.0278,
+      "step": 1560
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.48938611589214e-05,
+      "loss": 1.0368,
+      "step": 1565
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.503729202524383e-05,
+      "loss": 0.9063,
+      "step": 1570
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.5180722891566266e-05,
+      "loss": 0.9922,
+      "step": 1575
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.53241537578887e-05,
+      "loss": 0.9436,
+      "step": 1580
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.546758462421113e-05,
+      "loss": 1.1216,
+      "step": 1585
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.5611015490533566e-05,
+      "loss": 1.095,
+      "step": 1590
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.575444635685599e-05,
+      "loss": 1.0015,
+      "step": 1595
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.589787722317843e-05,
+      "loss": 0.9483,
+      "step": 1600
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.604130808950086e-05,
+      "loss": 0.9021,
+      "step": 1605
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.61847389558233e-05,
+      "loss": 0.9578,
+      "step": 1610
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.632816982214573e-05,
+      "loss": 1.0509,
+      "step": 1615
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.647160068846816e-05,
+      "loss": 0.9595,
+      "step": 1620
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.6615031554790593e-05,
+      "loss": 1.0539,
+      "step": 1625
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.675846242111303e-05,
+      "loss": 1.0104,
+      "step": 1630
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.690189328743546e-05,
+      "loss": 1.0165,
+      "step": 1635
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.7045324153757894e-05,
+      "loss": 0.9361,
+      "step": 1640
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.718875502008032e-05,
+      "loss": 0.928,
+      "step": 1645
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.733218588640276e-05,
+      "loss": 0.9363,
+      "step": 1650
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.747561675272519e-05,
+      "loss": 0.9143,
+      "step": 1655
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 0.9955,
+      "step": 1660
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.7762478485370054e-05,
+      "loss": 0.9988,
+      "step": 1665
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.790590935169249e-05,
+      "loss": 0.9563,
+      "step": 1670
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.804934021801492e-05,
+      "loss": 1.0039,
+      "step": 1675
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.8192771084337354e-05,
+      "loss": 1.1524,
+      "step": 1680
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.833620195065978e-05,
+      "loss": 1.0741,
+      "step": 1685
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.847963281698222e-05,
+      "loss": 0.9556,
+      "step": 1690
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.862306368330465e-05,
+      "loss": 0.8395,
+      "step": 1695
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.876649454962709e-05,
+      "loss": 1.0093,
+      "step": 1700
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.8909925415949514e-05,
+      "loss": 0.9998,
+      "step": 1705
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.905335628227195e-05,
+      "loss": 1.0079,
+      "step": 1710
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.919678714859438e-05,
+      "loss": 0.9187,
+      "step": 1715
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.9340218014916814e-05,
+      "loss": 1.0315,
+      "step": 1720
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.948364888123925e-05,
+      "loss": 0.833,
+      "step": 1725
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.962707974756168e-05,
+      "loss": 1.0263,
+      "step": 1730
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.977051061388411e-05,
+      "loss": 0.9684,
+      "step": 1735
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.991394148020654e-05,
+      "loss": 0.969,
+      "step": 1740
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.474609375,
+      "learning_rate": 5.0057372346528974e-05,
+      "loss": 1.0403,
+      "step": 1745
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.462890625,
+      "learning_rate": 5.020080321285141e-05,
+      "loss": 0.9688,
+      "step": 1750
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.47265625,
+      "learning_rate": 5.0344234079173834e-05,
+      "loss": 0.928,
+      "step": 1755
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.478515625,
+      "learning_rate": 5.0487664945496275e-05,
+      "loss": 0.8397,
+      "step": 1760
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.494140625,
+      "learning_rate": 5.063109581181871e-05,
+      "loss": 1.1001,
+      "step": 1765
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.077452667814114e-05,
+      "loss": 0.9803,
+      "step": 1770
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.091795754446357e-05,
+      "loss": 1.1502,
+      "step": 1775
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.48046875,
+      "learning_rate": 5.1061388410786e-05,
+      "loss": 0.9695,
+      "step": 1780
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.120481927710844e-05,
+      "loss": 0.969,
+      "step": 1785
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.48828125,
+      "learning_rate": 5.134825014343087e-05,
+      "loss": 1.1124,
+      "step": 1790
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.14916810097533e-05,
+      "loss": 0.9795,
+      "step": 1795
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.1635111876075735e-05,
+      "loss": 1.0894,
+      "step": 1800
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.48828125,
+      "learning_rate": 5.177854274239816e-05,
+      "loss": 0.9983,
+      "step": 1805
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.462890625,
+      "learning_rate": 5.1921973608720595e-05,
+      "loss": 0.9976,
+      "step": 1810
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.474609375,
+      "learning_rate": 5.2065404475043035e-05,
+      "loss": 1.0442,
+      "step": 1815
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5,
+      "learning_rate": 5.220883534136547e-05,
+      "loss": 1.1065,
+      "step": 1820
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.458984375,
+      "learning_rate": 5.2352266207687895e-05,
+      "loss": 0.9084,
+      "step": 1825
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.45703125,
+      "learning_rate": 5.249569707401033e-05,
+      "loss": 1.0342,
+      "step": 1830
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.263912794033277e-05,
+      "loss": 1.0483,
+      "step": 1835
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.278255880665519e-05,
+      "loss": 0.9213,
+      "step": 1840
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.451171875,
+      "learning_rate": 5.292598967297763e-05,
+      "loss": 1.0645,
+      "step": 1845
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.306942053930006e-05,
+      "loss": 1.0508,
+      "step": 1850
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.321285140562249e-05,
+      "loss": 0.8046,
+      "step": 1855
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.335628227194492e-05,
+      "loss": 0.8508,
+      "step": 1860
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4765625,
+      "learning_rate": 5.349971313826736e-05,
+      "loss": 1.0346,
+      "step": 1865
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4375,
+      "learning_rate": 5.364314400458978e-05,
+      "loss": 0.8905,
+      "step": 1870
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4921875,
+      "learning_rate": 5.378657487091222e-05,
+      "loss": 1.133,
+      "step": 1875
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.435546875,
+      "learning_rate": 5.3930005737234656e-05,
+      "loss": 1.1266,
+      "step": 1880
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4140625,
+      "learning_rate": 5.4073436603557096e-05,
+      "loss": 1.0006,
+      "step": 1885
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.451171875,
+      "learning_rate": 5.4216867469879516e-05,
+      "loss": 0.8921,
+      "step": 1890
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.4360298336201956e-05,
+      "loss": 1.0013,
+      "step": 1895
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.49609375,
+      "learning_rate": 5.450372920252439e-05,
+      "loss": 0.9136,
+      "step": 1900
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.44921875,
+      "learning_rate": 5.4647160068846816e-05,
+      "loss": 0.8111,
+      "step": 1905
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4765625,
+      "learning_rate": 5.479059093516925e-05,
+      "loss": 0.9691,
+      "step": 1910
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.359375,
+      "learning_rate": 5.493402180149169e-05,
+      "loss": 0.8102,
+      "step": 1915
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.47265625,
+      "learning_rate": 5.507745266781411e-05,
+      "loss": 0.9446,
+      "step": 1920
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.522088353413655e-05,
+      "loss": 0.9894,
+      "step": 1925
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.48046875,
+      "learning_rate": 5.536431440045898e-05,
+      "loss": 0.9952,
+      "step": 1930
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.482421875,
+      "learning_rate": 5.5507745266781416e-05,
+      "loss": 0.8921,
+      "step": 1935
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.565117613310384e-05,
+      "loss": 1.0289,
+      "step": 1940
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.579460699942628e-05,
+      "loss": 1.0987,
+      "step": 1945
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.455078125,
+      "learning_rate": 5.5938037865748716e-05,
+      "loss": 0.8982,
+      "step": 1950
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.608146873207114e-05,
+      "loss": 0.9589,
+      "step": 1955
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.6224899598393576e-05,
+      "loss": 1.0076,
+      "step": 1960
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4765625,
+      "learning_rate": 5.636833046471601e-05,
+      "loss": 1.0098,
+      "step": 1965
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4140625,
+      "learning_rate": 5.6511761331038436e-05,
+      "loss": 0.9846,
+      "step": 1970
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.6655192197360876e-05,
+      "loss": 1.0777,
+      "step": 1975
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.679862306368331e-05,
+      "loss": 0.9375,
+      "step": 1980
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.48046875,
+      "learning_rate": 5.694205393000574e-05,
+      "loss": 1.0631,
+      "step": 1985
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4609375,
+      "learning_rate": 5.708548479632817e-05,
+      "loss": 0.9297,
+      "step": 1990
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4140625,
+      "learning_rate": 5.72289156626506e-05,
+      "loss": 0.7953,
+      "step": 1995
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.46875,
+      "learning_rate": 5.737234652897304e-05,
+      "loss": 1.0055,
+      "step": 2000
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.435546875,
+      "learning_rate": 5.751577739529547e-05,
+      "loss": 1.0526,
+      "step": 2005
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.76592082616179e-05,
+      "loss": 0.9813,
+      "step": 2010
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.47265625,
+      "learning_rate": 5.780263912794034e-05,
+      "loss": 1.0185,
+      "step": 2015
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.474609375,
+      "learning_rate": 5.794606999426276e-05,
+      "loss": 0.9883,
+      "step": 2020
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.412109375,
+      "learning_rate": 5.80895008605852e-05,
+      "loss": 0.9408,
+      "step": 2025
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.462890625,
+      "learning_rate": 5.823293172690764e-05,
+      "loss": 1.0352,
+      "step": 2030
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.462890625,
+      "learning_rate": 5.8376362593230063e-05,
+      "loss": 0.8128,
+      "step": 2035
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.45703125,
+      "learning_rate": 5.85197934595525e-05,
+      "loss": 1.0551,
+      "step": 2040
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.44921875,
+      "learning_rate": 5.866322432587493e-05,
+      "loss": 0.9347,
+      "step": 2045
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.462890625,
+      "learning_rate": 5.880665519219737e-05,
+      "loss": 0.8477,
+      "step": 2050
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.895008605851979e-05,
+      "loss": 0.9384,
+      "step": 2055
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.478515625,
+      "learning_rate": 5.909351692484223e-05,
+      "loss": 1.0149,
+      "step": 2060
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.9236947791164664e-05,
+      "loss": 1.0872,
+      "step": 2065
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.443359375,
+      "learning_rate": 5.938037865748709e-05,
+      "loss": 0.8893,
+      "step": 2070
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.40625,
+      "learning_rate": 5.9523809523809524e-05,
+      "loss": 0.9899,
+      "step": 2075
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.447265625,
+      "learning_rate": 5.9667240390131964e-05,
+      "loss": 0.9359,
+      "step": 2080
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4296875,
+      "learning_rate": 5.9810671256454384e-05,
+      "loss": 0.9945,
+      "step": 2085
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.486328125,
+      "learning_rate": 5.9954102122776824e-05,
+      "loss": 0.8843,
+      "step": 2090
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.423828125,
+      "learning_rate": 6.009753298909926e-05,
+      "loss": 0.9516,
+      "step": 2095
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.484375,
+      "learning_rate": 6.02409638554217e-05,
+      "loss": 0.9624,
+      "step": 2100
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.43359375,
+      "learning_rate": 6.038439472174412e-05,
+      "loss": 0.8447,
+      "step": 2105
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.052782558806656e-05,
+      "loss": 0.979,
+      "step": 2110
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.46875,
+      "learning_rate": 6.067125645438899e-05,
+      "loss": 0.9816,
+      "step": 2115
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.453125,
+      "learning_rate": 6.081468732071142e-05,
+      "loss": 1.0292,
+      "step": 2120
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.408203125,
+      "learning_rate": 6.095811818703385e-05,
+      "loss": 1.0244,
+      "step": 2125
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.484375,
+      "learning_rate": 6.110154905335628e-05,
+      "loss": 1.1965,
+      "step": 2130
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4296875,
+      "learning_rate": 6.124497991967871e-05,
+      "loss": 0.9508,
+      "step": 2135
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.421875,
+      "learning_rate": 6.138841078600115e-05,
+      "loss": 1.2246,
+      "step": 2140
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3984375,
+      "learning_rate": 6.153184165232359e-05,
+      "loss": 0.9361,
+      "step": 2145
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.46484375,
+      "learning_rate": 6.1675272518646e-05,
+      "loss": 1.0158,
+      "step": 2150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.47265625,
+      "learning_rate": 6.181870338496844e-05,
+      "loss": 0.885,
+      "step": 2155
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.458984375,
+      "learning_rate": 6.196213425129088e-05,
+      "loss": 0.8979,
+      "step": 2160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.210556511761332e-05,
+      "loss": 1.0512,
+      "step": 2165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.404296875,
+      "learning_rate": 6.224899598393574e-05,
+      "loss": 0.9623,
+      "step": 2170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.486328125,
+      "learning_rate": 6.239242685025818e-05,
+      "loss": 0.8594,
+      "step": 2175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4453125,
+      "learning_rate": 6.253585771658062e-05,
+      "loss": 0.9633,
+      "step": 2180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.470703125,
+      "learning_rate": 6.267928858290304e-05,
+      "loss": 0.9947,
+      "step": 2185
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4609375,
+      "learning_rate": 6.282271944922547e-05,
+      "loss": 0.8975,
+      "step": 2190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.41796875,
+      "learning_rate": 6.296615031554791e-05,
+      "loss": 0.9396,
+      "step": 2195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.490234375,
+      "learning_rate": 6.310958118187034e-05,
+      "loss": 1.024,
+      "step": 2200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.39453125,
+      "learning_rate": 6.325301204819278e-05,
+      "loss": 1.0322,
+      "step": 2205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5,
+      "learning_rate": 6.33964429145152e-05,
+      "loss": 0.9946,
+      "step": 2210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.421875,
+      "learning_rate": 6.353987378083765e-05,
+      "loss": 1.0883,
+      "step": 2215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.455078125,
+      "learning_rate": 6.368330464716007e-05,
+      "loss": 1.0424,
+      "step": 2220
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.44140625,
+      "learning_rate": 6.382673551348251e-05,
+      "loss": 0.8524,
+      "step": 2225
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.40234375,
+      "learning_rate": 6.397016637980494e-05,
+      "loss": 0.9997,
+      "step": 2230
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.411359724612737e-05,
+      "loss": 0.824,
+      "step": 2235
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.435546875,
+      "learning_rate": 6.42570281124498e-05,
+      "loss": 1.0312,
+      "step": 2240
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.451171875,
+      "learning_rate": 6.440045897877223e-05,
+      "loss": 0.92,
+      "step": 2245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.423828125,
+      "learning_rate": 6.454388984509466e-05,
+      "loss": 0.879,
+      "step": 2250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.451171875,
+      "learning_rate": 6.46873207114171e-05,
+      "loss": 1.0229,
+      "step": 2255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3828125,
+      "learning_rate": 6.483075157773954e-05,
+      "loss": 1.0514,
+      "step": 2260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.408203125,
+      "learning_rate": 6.497418244406197e-05,
+      "loss": 0.9588,
+      "step": 2265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.470703125,
+      "learning_rate": 6.511761331038439e-05,
+      "loss": 1.0641,
+      "step": 2270
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.486328125,
+      "learning_rate": 6.526104417670683e-05,
+      "loss": 0.9099,
+      "step": 2275
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.462890625,
+      "learning_rate": 6.540447504302927e-05,
+      "loss": 1.0107,
+      "step": 2280
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.48046875,
+      "learning_rate": 6.55479059093517e-05,
+      "loss": 1.0007,
+      "step": 2285
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.462890625,
+      "learning_rate": 6.569133677567413e-05,
+      "loss": 1.0989,
+      "step": 2290
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.45703125,
+      "learning_rate": 6.583476764199657e-05,
+      "loss": 1.0797,
+      "step": 2295
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.466796875,
+      "learning_rate": 6.597819850831899e-05,
+      "loss": 0.8466,
+      "step": 2300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.447265625,
+      "learning_rate": 6.612162937464142e-05,
+      "loss": 1.0248,
+      "step": 2305
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.447265625,
+      "learning_rate": 6.626506024096386e-05,
+      "loss": 0.9221,
+      "step": 2310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.494140625,
+      "learning_rate": 6.640849110728629e-05,
+      "loss": 0.8596,
+      "step": 2315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4296875,
+      "learning_rate": 6.655192197360873e-05,
+      "loss": 1.0407,
+      "step": 2320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.494140625,
+      "learning_rate": 6.669535283993115e-05,
+      "loss": 1.082,
+      "step": 2325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4453125,
+      "learning_rate": 6.683878370625359e-05,
+      "loss": 1.0036,
+      "step": 2330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4296875,
+      "learning_rate": 6.698221457257602e-05,
+      "loss": 0.9212,
+      "step": 2335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.44921875,
+      "learning_rate": 6.712564543889846e-05,
+      "loss": 1.2177,
+      "step": 2340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.419921875,
+      "learning_rate": 6.726907630522089e-05,
+      "loss": 0.9034,
+      "step": 2345
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.40625,
+      "learning_rate": 6.741250717154331e-05,
+      "loss": 0.8041,
+      "step": 2350
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.46484375,
+      "learning_rate": 6.755593803786575e-05,
+      "loss": 1.0537,
+      "step": 2355
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4375,
+      "learning_rate": 6.769936890418819e-05,
+      "loss": 1.0058,
+      "step": 2360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.78427997705106e-05,
+      "loss": 0.8552,
+      "step": 2365
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.443359375,
+      "learning_rate": 6.798623063683305e-05,
+      "loss": 0.9354,
+      "step": 2370
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.466796875,
+      "learning_rate": 6.812966150315549e-05,
+      "loss": 0.9095,
+      "step": 2375
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.41015625,
+      "learning_rate": 6.827309236947793e-05,
+      "loss": 0.9488,
+      "step": 2380
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.841652323580034e-05,
+      "loss": 0.8641,
+      "step": 2385
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.855995410212278e-05,
+      "loss": 1.0436,
+      "step": 2390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.474609375,
+      "learning_rate": 6.870338496844522e-05,
+      "loss": 0.9638,
+      "step": 2395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.43359375,
+      "learning_rate": 6.884681583476765e-05,
+      "loss": 0.9359,
+      "step": 2400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.43359375,
+      "learning_rate": 6.899024670109007e-05,
+      "loss": 1.011,
+      "step": 2405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.470703125,
+      "learning_rate": 6.913367756741251e-05,
+      "loss": 1.0113,
+      "step": 2410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4296875,
+      "learning_rate": 6.927710843373494e-05,
+      "loss": 1.056,
+      "step": 2415
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.474609375,
+      "learning_rate": 6.942053930005738e-05,
+      "loss": 0.9155,
+      "step": 2420
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.95639701663798e-05,
+      "loss": 0.9205,
+      "step": 2425
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.423828125,
+      "learning_rate": 6.970740103270223e-05,
+      "loss": 0.9412,
+      "step": 2430
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.455078125,
+      "learning_rate": 6.985083189902467e-05,
+      "loss": 1.0245,
+      "step": 2435
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.421875,
+      "learning_rate": 6.999426276534711e-05,
+      "loss": 0.9259,
+      "step": 2440
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.392578125,
+      "learning_rate": 7.013769363166954e-05,
+      "loss": 0.8526,
+      "step": 2445
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.447265625,
+      "learning_rate": 7.028112449799197e-05,
+      "loss": 0.8853,
+      "step": 2450
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.435546875,
+      "learning_rate": 7.042455536431441e-05,
+      "loss": 0.9233,
+      "step": 2455
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.056798623063683e-05,
+      "loss": 0.8997,
+      "step": 2460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.486328125,
+      "learning_rate": 7.071141709695926e-05,
+      "loss": 1.0118,
+      "step": 2465
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.08548479632817e-05,
+      "loss": 1.0767,
+      "step": 2470
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.099827882960414e-05,
+      "loss": 0.9788,
+      "step": 2475
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.46484375,
+      "learning_rate": 7.114170969592657e-05,
+      "loss": 0.9593,
+      "step": 2480
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.466796875,
+      "learning_rate": 7.1285140562249e-05,
+      "loss": 0.9277,
+      "step": 2485
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4453125,
+      "learning_rate": 7.142857142857143e-05,
+      "loss": 1.0196,
+      "step": 2490
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.486328125,
+      "learning_rate": 7.157200229489387e-05,
+      "loss": 0.9658,
+      "step": 2495
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.44921875,
+      "learning_rate": 7.17154331612163e-05,
+      "loss": 0.986,
+      "step": 2500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.412109375,
+      "learning_rate": 7.185886402753873e-05,
+      "loss": 1.0481,
+      "step": 2505
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.470703125,
+      "learning_rate": 7.200229489386117e-05,
+      "loss": 1.1599,
+      "step": 2510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.21457257601836e-05,
+      "loss": 0.8659,
+      "step": 2515
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4765625,
+      "learning_rate": 7.228915662650602e-05,
+      "loss": 1.0566,
+      "step": 2520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.41796875,
+      "learning_rate": 7.243258749282846e-05,
+      "loss": 0.8413,
+      "step": 2525
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.257601835915089e-05,
+      "loss": 0.9329,
+      "step": 2530
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.484375,
+      "learning_rate": 7.271944922547333e-05,
+      "loss": 0.9888,
+      "step": 2535
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.286288009179575e-05,
+      "loss": 0.8587,
+      "step": 2540
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.453125,
+      "learning_rate": 7.30063109581182e-05,
+      "loss": 0.962,
+      "step": 2545
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.447265625,
+      "learning_rate": 7.314974182444062e-05,
+      "loss": 0.9628,
+      "step": 2550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.392578125,
+      "learning_rate": 7.329317269076306e-05,
+      "loss": 1.0271,
+      "step": 2555
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.343660355708549e-05,
+      "loss": 0.8804,
+      "step": 2560
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.380859375,
+      "learning_rate": 7.358003442340791e-05,
+      "loss": 0.8014,
+      "step": 2565
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.372346528973035e-05,
+      "loss": 0.9217,
+      "step": 2570
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.451171875,
+      "learning_rate": 7.38668961560528e-05,
+      "loss": 0.7977,
+      "step": 2575
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.484375,
+      "learning_rate": 7.401032702237521e-05,
+      "loss": 1.0367,
+      "step": 2580
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4921875,
+      "learning_rate": 7.415375788869765e-05,
+      "loss": 0.9485,
+      "step": 2585
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.439453125,
+      "learning_rate": 7.429718875502009e-05,
+      "loss": 1.1512,
+      "step": 2590
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.41015625,
+      "learning_rate": 7.444061962134251e-05,
+      "loss": 0.7874,
+      "step": 2595
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.43359375,
+      "learning_rate": 7.458405048766494e-05,
+      "loss": 1.0102,
+      "step": 2600
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.472748135398738e-05,
+      "loss": 1.0031,
+      "step": 2605
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.447265625,
+      "learning_rate": 7.487091222030982e-05,
+      "loss": 0.9819,
+      "step": 2610
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.46875,
+      "learning_rate": 7.501434308663225e-05,
+      "loss": 1.0839,
+      "step": 2615
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.515777395295467e-05,
+      "loss": 1.0095,
+      "step": 2620
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.44140625,
+      "learning_rate": 7.530120481927712e-05,
+      "loss": 0.87,
+      "step": 2625
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4609375,
+      "learning_rate": 7.544463568559954e-05,
+      "loss": 0.9675,
+      "step": 2630
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.423828125,
+      "learning_rate": 7.558806655192198e-05,
+      "loss": 0.9699,
+      "step": 2635
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.439453125,
+      "learning_rate": 7.573149741824441e-05,
+      "loss": 1.1126,
+      "step": 2640
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.587492828456684e-05,
+      "loss": 0.8104,
+      "step": 2645
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.443359375,
+      "learning_rate": 7.601835915088928e-05,
+      "loss": 0.951,
+      "step": 2650
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.39453125,
+      "learning_rate": 7.616179001721172e-05,
+      "loss": 1.0582,
+      "step": 2655
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4765625,
+      "learning_rate": 7.630522088353414e-05,
+      "loss": 1.0965,
+      "step": 2660
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.439453125,
+      "learning_rate": 7.644865174985657e-05,
+      "loss": 0.9515,
+      "step": 2665
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.45703125,
+      "learning_rate": 7.659208261617901e-05,
+      "loss": 1.0292,
+      "step": 2670
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.443359375,
+      "learning_rate": 7.673551348250144e-05,
+      "loss": 1.0915,
+      "step": 2675
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4140625,
+      "learning_rate": 7.687894434882386e-05,
+      "loss": 1.055,
+      "step": 2680
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.70223752151463e-05,
+      "loss": 1.0428,
+      "step": 2685
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.716580608146874e-05,
+      "loss": 0.9814,
+      "step": 2690
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.470703125,
+      "learning_rate": 7.730923694779117e-05,
+      "loss": 0.8716,
+      "step": 2695
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.74526678141136e-05,
+      "loss": 0.8861,
+      "step": 2700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.466796875,
+      "learning_rate": 7.759609868043604e-05,
+      "loss": 1.0297,
+      "step": 2705
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.435546875,
+      "learning_rate": 7.773952954675848e-05,
+      "loss": 0.9579,
+      "step": 2710
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.482421875,
+      "learning_rate": 7.78829604130809e-05,
+      "loss": 0.9681,
+      "step": 2715
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.802639127940333e-05,
+      "loss": 0.829,
+      "step": 2720
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.388671875,
+      "learning_rate": 7.816982214572577e-05,
+      "loss": 0.9794,
+      "step": 2725
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.46484375,
+      "learning_rate": 7.83132530120482e-05,
+      "loss": 1.0554,
+      "step": 2730
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.41796875,
+      "learning_rate": 7.845668387837064e-05,
+      "loss": 0.8973,
+      "step": 2735
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.860011474469306e-05,
+      "loss": 1.0593,
+      "step": 2740
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4453125,
+      "learning_rate": 7.874354561101549e-05,
+      "loss": 1.0436,
+      "step": 2745
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.462890625,
+      "learning_rate": 7.888697647733793e-05,
+      "loss": 1.091,
+      "step": 2750
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.443359375,
+      "learning_rate": 7.903040734366036e-05,
+      "loss": 1.1119,
+      "step": 2755
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.43359375,
+      "learning_rate": 7.917383820998278e-05,
+      "loss": 1.0251,
+      "step": 2760
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4453125,
+      "learning_rate": 7.931726907630522e-05,
+      "loss": 0.888,
+      "step": 2765
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.458984375,
+      "learning_rate": 7.946069994262766e-05,
+      "loss": 0.9257,
+      "step": 2770
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.48046875,
+      "learning_rate": 7.960413080895009e-05,
+      "loss": 0.997,
+      "step": 2775
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.376953125,
+      "learning_rate": 7.974756167527252e-05,
+      "loss": 0.9598,
+      "step": 2780
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.443359375,
+      "learning_rate": 7.989099254159496e-05,
+      "loss": 0.8065,
+      "step": 2785
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.00344234079174e-05,
+      "loss": 0.9678,
+      "step": 2790
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.38671875,
+      "learning_rate": 8.017785427423982e-05,
+      "loss": 0.9295,
+      "step": 2795
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.45703125,
+      "learning_rate": 8.032128514056225e-05,
+      "loss": 0.8496,
+      "step": 2800
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.453125,
+      "learning_rate": 8.046471600688469e-05,
+      "loss": 0.9453,
+      "step": 2805
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.060814687320712e-05,
+      "loss": 0.9759,
+      "step": 2810
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.44140625,
+      "learning_rate": 8.075157773952954e-05,
+      "loss": 0.8887,
+      "step": 2815
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.408203125,
+      "learning_rate": 8.089500860585198e-05,
+      "loss": 0.9728,
+      "step": 2820
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.40625,
+      "learning_rate": 8.103843947217442e-05,
+      "loss": 0.8913,
+      "step": 2825
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.365234375,
+      "learning_rate": 8.118187033849685e-05,
+      "loss": 0.9973,
+      "step": 2830
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.132530120481928e-05,
+      "loss": 1.0904,
+      "step": 2835
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.146873207114172e-05,
+      "loss": 0.8934,
+      "step": 2840
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.458984375,
+      "learning_rate": 8.161216293746414e-05,
+      "loss": 1.0938,
+      "step": 2845
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.175559380378658e-05,
+      "loss": 0.9017,
+      "step": 2850
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.470703125,
+      "learning_rate": 8.189902467010901e-05,
+      "loss": 1.0773,
+      "step": 2855
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.204245553643144e-05,
+      "loss": 0.9509,
+      "step": 2860
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.458984375,
+      "learning_rate": 8.218588640275388e-05,
+      "loss": 0.8704,
+      "step": 2865
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.416015625,
+      "learning_rate": 8.232931726907632e-05,
+      "loss": 0.9692,
+      "step": 2870
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.247274813539873e-05,
+      "loss": 1.0793,
+      "step": 2875
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.46875,
+      "learning_rate": 8.261617900172117e-05,
+      "loss": 0.8818,
+      "step": 2880
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.392578125,
+      "learning_rate": 8.275960986804361e-05,
+      "loss": 1.0164,
+      "step": 2885
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.290304073436605e-05,
+      "loss": 0.9334,
+      "step": 2890
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.380859375,
+      "learning_rate": 8.304647160068846e-05,
+      "loss": 1.1717,
+      "step": 2895
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.451171875,
+      "learning_rate": 8.31899024670109e-05,
+      "loss": 1.071,
+      "step": 2900
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.482421875,
+      "learning_rate": 8.333333333333334e-05,
+      "loss": 0.92,
+      "step": 2905
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.458984375,
+      "learning_rate": 8.347676419965577e-05,
+      "loss": 1.0631,
+      "step": 2910
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.36201950659782e-05,
+      "loss": 0.9019,
+      "step": 2915
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.376362593230064e-05,
+      "loss": 0.8708,
+      "step": 2920
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.390705679862306e-05,
+      "loss": 0.9173,
+      "step": 2925
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.49609375,
+      "learning_rate": 8.40504876649455e-05,
+      "loss": 1.0932,
+      "step": 2930
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.46875,
+      "learning_rate": 8.419391853126793e-05,
+      "loss": 0.9486,
+      "step": 2935
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.433734939759037e-05,
+      "loss": 1.0511,
+      "step": 2940
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.451171875,
+      "learning_rate": 8.44807802639128e-05,
+      "loss": 0.9129,
+      "step": 2945
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3828125,
+      "learning_rate": 8.462421113023524e-05,
+      "loss": 1.0939,
+      "step": 2950
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.476764199655766e-05,
+      "loss": 0.8591,
+      "step": 2955
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.384765625,
+      "learning_rate": 8.491107286288009e-05,
+      "loss": 1.0119,
+      "step": 2960
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.451171875,
+      "learning_rate": 8.505450372920253e-05,
+      "loss": 1.0331,
+      "step": 2965
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.462890625,
+      "learning_rate": 8.519793459552496e-05,
+      "loss": 1.0119,
+      "step": 2970
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.534136546184738e-05,
+      "loss": 0.9088,
+      "step": 2975
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.548479632816982e-05,
+      "loss": 0.9238,
+      "step": 2980
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.48828125,
+      "learning_rate": 8.562822719449226e-05,
+      "loss": 0.9416,
+      "step": 2985
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.577165806081469e-05,
+      "loss": 1.0492,
+      "step": 2990
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.591508892713712e-05,
+      "loss": 1.008,
+      "step": 2995
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.44921875,
+      "learning_rate": 8.605851979345956e-05,
+      "loss": 1.0736,
+      "step": 3000
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.6201950659782e-05,
+      "loss": 0.8645,
+      "step": 3005
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.404296875,
+      "learning_rate": 8.634538152610442e-05,
+      "loss": 1.0537,
+      "step": 3010
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.648881239242685e-05,
+      "loss": 1.0242,
+      "step": 3015
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.663224325874929e-05,
+      "loss": 1.052,
+      "step": 3020
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.484375,
+      "learning_rate": 8.677567412507172e-05,
+      "loss": 1.1053,
+      "step": 3025
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4609375,
+      "learning_rate": 8.691910499139414e-05,
+      "loss": 1.0429,
+      "step": 3030
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.706253585771659e-05,
+      "loss": 0.9297,
+      "step": 3035
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.720596672403901e-05,
+      "loss": 0.9675,
+      "step": 3040
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.44921875,
+      "learning_rate": 8.734939759036145e-05,
+      "loss": 1.048,
+      "step": 3045
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.453125,
+      "learning_rate": 8.749282845668388e-05,
+      "loss": 0.8463,
+      "step": 3050
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.369140625,
+      "learning_rate": 8.763625932300632e-05,
+      "loss": 0.8302,
+      "step": 3055
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.484375,
+      "learning_rate": 8.777969018932875e-05,
+      "loss": 0.9625,
+      "step": 3060
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.792312105565119e-05,
+      "loss": 1.0129,
+      "step": 3065
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.806655192197361e-05,
+      "loss": 0.8039,
+      "step": 3070
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.41015625,
+      "learning_rate": 8.820998278829604e-05,
+      "loss": 0.9156,
+      "step": 3075
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.421875,
+      "learning_rate": 8.835341365461848e-05,
+      "loss": 1.033,
+      "step": 3080
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.44140625,
+      "learning_rate": 8.849684452094092e-05,
+      "loss": 1.03,
+      "step": 3085
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.864027538726333e-05,
+      "loss": 1.0026,
+      "step": 3090
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.458984375,
+      "learning_rate": 8.878370625358577e-05,
+      "loss": 1.0639,
+      "step": 3095
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4609375,
+      "learning_rate": 8.892713711990821e-05,
+      "loss": 0.9634,
+      "step": 3100
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.490234375,
+      "learning_rate": 8.907056798623065e-05,
+      "loss": 0.9578,
+      "step": 3105
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.48828125,
+      "learning_rate": 8.921399885255307e-05,
+      "loss": 0.9326,
+      "step": 3110
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.421875,
+      "learning_rate": 8.93574297188755e-05,
+      "loss": 0.9253,
+      "step": 3115
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.950086058519795e-05,
+      "loss": 0.918,
+      "step": 3120
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.964429145152037e-05,
+      "loss": 1.0224,
+      "step": 3125
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.388671875,
+      "learning_rate": 8.97877223178428e-05,
+      "loss": 0.898,
+      "step": 3130
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.404296875,
+      "learning_rate": 8.993115318416524e-05,
+      "loss": 1.1698,
+      "step": 3135
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.007458405048767e-05,
+      "loss": 0.9055,
+      "step": 3140
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4609375,
+      "learning_rate": 9.02180149168101e-05,
+      "loss": 1.0444,
+      "step": 3145
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.439453125,
+      "learning_rate": 9.036144578313253e-05,
+      "loss": 1.0148,
+      "step": 3150
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.050487664945496e-05,
+      "loss": 0.8328,
+      "step": 3155
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.06483075157774e-05,
+      "loss": 0.9466,
+      "step": 3160
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.40625,
+      "learning_rate": 9.079173838209984e-05,
+      "loss": 0.9638,
+      "step": 3165
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.093516924842227e-05,
+      "loss": 0.9799,
+      "step": 3170
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.107860011474469e-05,
+      "loss": 1.0191,
+      "step": 3175
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45703125,
+      "learning_rate": 9.122203098106713e-05,
+      "loss": 1.0255,
+      "step": 3180
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.136546184738956e-05,
+      "loss": 0.9925,
+      "step": 3185
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.150889271371199e-05,
+      "loss": 0.8852,
+      "step": 3190
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45703125,
+      "learning_rate": 9.165232358003443e-05,
+      "loss": 1.1207,
+      "step": 3195
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.179575444635687e-05,
+      "loss": 1.052,
+      "step": 3200
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.390625,
+      "learning_rate": 9.193918531267929e-05,
+      "loss": 0.918,
+      "step": 3205
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.208261617900172e-05,
+      "loss": 0.9535,
+      "step": 3210
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.447265625,
+      "learning_rate": 9.222604704532416e-05,
+      "loss": 1.1085,
+      "step": 3215
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9.23694779116466e-05,
+      "loss": 1.0227,
+      "step": 3220
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.251290877796903e-05,
+      "loss": 0.9191,
+      "step": 3225
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.490234375,
+      "learning_rate": 9.265633964429145e-05,
+      "loss": 1.0532,
+      "step": 3230
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.27997705106139e-05,
+      "loss": 0.9121,
+      "step": 3235
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.478515625,
+      "learning_rate": 9.294320137693632e-05,
+      "loss": 1.0063,
+      "step": 3240
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.408203125,
+      "learning_rate": 9.308663224325875e-05,
+      "loss": 1.1222,
+      "step": 3245
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.390625,
+      "learning_rate": 9.323006310958119e-05,
+      "loss": 0.8751,
+      "step": 3250
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.337349397590361e-05,
+      "loss": 0.9919,
+      "step": 3255
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.462890625,
+      "learning_rate": 9.351692484222605e-05,
+      "loss": 0.965,
+      "step": 3260
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.366035570854848e-05,
+      "loss": 0.9367,
+      "step": 3265
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.439453125,
+      "learning_rate": 9.380378657487092e-05,
+      "loss": 0.9339,
+      "step": 3270
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44921875,
+      "learning_rate": 9.394721744119335e-05,
+      "loss": 1.1373,
+      "step": 3275
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.409064830751579e-05,
+      "loss": 0.9003,
+      "step": 3280
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.48046875,
+      "learning_rate": 9.423407917383821e-05,
+      "loss": 1.0328,
+      "step": 3285
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.455078125,
+      "learning_rate": 9.437751004016064e-05,
+      "loss": 0.9807,
+      "step": 3290
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.45703125,
+      "learning_rate": 9.452094090648308e-05,
+      "loss": 1.1077,
+      "step": 3295
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.490234375,
+      "learning_rate": 9.466437177280552e-05,
+      "loss": 1.0933,
+      "step": 3300
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44921875,
+      "learning_rate": 9.480780263912795e-05,
+      "loss": 0.9658,
+      "step": 3305
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.495123350545037e-05,
+      "loss": 0.9943,
+      "step": 3310
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.408203125,
+      "learning_rate": 9.509466437177281e-05,
+      "loss": 1.0197,
+      "step": 3315
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.523809523809524e-05,
+      "loss": 1.0082,
+      "step": 3320
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.538152610441767e-05,
+      "loss": 0.9112,
+      "step": 3325
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46484375,
+      "learning_rate": 9.552495697074011e-05,
+      "loss": 0.9961,
+      "step": 3330
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.566838783706255e-05,
+      "loss": 1.0198,
+      "step": 3335
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.474609375,
+      "learning_rate": 9.581181870338497e-05,
+      "loss": 0.9685,
+      "step": 3340
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.474609375,
+      "learning_rate": 9.59552495697074e-05,
+      "loss": 1.0192,
+      "step": 3345
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.408203125,
+      "learning_rate": 9.609868043602984e-05,
+      "loss": 1.1994,
+      "step": 3350
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.624211130235227e-05,
+      "loss": 0.954,
+      "step": 3355
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44921875,
+      "learning_rate": 9.638554216867471e-05,
+      "loss": 0.8963,
+      "step": 3360
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.652897303499713e-05,
+      "loss": 0.9157,
+      "step": 3365
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.667240390131956e-05,
+      "loss": 0.8979,
+      "step": 3370
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.6815834767642e-05,
+      "loss": 1.1164,
+      "step": 3375
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.447265625,
+      "learning_rate": 9.695926563396444e-05,
+      "loss": 1.0539,
+      "step": 3380
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.427734375,
+      "learning_rate": 9.710269650028687e-05,
+      "loss": 0.9551,
+      "step": 3385
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.72461273666093e-05,
+      "loss": 1.0195,
+      "step": 3390
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.738955823293173e-05,
+      "loss": 0.9593,
+      "step": 3395
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.753298909925417e-05,
+      "loss": 0.8717,
+      "step": 3400
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.767641996557659e-05,
+      "loss": 0.8758,
+      "step": 3405
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.427734375,
+      "learning_rate": 9.781985083189903e-05,
+      "loss": 0.9862,
+      "step": 3410
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9.796328169822147e-05,
+      "loss": 0.957,
+      "step": 3415
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.81067125645439e-05,
+      "loss": 0.894,
+      "step": 3420
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.825014343086632e-05,
+      "loss": 0.9824,
+      "step": 3425
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9.839357429718876e-05,
+      "loss": 1.0242,
+      "step": 3430
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.853700516351119e-05,
+      "loss": 0.8554,
+      "step": 3435
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.466796875,
+      "learning_rate": 9.868043602983363e-05,
+      "loss": 1.0288,
+      "step": 3440
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5,
+      "learning_rate": 9.882386689615606e-05,
+      "loss": 1.0798,
+      "step": 3445
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.89672977624785e-05,
+      "loss": 0.8537,
+      "step": 3450
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.443359375,
+      "learning_rate": 9.911072862880092e-05,
+      "loss": 0.9823,
+      "step": 3455
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.38671875,
+      "learning_rate": 9.925415949512336e-05,
+      "loss": 0.9169,
+      "step": 3460
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.939759036144579e-05,
+      "loss": 0.8519,
+      "step": 3465
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.954102122776822e-05,
+      "loss": 0.9385,
+      "step": 3470
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.376953125,
+      "learning_rate": 9.968445209409066e-05,
+      "loss": 0.936,
+      "step": 3475
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.982788296041308e-05,
+      "loss": 0.982,
+      "step": 3480
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.453125,
+      "learning_rate": 9.997131382673551e-05,
+      "loss": 0.8798,
+      "step": 3485
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00010011474469305795,
+      "loss": 1.1016,
+      "step": 3490
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00010025817555938038,
+      "loss": 1.2804,
+      "step": 3495
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00010040160642570282,
+      "loss": 1.0197,
+      "step": 3500
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00010054503729202524,
+      "loss": 1.0637,
+      "step": 3505
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00010068846815834767,
+      "loss": 0.9281,
+      "step": 3510
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00010083189902467012,
+      "loss": 0.9208,
+      "step": 3515
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00010097532989099255,
+      "loss": 0.9596,
+      "step": 3520
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00010111876075731499,
+      "loss": 0.917,
+      "step": 3525
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00010126219162363742,
+      "loss": 0.9712,
+      "step": 3530
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00010140562248995984,
+      "loss": 1.081,
+      "step": 3535
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00010154905335628228,
+      "loss": 1.0385,
+      "step": 3540
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00010169248422260471,
+      "loss": 0.9869,
+      "step": 3545
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00010183591508892714,
+      "loss": 0.8981,
+      "step": 3550
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00010197934595524959,
+      "loss": 0.8882,
+      "step": 3555
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.000102122776821572,
+      "loss": 0.8624,
+      "step": 3560
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00010226620768789443,
+      "loss": 0.9483,
+      "step": 3565
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00010240963855421688,
+      "loss": 0.9354,
+      "step": 3570
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00010255306942053931,
+      "loss": 0.9673,
+      "step": 3575
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00010269650028686174,
+      "loss": 0.9585,
+      "step": 3580
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00010283993115318418,
+      "loss": 1.0179,
+      "step": 3585
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001029833620195066,
+      "loss": 0.9306,
+      "step": 3590
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00010312679288582903,
+      "loss": 1.0562,
+      "step": 3595
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00010327022375215147,
+      "loss": 0.8476,
+      "step": 3600
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0001034136546184739,
+      "loss": 1.0539,
+      "step": 3605
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00010355708548479632,
+      "loss": 0.9696,
+      "step": 3610
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00010370051635111878,
+      "loss": 1.0788,
+      "step": 3615
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00010384394721744119,
+      "loss": 1.1724,
+      "step": 3620
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00010398737808376362,
+      "loss": 1.0326,
+      "step": 3625
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00010413080895008607,
+      "loss": 0.9767,
+      "step": 3630
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001042742398164085,
+      "loss": 0.9976,
+      "step": 3635
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00010441767068273094,
+      "loss": 0.898,
+      "step": 3640
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00010456110154905336,
+      "loss": 0.8513,
+      "step": 3645
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00010470453241537579,
+      "loss": 1.0365,
+      "step": 3650
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00010484796328169823,
+      "loss": 1.0731,
+      "step": 3655
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00010499139414802066,
+      "loss": 0.9298,
+      "step": 3660
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00010513482501434308,
+      "loss": 1.0778,
+      "step": 3665
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00010527825588066554,
+      "loss": 0.9948,
+      "step": 3670
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010542168674698796,
+      "loss": 0.9008,
+      "step": 3675
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00010556511761331038,
+      "loss": 0.9447,
+      "step": 3680
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00010570854847963283,
+      "loss": 0.816,
+      "step": 3685
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00010585197934595526,
+      "loss": 0.9181,
+      "step": 3690
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00010599541021227768,
+      "loss": 0.9687,
+      "step": 3695
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00010613884107860012,
+      "loss": 0.8236,
+      "step": 3700
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00010628227194492255,
+      "loss": 0.862,
+      "step": 3705
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00010642570281124498,
+      "loss": 1.0777,
+      "step": 3710
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00010656913367756742,
+      "loss": 0.8843,
+      "step": 3715
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00010671256454388984,
+      "loss": 0.8764,
+      "step": 3720
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00010685599541021227,
+      "loss": 0.9723,
+      "step": 3725
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00010699942627653472,
+      "loss": 0.9971,
+      "step": 3730
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00010714285714285715,
+      "loss": 1.03,
+      "step": 3735
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00010728628800917956,
+      "loss": 1.0012,
+      "step": 3740
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00010742971887550202,
+      "loss": 0.9739,
+      "step": 3745
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00010757314974182444,
+      "loss": 1.0166,
+      "step": 3750
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00010771658060814688,
+      "loss": 1.0096,
+      "step": 3755
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00010786001147446931,
+      "loss": 0.9461,
+      "step": 3760
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00010800344234079174,
+      "loss": 0.802,
+      "step": 3765
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00010814687320711419,
+      "loss": 1.0664,
+      "step": 3770
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0001082903040734366,
+      "loss": 0.9477,
+      "step": 3775
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00010843373493975903,
+      "loss": 1.0207,
+      "step": 3780
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00010857716580608148,
+      "loss": 0.979,
+      "step": 3785
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00010872059667240391,
+      "loss": 0.896,
+      "step": 3790
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00010886402753872634,
+      "loss": 1.0602,
+      "step": 3795
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00010900745840504878,
+      "loss": 0.8848,
+      "step": 3800
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001091508892713712,
+      "loss": 0.9814,
+      "step": 3805
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00010929432013769363,
+      "loss": 0.8746,
+      "step": 3810
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00010943775100401607,
+      "loss": 0.9627,
+      "step": 3815
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001095811818703385,
+      "loss": 0.943,
+      "step": 3820
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00010972461273666092,
+      "loss": 0.9082,
+      "step": 3825
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00010986804360298338,
+      "loss": 1.1245,
+      "step": 3830
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00011001147446930579,
+      "loss": 1.0683,
+      "step": 3835
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011015490533562822,
+      "loss": 0.9948,
+      "step": 3840
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00011029833620195067,
+      "loss": 0.933,
+      "step": 3845
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0001104417670682731,
+      "loss": 1.0216,
+      "step": 3850
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00011058519793459554,
+      "loss": 0.999,
+      "step": 3855
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00011072862880091797,
+      "loss": 1.0413,
+      "step": 3860
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00011087205966724039,
+      "loss": 1.0061,
+      "step": 3865
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00011101549053356283,
+      "loss": 0.8976,
+      "step": 3870
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00011115892139988526,
+      "loss": 0.9141,
+      "step": 3875
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00011130235226620769,
+      "loss": 1.0766,
+      "step": 3880
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00011144578313253014,
+      "loss": 0.9397,
+      "step": 3885
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00011158921399885257,
+      "loss": 1.0154,
+      "step": 3890
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00011173264486517498,
+      "loss": 0.9654,
+      "step": 3895
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00011187607573149743,
+      "loss": 1.0002,
+      "step": 3900
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00011201950659781986,
+      "loss": 0.9276,
+      "step": 3905
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00011216293746414229,
+      "loss": 0.9197,
+      "step": 3910
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00011230636833046473,
+      "loss": 1.0522,
+      "step": 3915
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00011244979919678715,
+      "loss": 0.9447,
+      "step": 3920
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00011259323006310958,
+      "loss": 0.9358,
+      "step": 3925
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00011273666092943202,
+      "loss": 1.0351,
+      "step": 3930
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00011288009179575445,
+      "loss": 0.9158,
+      "step": 3935
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00011302352266207687,
+      "loss": 0.9975,
+      "step": 3940
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00011316695352839933,
+      "loss": 0.9493,
+      "step": 3945
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00011331038439472175,
+      "loss": 1.0005,
+      "step": 3950
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00011345381526104417,
+      "loss": 1.1862,
+      "step": 3955
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00011359724612736662,
+      "loss": 1.122,
+      "step": 3960
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00011374067699368905,
+      "loss": 1.0702,
+      "step": 3965
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00011388410786001149,
+      "loss": 0.8827,
+      "step": 3970
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00011402753872633391,
+      "loss": 0.8097,
+      "step": 3975
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00011417096959265634,
+      "loss": 0.935,
+      "step": 3980
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00011431440045897879,
+      "loss": 0.9727,
+      "step": 3985
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0001144578313253012,
+      "loss": 0.9867,
+      "step": 3990
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00011460126219162363,
+      "loss": 1.0205,
+      "step": 3995
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00011474469305794609,
+      "loss": 0.9801,
+      "step": 4000
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011488812392426851,
+      "loss": 0.9266,
+      "step": 4005
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00011503155479059094,
+      "loss": 0.9448,
+      "step": 4010
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00011517498565691338,
+      "loss": 0.9925,
+      "step": 4015
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001153184165232358,
+      "loss": 1.0278,
+      "step": 4020
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00011546184738955823,
+      "loss": 0.942,
+      "step": 4025
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00011560527825588067,
+      "loss": 0.9767,
+      "step": 4030
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001157487091222031,
+      "loss": 0.9365,
+      "step": 4035
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00011589213998852553,
+      "loss": 0.9618,
+      "step": 4040
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00011603557085484798,
+      "loss": 0.8656,
+      "step": 4045
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001161790017211704,
+      "loss": 0.8396,
+      "step": 4050
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00011632243258749282,
+      "loss": 0.9391,
+      "step": 4055
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00011646586345381527,
+      "loss": 0.9302,
+      "step": 4060
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001166092943201377,
+      "loss": 1.0018,
+      "step": 4065
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00011675272518646013,
+      "loss": 0.9767,
+      "step": 4070
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00011689615605278257,
+      "loss": 0.9765,
+      "step": 4075
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.000117039586919105,
+      "loss": 1.0398,
+      "step": 4080
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00011718301778542743,
+      "loss": 0.8486,
+      "step": 4085
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00011732644865174986,
+      "loss": 0.8443,
+      "step": 4090
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00011746987951807229,
+      "loss": 0.9544,
+      "step": 4095
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00011761331038439474,
+      "loss": 1.1874,
+      "step": 4100
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00011775674125071717,
+      "loss": 1.1332,
+      "step": 4105
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00011790017211703958,
+      "loss": 0.9719,
+      "step": 4110
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00011804360298336203,
+      "loss": 0.909,
+      "step": 4115
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00011818703384968446,
+      "loss": 1.0488,
+      "step": 4120
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00011833046471600689,
+      "loss": 1.1675,
+      "step": 4125
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00011847389558232933,
+      "loss": 0.9763,
+      "step": 4130
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00011861732644865175,
+      "loss": 0.9416,
+      "step": 4135
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00011876075731497418,
+      "loss": 0.9931,
+      "step": 4140
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00011890418818129662,
+      "loss": 0.8678,
+      "step": 4145
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00011904761904761905,
+      "loss": 1.0077,
+      "step": 4150
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00011919104991394147,
+      "loss": 0.8943,
+      "step": 4155
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00011933448078026393,
+      "loss": 1.0197,
+      "step": 4160
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00011947791164658635,
+      "loss": 0.9257,
+      "step": 4165
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00011962134251290877,
+      "loss": 0.9073,
+      "step": 4170
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00011976477337923122,
+      "loss": 0.9657,
+      "step": 4175
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00011990820424555365,
+      "loss": 0.9148,
+      "step": 4180
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00012005163511187607,
+      "loss": 0.9801,
+      "step": 4185
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012019506597819851,
+      "loss": 1.0169,
+      "step": 4190
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00012033849684452094,
+      "loss": 0.9711,
+      "step": 4195
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0001204819277108434,
+      "loss": 1.0298,
+      "step": 4200
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00012062535857716581,
+      "loss": 0.9817,
+      "step": 4205
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00012076878944348823,
+      "loss": 0.8817,
+      "step": 4210
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00012091222030981069,
+      "loss": 0.8855,
+      "step": 4215
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00012105565117613311,
+      "loss": 1.0517,
+      "step": 4220
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00012119908204245554,
+      "loss": 0.9723,
+      "step": 4225
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00012134251290877798,
+      "loss": 0.8449,
+      "step": 4230
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00012148594377510041,
+      "loss": 0.9528,
+      "step": 4235
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00012162937464142283,
+      "loss": 1.0405,
+      "step": 4240
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00012177280550774528,
+      "loss": 0.9225,
+      "step": 4245
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0001219162363740677,
+      "loss": 0.8293,
+      "step": 4250
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00012205966724039013,
+      "loss": 0.8751,
+      "step": 4255
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00012220309810671257,
+      "loss": 1.0213,
+      "step": 4260
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.000122346528973035,
+      "loss": 0.9466,
+      "step": 4265
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00012248995983935742,
+      "loss": 0.8652,
+      "step": 4270
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012263339070567986,
+      "loss": 0.9067,
+      "step": 4275
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0001227768215720023,
+      "loss": 0.8576,
+      "step": 4280
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00012292025243832472,
+      "loss": 1.1626,
+      "step": 4285
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00012306368330464718,
+      "loss": 0.9502,
+      "step": 4290
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001232071141709696,
+      "loss": 1.006,
+      "step": 4295
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.375,
+      "learning_rate": 0.000123350545037292,
+      "loss": 0.9994,
+      "step": 4300
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012349397590361448,
+      "loss": 1.0563,
+      "step": 4305
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001236374067699369,
+      "loss": 0.9123,
+      "step": 4310
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012378083763625933,
+      "loss": 0.9732,
+      "step": 4315
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012392426850258177,
+      "loss": 0.9571,
+      "step": 4320
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00012406769936890418,
+      "loss": 0.9954,
+      "step": 4325
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00012421113023522665,
+      "loss": 1.0072,
+      "step": 4330
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00012435456110154906,
+      "loss": 0.9763,
+      "step": 4335
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00012449799196787148,
+      "loss": 1.0008,
+      "step": 4340
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00012464142283419394,
+      "loss": 1.0088,
+      "step": 4345
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00012478485370051636,
+      "loss": 1.1205,
+      "step": 4350
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00012492828456683877,
+      "loss": 0.991,
+      "step": 4355
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00012507171543316124,
+      "loss": 0.8901,
+      "step": 4360
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00012521514629948365,
+      "loss": 1.1043,
+      "step": 4365
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001253585771658061,
+      "loss": 1.0119,
+      "step": 4370
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00012550200803212853,
+      "loss": 1.0429,
+      "step": 4375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.015625,
+      "learning_rate": 0.00012564543889845094,
+      "loss": 1.0057,
+      "step": 4380
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00012578886976477338,
+      "loss": 0.9982,
+      "step": 4385
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00012593230063109582,
+      "loss": 0.9193,
+      "step": 4390
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00012607573149741824,
+      "loss": 0.9164,
+      "step": 4395
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00012621916236374068,
+      "loss": 1.0736,
+      "step": 4400
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00012636259323006312,
+      "loss": 1.0426,
+      "step": 4405
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00012650602409638556,
+      "loss": 0.9148,
+      "step": 4410
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.000126649454962708,
+      "loss": 1.1345,
+      "step": 4415
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001267928858290304,
+      "loss": 0.9112,
+      "step": 4420
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00012693631669535285,
+      "loss": 0.9311,
+      "step": 4425
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001270797475616753,
+      "loss": 1.0584,
+      "step": 4430
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001272231784279977,
+      "loss": 1.0231,
+      "step": 4435
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00012736660929432014,
+      "loss": 1.1119,
+      "step": 4440
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00012751004016064258,
+      "loss": 0.9418,
+      "step": 4445
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00012765347102696502,
+      "loss": 0.9222,
+      "step": 4450
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00012779690189328744,
+      "loss": 1.0592,
+      "step": 4455
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00012794033275960988,
+      "loss": 1.0891,
+      "step": 4460
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00012808376362593232,
+      "loss": 1.0498,
+      "step": 4465
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00012822719449225473,
+      "loss": 1.0035,
+      "step": 4470
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00012837062535857717,
+      "loss": 0.8811,
+      "step": 4475
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001285140562248996,
+      "loss": 1.0189,
+      "step": 4480
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00012865748709122202,
+      "loss": 1.0349,
+      "step": 4485
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012880091795754446,
+      "loss": 0.954,
+      "step": 4490
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001289443488238669,
+      "loss": 0.8879,
+      "step": 4495
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00012908777969018932,
+      "loss": 0.9027,
+      "step": 4500
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00012923121055651178,
+      "loss": 0.8992,
+      "step": 4505
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001293746414228342,
+      "loss": 1.3137,
+      "step": 4510
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0001295180722891566,
+      "loss": 1.0226,
+      "step": 4515
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00012966150315547908,
+      "loss": 0.9629,
+      "step": 4520
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001298049340218015,
+      "loss": 0.9313,
+      "step": 4525
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00012994836488812393,
+      "loss": 0.9058,
+      "step": 4530
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00013009179575444637,
+      "loss": 1.0217,
+      "step": 4535
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00013023522662076878,
+      "loss": 0.9383,
+      "step": 4540
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00013037865748709125,
+      "loss": 1.0084,
+      "step": 4545
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00013052208835341366,
+      "loss": 1.0465,
+      "step": 4550
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00013066551921973608,
+      "loss": 0.8307,
+      "step": 4555
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00013080895008605854,
+      "loss": 0.9803,
+      "step": 4560
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00013095238095238096,
+      "loss": 1.0079,
+      "step": 4565
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0001310958118187034,
+      "loss": 1.0062,
+      "step": 4570
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00013123924268502584,
+      "loss": 1.0667,
+      "step": 4575
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00013138267355134825,
+      "loss": 0.9888,
+      "step": 4580
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001315261044176707,
+      "loss": 1.1235,
+      "step": 4585
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00013166953528399313,
+      "loss": 0.998,
+      "step": 4590
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00013181296615031554,
+      "loss": 1.0042,
+      "step": 4595
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00013195639701663798,
+      "loss": 0.9648,
+      "step": 4600
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00013209982788296042,
+      "loss": 0.8943,
+      "step": 4605
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00013224325874928284,
+      "loss": 0.9704,
+      "step": 4610
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00013238668961560528,
+      "loss": 1.0426,
+      "step": 4615
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00013253012048192772,
+      "loss": 0.9591,
+      "step": 4620
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00013267355134825016,
+      "loss": 0.991,
+      "step": 4625
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00013281698221457257,
+      "loss": 0.971,
+      "step": 4630
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.000132960413080895,
+      "loss": 1.1203,
+      "step": 4635
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00013310384394721745,
+      "loss": 1.0527,
+      "step": 4640
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001332472748135399,
+      "loss": 1.057,
+      "step": 4645
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001333907056798623,
+      "loss": 0.889,
+      "step": 4650
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00013353413654618475,
+      "loss": 0.952,
+      "step": 4655
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00013367756741250719,
+      "loss": 0.9844,
+      "step": 4660
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00013382099827882963,
+      "loss": 0.899,
+      "step": 4665
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00013396442914515204,
+      "loss": 1.0083,
+      "step": 4670
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00013410786001147448,
+      "loss": 1.1103,
+      "step": 4675
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00013425129087779692,
+      "loss": 1.0983,
+      "step": 4680
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00013439472174411933,
+      "loss": 0.9257,
+      "step": 4685
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00013453815261044177,
+      "loss": 0.9193,
+      "step": 4690
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001346815834767642,
+      "loss": 0.9014,
+      "step": 4695
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00013482501434308663,
+      "loss": 0.9115,
+      "step": 4700
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00013496844520940907,
+      "loss": 0.9923,
+      "step": 4705
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001351118760757315,
+      "loss": 0.9589,
+      "step": 4710
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00013525530694205392,
+      "loss": 0.8925,
+      "step": 4715
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00013539873780837639,
+      "loss": 1.1727,
+      "step": 4720
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001355421686746988,
+      "loss": 1.0245,
+      "step": 4725
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0001356855995410212,
+      "loss": 0.9471,
+      "step": 4730
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00013582903040734368,
+      "loss": 0.8213,
+      "step": 4735
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001359724612736661,
+      "loss": 0.9894,
+      "step": 4740
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00013611589213998853,
+      "loss": 0.9985,
+      "step": 4745
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00013625932300631097,
+      "loss": 0.9631,
+      "step": 4750
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00013640275387263339,
+      "loss": 0.9722,
+      "step": 4755
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00013654618473895585,
+      "loss": 0.9535,
+      "step": 4760
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00013668961560527827,
+      "loss": 0.8753,
+      "step": 4765
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00013683304647160068,
+      "loss": 0.9648,
+      "step": 4770
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00013697647733792315,
+      "loss": 0.8829,
+      "step": 4775
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00013711990820424556,
+      "loss": 1.0206,
+      "step": 4780
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000137263339070568,
+      "loss": 0.9263,
+      "step": 4785
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00013740676993689044,
+      "loss": 0.9042,
+      "step": 4790
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00013755020080321285,
+      "loss": 0.9451,
+      "step": 4795
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001376936316695353,
+      "loss": 1.0014,
+      "step": 4800
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00013783706253585773,
+      "loss": 0.9152,
+      "step": 4805
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00013798049340218015,
+      "loss": 1.0523,
+      "step": 4810
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00013812392426850259,
+      "loss": 1.0063,
+      "step": 4815
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00013826735513482503,
+      "loss": 0.9768,
+      "step": 4820
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00013841078600114744,
+      "loss": 1.2043,
+      "step": 4825
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00013855421686746988,
+      "loss": 0.9456,
+      "step": 4830
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00013869764773379232,
+      "loss": 0.8694,
+      "step": 4835
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00013884107860011476,
+      "loss": 0.89,
+      "step": 4840
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00013898450946643717,
+      "loss": 0.9385,
+      "step": 4845
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001391279403327596,
+      "loss": 0.8744,
+      "step": 4850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00013927137119908205,
+      "loss": 0.9686,
+      "step": 4855
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00013941480206540447,
+      "loss": 0.9721,
+      "step": 4860
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001395582329317269,
+      "loss": 0.9944,
+      "step": 4865
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00013970166379804935,
+      "loss": 1.0323,
+      "step": 4870
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0001398450946643718,
+      "loss": 1.0255,
+      "step": 4875
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00013998852553069423,
+      "loss": 0.9475,
+      "step": 4880
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00014013195639701664,
+      "loss": 1.0105,
+      "step": 4885
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00014027538726333908,
+      "loss": 0.9109,
+      "step": 4890
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00014041881812966152,
+      "loss": 0.9437,
+      "step": 4895
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00014056224899598393,
+      "loss": 1.0219,
+      "step": 4900
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00014070567986230637,
+      "loss": 0.9312,
+      "step": 4905
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00014084911072862881,
+      "loss": 1.0337,
+      "step": 4910
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00014099254159495123,
+      "loss": 1.1155,
+      "step": 4915
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014113597246127367,
+      "loss": 1.0321,
+      "step": 4920
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0001412794033275961,
+      "loss": 0.9464,
+      "step": 4925
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00014142283419391852,
+      "loss": 1.0167,
+      "step": 4930
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.000141566265060241,
+      "loss": 1.1417,
+      "step": 4935
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001417096959265634,
+      "loss": 1.0258,
+      "step": 4940
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00014185312679288581,
+      "loss": 0.9996,
+      "step": 4945
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00014199655765920828,
+      "loss": 1.102,
+      "step": 4950
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001421399885255307,
+      "loss": 0.9078,
+      "step": 4955
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00014228341939185313,
+      "loss": 0.9108,
+      "step": 4960
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00014242685025817557,
+      "loss": 0.9866,
+      "step": 4965
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.000142570281124498,
+      "loss": 1.0493,
+      "step": 4970
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00014271371199082045,
+      "loss": 0.8932,
+      "step": 4975
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00014285714285714287,
+      "loss": 1.0283,
+      "step": 4980
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00014300057372346528,
+      "loss": 0.9218,
+      "step": 4985
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00014314400458978775,
+      "loss": 0.8918,
+      "step": 4990
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00014328743545611016,
+      "loss": 0.9241,
+      "step": 4995
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001434308663224326,
+      "loss": 1.0046,
+      "step": 5000
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00014357429718875504,
+      "loss": 0.9486,
+      "step": 5005
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00014371772805507745,
+      "loss": 0.8616,
+      "step": 5010
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001438611589213999,
+      "loss": 0.952,
+      "step": 5015
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00014400458978772233,
+      "loss": 0.867,
+      "step": 5020
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00014414802065404475,
+      "loss": 1.0847,
+      "step": 5025
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0001442914515203672,
+      "loss": 0.8939,
+      "step": 5030
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00014443488238668963,
+      "loss": 0.9756,
+      "step": 5035
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00014457831325301204,
+      "loss": 1.0356,
+      "step": 5040
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00014472174411933448,
+      "loss": 0.8602,
+      "step": 5045
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00014486517498565692,
+      "loss": 0.8666,
+      "step": 5050
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00014500860585197936,
+      "loss": 0.9651,
+      "step": 5055
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00014515203671830177,
+      "loss": 1.02,
+      "step": 5060
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00014529546758462422,
+      "loss": 1.0457,
+      "step": 5065
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00014543889845094666,
+      "loss": 1.1339,
+      "step": 5070
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00014558232931726907,
+      "loss": 0.8031,
+      "step": 5075
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001457257601835915,
+      "loss": 0.9769,
+      "step": 5080
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00014586919104991395,
+      "loss": 0.8861,
+      "step": 5085
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001460126219162364,
+      "loss": 0.9878,
+      "step": 5090
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00014615605278255883,
+      "loss": 0.904,
+      "step": 5095
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00014629948364888124,
+      "loss": 0.8937,
+      "step": 5100
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014644291451520368,
+      "loss": 1.0381,
+      "step": 5105
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00014658634538152612,
+      "loss": 1.0496,
+      "step": 5110
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00014672977624784854,
+      "loss": 1.0499,
+      "step": 5115
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00014687320711417098,
+      "loss": 1.0203,
+      "step": 5120
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00014701663798049342,
+      "loss": 1.0739,
+      "step": 5125
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00014716006884681583,
+      "loss": 0.9814,
+      "step": 5130
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00014730349971313827,
+      "loss": 1.0477,
+      "step": 5135
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001474469305794607,
+      "loss": 0.9457,
+      "step": 5140
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00014759036144578312,
+      "loss": 0.869,
+      "step": 5145
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001477337923121056,
+      "loss": 0.9091,
+      "step": 5150
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.000147877223178428,
+      "loss": 0.8955,
+      "step": 5155
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00014802065404475042,
+      "loss": 0.9003,
+      "step": 5160
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00014816408491107288,
+      "loss": 0.8217,
+      "step": 5165
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0001483075157773953,
+      "loss": 0.9767,
+      "step": 5170
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00014845094664371774,
+      "loss": 1.0299,
+      "step": 5175
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00014859437751004018,
+      "loss": 1.0436,
+      "step": 5180
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0001487378083763626,
+      "loss": 0.9742,
+      "step": 5185
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00014888123924268503,
+      "loss": 0.9181,
+      "step": 5190
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014902467010900747,
+      "loss": 0.9368,
+      "step": 5195
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00014916810097532988,
+      "loss": 0.9426,
+      "step": 5200
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00014931153184165235,
+      "loss": 1.0936,
+      "step": 5205
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00014945496270797476,
+      "loss": 1.0672,
+      "step": 5210
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001495983935742972,
+      "loss": 0.9419,
+      "step": 5215
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00014974182444061964,
+      "loss": 0.9147,
+      "step": 5220
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00014988525530694206,
+      "loss": 1.168,
+      "step": 5225
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0001500286861732645,
+      "loss": 1.0776,
+      "step": 5230
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00015017211703958694,
+      "loss": 0.8433,
+      "step": 5235
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00015031554790590935,
+      "loss": 0.959,
+      "step": 5240
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001504589787722318,
+      "loss": 0.9664,
+      "step": 5245
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00015060240963855423,
+      "loss": 0.9351,
+      "step": 5250
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00015074584050487664,
+      "loss": 1.0872,
+      "step": 5255
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00015088927137119908,
+      "loss": 1.0728,
+      "step": 5260
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00015103270223752152,
+      "loss": 1.07,
+      "step": 5265
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00015117613310384396,
+      "loss": 1.003,
+      "step": 5270
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00015131956397016638,
+      "loss": 0.9776,
+      "step": 5275
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00015146299483648882,
+      "loss": 0.908,
+      "step": 5280
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00015160642570281126,
+      "loss": 0.8376,
+      "step": 5285
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00015174985656913367,
+      "loss": 0.9636,
+      "step": 5290
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001518932874354561,
+      "loss": 0.9067,
+      "step": 5295
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00015203671830177855,
+      "loss": 1.0184,
+      "step": 5300
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00015218014916810096,
+      "loss": 0.9313,
+      "step": 5305
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00015232358003442343,
+      "loss": 1.1231,
+      "step": 5310
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00015246701090074584,
+      "loss": 1.0447,
+      "step": 5315
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00015261044176706828,
+      "loss": 1.0549,
+      "step": 5320
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00015275387263339072,
+      "loss": 0.9707,
+      "step": 5325
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00015289730349971314,
+      "loss": 0.9855,
+      "step": 5330
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00015304073436603558,
+      "loss": 1.1058,
+      "step": 5335
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00015318416523235802,
+      "loss": 0.8579,
+      "step": 5340
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00015332759609868043,
+      "loss": 1.1284,
+      "step": 5345
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00015347102696500287,
+      "loss": 1.0177,
+      "step": 5350
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001536144578313253,
+      "loss": 0.9577,
+      "step": 5355
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00015375788869764772,
+      "loss": 1.0579,
+      "step": 5360
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001539013195639702,
+      "loss": 0.9464,
+      "step": 5365
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001540447504302926,
+      "loss": 0.9591,
+      "step": 5370
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00015418818129661502,
+      "loss": 0.9479,
+      "step": 5375
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00015433161216293748,
+      "loss": 0.9686,
+      "step": 5380
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001544750430292599,
+      "loss": 0.9715,
+      "step": 5385
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00015461847389558234,
+      "loss": 0.9483,
+      "step": 5390
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00015476190476190478,
+      "loss": 0.8867,
+      "step": 5395
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001549053356282272,
+      "loss": 0.9678,
+      "step": 5400
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00015504876649454963,
+      "loss": 1.0233,
+      "step": 5405
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00015519219736087207,
+      "loss": 1.0728,
+      "step": 5410
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00015533562822719448,
+      "loss": 0.9649,
+      "step": 5415
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00015547905909351695,
+      "loss": 1.0888,
+      "step": 5420
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00015562248995983936,
+      "loss": 0.9936,
+      "step": 5425
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001557659208261618,
+      "loss": 0.9588,
+      "step": 5430
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00015590935169248425,
+      "loss": 0.9805,
+      "step": 5435
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00015605278255880666,
+      "loss": 1.0008,
+      "step": 5440
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0001561962134251291,
+      "loss": 1.0179,
+      "step": 5445
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00015633964429145154,
+      "loss": 0.8855,
+      "step": 5450
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00015648307515777395,
+      "loss": 0.958,
+      "step": 5455
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001566265060240964,
+      "loss": 0.9418,
+      "step": 5460
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00015676993689041883,
+      "loss": 0.9573,
+      "step": 5465
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00015691336775674127,
+      "loss": 1.1149,
+      "step": 5470
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00015705679862306369,
+      "loss": 0.995,
+      "step": 5475
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00015720022948938613,
+      "loss": 0.9856,
+      "step": 5480
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00015734366035570857,
+      "loss": 0.9135,
+      "step": 5485
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00015748709122203098,
+      "loss": 0.8655,
+      "step": 5490
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00015763052208835342,
+      "loss": 1.0002,
+      "step": 5495
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00015777395295467586,
+      "loss": 0.9998,
+      "step": 5500
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00015791738382099827,
+      "loss": 0.9179,
+      "step": 5505
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0001580608146873207,
+      "loss": 0.7862,
+      "step": 5510
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00015820424555364315,
+      "loss": 1.0359,
+      "step": 5515
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00015834767641996557,
+      "loss": 0.9956,
+      "step": 5520
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015849110728628803,
+      "loss": 0.8596,
+      "step": 5525
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00015863453815261045,
+      "loss": 1.1384,
+      "step": 5530
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00015877796901893289,
+      "loss": 1.0153,
+      "step": 5535
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00015892139988525533,
+      "loss": 1.0316,
+      "step": 5540
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00015906483075157774,
+      "loss": 1.0093,
+      "step": 5545
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00015920826161790018,
+      "loss": 1.047,
+      "step": 5550
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00015935169248422262,
+      "loss": 0.9798,
+      "step": 5555
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00015949512335054503,
+      "loss": 0.9495,
+      "step": 5560
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0001596385542168675,
+      "loss": 0.994,
+      "step": 5565
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001597819850831899,
+      "loss": 0.97,
+      "step": 5570
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00015992541594951233,
+      "loss": 0.9392,
+      "step": 5575
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001600688468158348,
+      "loss": 0.9793,
+      "step": 5580
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001602122776821572,
+      "loss": 1.0855,
+      "step": 5585
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00016035570854847965,
+      "loss": 1.0543,
+      "step": 5590
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00016049913941480209,
+      "loss": 1.0799,
+      "step": 5595
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0001606425702811245,
+      "loss": 0.9207,
+      "step": 5600
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00016078600114744694,
+      "loss": 0.9722,
+      "step": 5605
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00016092943201376938,
+      "loss": 1.0413,
+      "step": 5610
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001610728628800918,
+      "loss": 0.9253,
+      "step": 5615
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00016121629374641423,
+      "loss": 1.0691,
+      "step": 5620
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00016135972461273667,
+      "loss": 0.9382,
+      "step": 5625
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00016150315547905909,
+      "loss": 0.9178,
+      "step": 5630
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00016164658634538153,
+      "loss": 1.0859,
+      "step": 5635
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016179001721170397,
+      "loss": 0.9187,
+      "step": 5640
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0001619334480780264,
+      "loss": 0.9841,
+      "step": 5645
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00016207687894434885,
+      "loss": 0.9436,
+      "step": 5650
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00016222030981067126,
+      "loss": 0.9063,
+      "step": 5655
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001623637406769937,
+      "loss": 1.0031,
+      "step": 5660
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016250717154331614,
+      "loss": 0.968,
+      "step": 5665
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00016265060240963855,
+      "loss": 1.1923,
+      "step": 5670
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.000162794033275961,
+      "loss": 0.85,
+      "step": 5675
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00016293746414228343,
+      "loss": 1.0102,
+      "step": 5680
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00016308089500860587,
+      "loss": 1.0054,
+      "step": 5685
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001632243258749283,
+      "loss": 0.8971,
+      "step": 5690
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00016336775674125073,
+      "loss": 1.0443,
+      "step": 5695
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00016351118760757317,
+      "loss": 1.0344,
+      "step": 5700
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00016365461847389558,
+      "loss": 0.9438,
+      "step": 5705
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00016379804934021802,
+      "loss": 1.0228,
+      "step": 5710
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00016394148020654046,
+      "loss": 0.9256,
+      "step": 5715
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00016408491107286287,
+      "loss": 0.9056,
+      "step": 5720
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00016422834193918531,
+      "loss": 1.0632,
+      "step": 5725
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00016437177280550775,
+      "loss": 0.8901,
+      "step": 5730
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00016451520367183017,
+      "loss": 0.9466,
+      "step": 5735
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00016465863453815263,
+      "loss": 1.1376,
+      "step": 5740
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00016480206540447505,
+      "loss": 0.8577,
+      "step": 5745
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00016494549627079746,
+      "loss": 0.9745,
+      "step": 5750
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00016508892713711993,
+      "loss": 1.0465,
+      "step": 5755
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00016523235800344234,
+      "loss": 1.1213,
+      "step": 5760
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00016537578886976478,
+      "loss": 0.9917,
+      "step": 5765
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00016551921973608722,
+      "loss": 0.8194,
+      "step": 5770
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00016566265060240963,
+      "loss": 0.9969,
+      "step": 5775
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001658060814687321,
+      "loss": 0.8893,
+      "step": 5780
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00016594951233505451,
+      "loss": 0.9696,
+      "step": 5785
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00016609294320137693,
+      "loss": 1.1031,
+      "step": 5790
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001662363740676994,
+      "loss": 1.0955,
+      "step": 5795
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001663798049340218,
+      "loss": 1.0878,
+      "step": 5800
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00016652323580034425,
+      "loss": 0.9839,
+      "step": 5805
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001666666666666667,
+      "loss": 1.1202,
+      "step": 5810
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001668100975329891,
+      "loss": 0.9917,
+      "step": 5815
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00016695352839931154,
+      "loss": 0.9115,
+      "step": 5820
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00016709695926563398,
+      "loss": 0.9164,
+      "step": 5825
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001672403901319564,
+      "loss": 0.995,
+      "step": 5830
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016738382099827883,
+      "loss": 1.091,
+      "step": 5835
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00016752725186460127,
+      "loss": 0.9019,
+      "step": 5840
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001676706827309237,
+      "loss": 0.8586,
+      "step": 5845
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00016781411359724613,
+      "loss": 0.9595,
+      "step": 5850
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00016795754446356857,
+      "loss": 1.1204,
+      "step": 5855
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.000168100975329891,
+      "loss": 0.8435,
+      "step": 5860
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00016824440619621342,
+      "loss": 0.9602,
+      "step": 5865
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00016838783706253586,
+      "loss": 0.9821,
+      "step": 5870
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0001685312679288583,
+      "loss": 0.864,
+      "step": 5875
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00016867469879518074,
+      "loss": 0.979,
+      "step": 5880
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00016881812966150316,
+      "loss": 0.9652,
+      "step": 5885
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001689615605278256,
+      "loss": 0.9546,
+      "step": 5890
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00016910499139414804,
+      "loss": 0.9793,
+      "step": 5895
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00016924842226047048,
+      "loss": 0.9838,
+      "step": 5900
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0001693918531267929,
+      "loss": 0.9815,
+      "step": 5905
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00016953528399311533,
+      "loss": 0.9909,
+      "step": 5910
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016967871485943777,
+      "loss": 1.0473,
+      "step": 5915
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00016982214572576018,
+      "loss": 1.0577,
+      "step": 5920
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00016996557659208262,
+      "loss": 1.0259,
+      "step": 5925
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017010900745840506,
+      "loss": 0.9825,
+      "step": 5930
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00017025243832472748,
+      "loss": 1.172,
+      "step": 5935
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00017039586919104992,
+      "loss": 0.8524,
+      "step": 5940
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00017053930005737236,
+      "loss": 0.8916,
+      "step": 5945
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017068273092369477,
+      "loss": 0.9652,
+      "step": 5950
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00017082616179001724,
+      "loss": 0.9527,
+      "step": 5955
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00017096959265633965,
+      "loss": 0.8899,
+      "step": 5960
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00017111302352266206,
+      "loss": 0.9364,
+      "step": 5965
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00017125645438898453,
+      "loss": 0.9621,
+      "step": 5970
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00017139988525530694,
+      "loss": 0.9841,
+      "step": 5975
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00017154331612162938,
+      "loss": 1.0999,
+      "step": 5980
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00017168674698795182,
+      "loss": 0.883,
+      "step": 5985
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017183017785427424,
+      "loss": 1.0271,
+      "step": 5990
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001719736087205967,
+      "loss": 1.0657,
+      "step": 5995
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00017211703958691912,
+      "loss": 1.0207,
+      "step": 6000
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00017226047045324153,
+      "loss": 1.2019,
+      "step": 6005
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.000172403901319564,
+      "loss": 1.1323,
+      "step": 6010
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0001725473321858864,
+      "loss": 0.8854,
+      "step": 6015
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00017269076305220885,
+      "loss": 1.0004,
+      "step": 6020
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001728341939185313,
+      "loss": 0.9655,
+      "step": 6025
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001729776247848537,
+      "loss": 0.9061,
+      "step": 6030
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00017312105565117614,
+      "loss": 0.9962,
+      "step": 6035
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017326448651749858,
+      "loss": 0.9991,
+      "step": 6040
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.000173407917383821,
+      "loss": 1.0361,
+      "step": 6045
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00017355134825014344,
+      "loss": 0.9285,
+      "step": 6050
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00017369477911646588,
+      "loss": 1.048,
+      "step": 6055
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001738382099827883,
+      "loss": 0.8781,
+      "step": 6060
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00017398164084911073,
+      "loss": 1.0609,
+      "step": 6065
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00017412507171543317,
+      "loss": 0.8504,
+      "step": 6070
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001742685025817556,
+      "loss": 1.1307,
+      "step": 6075
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017441193344807802,
+      "loss": 1.0649,
+      "step": 6080
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00017455536431440046,
+      "loss": 0.8316,
+      "step": 6085
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0001746987951807229,
+      "loss": 0.9775,
+      "step": 6090
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00017484222604704534,
+      "loss": 1.0049,
+      "step": 6095
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00017498565691336776,
+      "loss": 0.965,
+      "step": 6100
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0001751290877796902,
+      "loss": 1.0832,
+      "step": 6105
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00017527251864601264,
+      "loss": 0.8216,
+      "step": 6110
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00017541594951233508,
+      "loss": 0.9725,
+      "step": 6115
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0001755593803786575,
+      "loss": 0.8486,
+      "step": 6120
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00017570281124497993,
+      "loss": 1.067,
+      "step": 6125
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017584624211130237,
+      "loss": 0.9053,
+      "step": 6130
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00017598967297762478,
+      "loss": 0.8203,
+      "step": 6135
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00017613310384394722,
+      "loss": 1.0192,
+      "step": 6140
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017627653471026966,
+      "loss": 0.9056,
+      "step": 6145
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00017641996557659208,
+      "loss": 1.0876,
+      "step": 6150
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00017656339644291452,
+      "loss": 1.0021,
+      "step": 6155
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00017670682730923696,
+      "loss": 1.0436,
+      "step": 6160
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00017685025817555937,
+      "loss": 0.874,
+      "step": 6165
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00017699368904188184,
+      "loss": 0.9586,
+      "step": 6170
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00017713711990820425,
+      "loss": 1.0107,
+      "step": 6175
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017728055077452666,
+      "loss": 0.9436,
+      "step": 6180
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017742398164084913,
+      "loss": 0.9741,
+      "step": 6185
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00017756741250717154,
+      "loss": 0.9436,
+      "step": 6190
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017771084337349398,
+      "loss": 0.9143,
+      "step": 6195
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00017785427423981642,
+      "loss": 1.0719,
+      "step": 6200
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00017799770510613884,
+      "loss": 0.795,
+      "step": 6205
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001781411359724613,
+      "loss": 0.8589,
+      "step": 6210
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017828456683878372,
+      "loss": 1.1208,
+      "step": 6215
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00017842799770510613,
+      "loss": 1.0982,
+      "step": 6220
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001785714285714286,
+      "loss": 0.9124,
+      "step": 6225
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.000178714859437751,
+      "loss": 1.1237,
+      "step": 6230
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00017885829030407345,
+      "loss": 0.8831,
+      "step": 6235
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001790017211703959,
+      "loss": 0.9529,
+      "step": 6240
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0001791451520367183,
+      "loss": 1.0183,
+      "step": 6245
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00017928858290304074,
+      "loss": 1.0009,
+      "step": 6250
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00017943201376936319,
+      "loss": 0.79,
+      "step": 6255
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001795754446356856,
+      "loss": 0.9302,
+      "step": 6260
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00017971887550200804,
+      "loss": 1.0073,
+      "step": 6265
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00017986230636833048,
+      "loss": 0.9439,
+      "step": 6270
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001800057372346529,
+      "loss": 0.9953,
+      "step": 6275
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00018014916810097533,
+      "loss": 0.9232,
+      "step": 6280
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018029259896729777,
+      "loss": 0.9308,
+      "step": 6285
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0001804360298336202,
+      "loss": 0.9297,
+      "step": 6290
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00018057946069994263,
+      "loss": 0.9201,
+      "step": 6295
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00018072289156626507,
+      "loss": 1.0003,
+      "step": 6300
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001808663224325875,
+      "loss": 1.0814,
+      "step": 6305
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00018100975329890992,
+      "loss": 1.1829,
+      "step": 6310
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00018115318416523236,
+      "loss": 0.7843,
+      "step": 6315
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001812966150315548,
+      "loss": 1.071,
+      "step": 6320
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00018144004589787724,
+      "loss": 0.766,
+      "step": 6325
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00018158347676419968,
+      "loss": 0.9826,
+      "step": 6330
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0001817269076305221,
+      "loss": 0.9046,
+      "step": 6335
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018187033849684453,
+      "loss": 0.9938,
+      "step": 6340
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00018201376936316697,
+      "loss": 0.9356,
+      "step": 6345
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00018215720022948939,
+      "loss": 0.9787,
+      "step": 6350
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00018230063109581183,
+      "loss": 0.9114,
+      "step": 6355
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00018244406196213427,
+      "loss": 1.0716,
+      "step": 6360
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018258749282845668,
+      "loss": 0.8913,
+      "step": 6365
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018273092369477912,
+      "loss": 1.1467,
+      "step": 6370
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018287435456110156,
+      "loss": 0.9404,
+      "step": 6375
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00018301778542742397,
+      "loss": 0.9886,
+      "step": 6380
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018316121629374644,
+      "loss": 0.847,
+      "step": 6385
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018330464716006885,
+      "loss": 0.8849,
+      "step": 6390
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018344807802639127,
+      "loss": 1.2015,
+      "step": 6395
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00018359150889271373,
+      "loss": 0.9451,
+      "step": 6400
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00018373493975903615,
+      "loss": 0.7917,
+      "step": 6405
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00018387837062535859,
+      "loss": 1.0318,
+      "step": 6410
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00018402180149168103,
+      "loss": 0.9183,
+      "step": 6415
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00018416523235800344,
+      "loss": 0.9658,
+      "step": 6420
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0001843086632243259,
+      "loss": 0.9126,
+      "step": 6425
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00018445209409064832,
+      "loss": 0.9077,
+      "step": 6430
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00018459552495697073,
+      "loss": 0.9502,
+      "step": 6435
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001847389558232932,
+      "loss": 0.8701,
+      "step": 6440
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0001848823866896156,
+      "loss": 0.9623,
+      "step": 6445
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00018502581755593805,
+      "loss": 1.0634,
+      "step": 6450
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001851692484222605,
+      "loss": 0.8987,
+      "step": 6455
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001853126792885829,
+      "loss": 1.0042,
+      "step": 6460
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018545611015490535,
+      "loss": 0.9661,
+      "step": 6465
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001855995410212278,
+      "loss": 0.9081,
+      "step": 6470
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001857429718875502,
+      "loss": 1.1281,
+      "step": 6475
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00018588640275387264,
+      "loss": 0.9521,
+      "step": 6480
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00018602983362019508,
+      "loss": 0.8882,
+      "step": 6485
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0001861732644865175,
+      "loss": 0.9865,
+      "step": 6490
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018631669535283993,
+      "loss": 0.8568,
+      "step": 6495
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018646012621916237,
+      "loss": 1.1288,
+      "step": 6500
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018660355708548481,
+      "loss": 0.9184,
+      "step": 6505
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018674698795180723,
+      "loss": 0.9487,
+      "step": 6510
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00018689041881812967,
+      "loss": 0.8626,
+      "step": 6515
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001870338496844521,
+      "loss": 0.926,
+      "step": 6520
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00018717728055077452,
+      "loss": 0.9192,
+      "step": 6525
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00018732071141709696,
+      "loss": 0.9383,
+      "step": 6530
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001874641422834194,
+      "loss": 0.8595,
+      "step": 6535
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00018760757314974184,
+      "loss": 0.8997,
+      "step": 6540
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00018775100401606428,
+      "loss": 0.9123,
+      "step": 6545
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0001878944348823867,
+      "loss": 0.9546,
+      "step": 6550
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00018803786574870913,
+      "loss": 0.9023,
+      "step": 6555
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018818129661503157,
+      "loss": 0.9866,
+      "step": 6560
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.000188324727481354,
+      "loss": 1.1346,
+      "step": 6565
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00018846815834767643,
+      "loss": 0.946,
+      "step": 6570
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018861158921399887,
+      "loss": 0.9296,
+      "step": 6575
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00018875502008032128,
+      "loss": 0.9713,
+      "step": 6580
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00018889845094664375,
+      "loss": 1.0274,
+      "step": 6585
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018904188181296616,
+      "loss": 0.8749,
+      "step": 6590
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00018918531267928857,
+      "loss": 0.989,
+      "step": 6595
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00018932874354561104,
+      "loss": 0.9316,
+      "step": 6600
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018947217441193345,
+      "loss": 0.9591,
+      "step": 6605
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001896156052782559,
+      "loss": 1.045,
+      "step": 6610
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018975903614457833,
+      "loss": 1.0241,
+      "step": 6615
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018990246701090075,
+      "loss": 1.0077,
+      "step": 6620
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001900458978772232,
+      "loss": 1.047,
+      "step": 6625
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019018932874354563,
+      "loss": 0.986,
+      "step": 6630
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00019033275960986804,
+      "loss": 0.8676,
+      "step": 6635
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00019047619047619048,
+      "loss": 0.9166,
+      "step": 6640
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00019061962134251292,
+      "loss": 1.1594,
+      "step": 6645
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019076305220883533,
+      "loss": 0.9954,
+      "step": 6650
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001909064830751578,
+      "loss": 0.9052,
+      "step": 6655
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019104991394148021,
+      "loss": 0.9444,
+      "step": 6660
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019119334480780266,
+      "loss": 0.9136,
+      "step": 6665
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001913367756741251,
+      "loss": 0.9398,
+      "step": 6670
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001914802065404475,
+      "loss": 1.1613,
+      "step": 6675
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019162363740676995,
+      "loss": 1.0539,
+      "step": 6680
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001917670682730924,
+      "loss": 1.0843,
+      "step": 6685
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001919104991394148,
+      "loss": 0.8968,
+      "step": 6690
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019205393000573724,
+      "loss": 0.9034,
+      "step": 6695
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019219736087205968,
+      "loss": 1.011,
+      "step": 6700
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019234079173838212,
+      "loss": 1.0446,
+      "step": 6705
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019248422260470454,
+      "loss": 0.8931,
+      "step": 6710
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019262765347102698,
+      "loss": 0.8736,
+      "step": 6715
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00019277108433734942,
+      "loss": 0.9596,
+      "step": 6720
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00019291451520367183,
+      "loss": 0.9351,
+      "step": 6725
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019305794606999427,
+      "loss": 0.9922,
+      "step": 6730
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001932013769363167,
+      "loss": 1.0471,
+      "step": 6735
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00019334480780263912,
+      "loss": 0.9003,
+      "step": 6740
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00019348823866896156,
+      "loss": 0.8397,
+      "step": 6745
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.000193631669535284,
+      "loss": 0.9666,
+      "step": 6750
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019377510040160642,
+      "loss": 0.9207,
+      "step": 6755
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019391853126792888,
+      "loss": 0.997,
+      "step": 6760
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0001940619621342513,
+      "loss": 0.9651,
+      "step": 6765
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019420539300057374,
+      "loss": 0.9569,
+      "step": 6770
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019434882386689618,
+      "loss": 1.0032,
+      "step": 6775
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0001944922547332186,
+      "loss": 0.8701,
+      "step": 6780
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019463568559954103,
+      "loss": 0.8503,
+      "step": 6785
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019477911646586347,
+      "loss": 0.9847,
+      "step": 6790
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00019492254733218588,
+      "loss": 0.9694,
+      "step": 6795
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00019506597819850835,
+      "loss": 0.8473,
+      "step": 6800
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019520940906483076,
+      "loss": 0.8632,
+      "step": 6805
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019535283993115318,
+      "loss": 1.0169,
+      "step": 6810
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019549627079747564,
+      "loss": 0.9015,
+      "step": 6815
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019563970166379806,
+      "loss": 1.1592,
+      "step": 6820
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001957831325301205,
+      "loss": 0.9537,
+      "step": 6825
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019592656339644294,
+      "loss": 1.0179,
+      "step": 6830
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019606999426276535,
+      "loss": 0.865,
+      "step": 6835
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001962134251290878,
+      "loss": 1.0772,
+      "step": 6840
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019635685599541023,
+      "loss": 0.9969,
+      "step": 6845
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00019650028686173264,
+      "loss": 0.8717,
+      "step": 6850
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019664371772805508,
+      "loss": 0.9652,
+      "step": 6855
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019678714859437752,
+      "loss": 1.0097,
+      "step": 6860
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019693057946069994,
+      "loss": 0.9766,
+      "step": 6865
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019707401032702238,
+      "loss": 1.0268,
+      "step": 6870
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019721744119334482,
+      "loss": 0.924,
+      "step": 6875
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00019736087205966726,
+      "loss": 0.9296,
+      "step": 6880
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001975043029259897,
+      "loss": 1.0601,
+      "step": 6885
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001976477337923121,
+      "loss": 0.9167,
+      "step": 6890
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019779116465863455,
+      "loss": 0.9092,
+      "step": 6895
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.000197934595524957,
+      "loss": 0.9494,
+      "step": 6900
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0001980780263912794,
+      "loss": 0.8921,
+      "step": 6905
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019822145725760184,
+      "loss": 0.9716,
+      "step": 6910
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019836488812392428,
+      "loss": 0.9527,
+      "step": 6915
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019850831899024672,
+      "loss": 1.0112,
+      "step": 6920
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019865174985656914,
+      "loss": 1.0535,
+      "step": 6925
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019879518072289158,
+      "loss": 0.8874,
+      "step": 6930
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019893861158921402,
+      "loss": 1.0039,
+      "step": 6935
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019908204245553643,
+      "loss": 1.0371,
+      "step": 6940
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019922547332185887,
+      "loss": 1.2052,
+      "step": 6945
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001993689041881813,
+      "loss": 0.942,
+      "step": 6950
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019951233505450372,
+      "loss": 1.0566,
+      "step": 6955
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019965576592082616,
+      "loss": 0.9806,
+      "step": 6960
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0001997991967871486,
+      "loss": 0.8986,
+      "step": 6965
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019994262765347102,
+      "loss": 1.0299,
+      "step": 6970
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001999999988717395,
+      "loss": 0.977,
+      "step": 6975
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0001999999919768143,
+      "loss": 0.8963,
+      "step": 6980
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001999999788137757,
+      "loss": 1.0324,
+      "step": 6985
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00019999995938262455,
+      "loss": 0.9285,
+      "step": 6990
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019999993368336204,
+      "loss": 0.878,
+      "step": 6995
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001999999017159898,
+      "loss": 1.0065,
+      "step": 7000
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019999986348050983,
+      "loss": 1.0322,
+      "step": 7005
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019999981897692452,
+      "loss": 0.9879,
+      "step": 7010
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019999976820523667,
+      "loss": 0.8922,
+      "step": 7015
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019999971116544947,
+      "loss": 1.0516,
+      "step": 7020
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001999996478575665,
+      "loss": 0.9043,
+      "step": 7025
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001999995782815917,
+      "loss": 1.0973,
+      "step": 7030
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001999995024375294,
+      "loss": 0.9128,
+      "step": 7035
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019999942032538444,
+      "loss": 1.0025,
+      "step": 7040
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0001999993319451619,
+      "loss": 0.8311,
+      "step": 7045
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019999923729686737,
+      "loss": 1.0452,
+      "step": 7050
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019999913638050674,
+      "loss": 1.0823,
+      "step": 7055
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00019999902919608635,
+      "loss": 0.8616,
+      "step": 7060
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019999891574361294,
+      "loss": 0.8661,
+      "step": 7065
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019999879602309358,
+      "loss": 1.0924,
+      "step": 7070
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001999986700345358,
+      "loss": 0.9282,
+      "step": 7075
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001999985377779475,
+      "loss": 0.8499,
+      "step": 7080
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00019999839925333697,
+      "loss": 0.9638,
+      "step": 7085
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001999982544607129,
+      "loss": 1.019,
+      "step": 7090
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00019999810340008432,
+      "loss": 1.0765,
+      "step": 7095
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00019999794607146072,
+      "loss": 0.9585,
+      "step": 7100
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019999778247485202,
+      "loss": 0.9834,
+      "step": 7105
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019999761261026838,
+      "loss": 1.0673,
+      "step": 7110
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019999743647772055,
+      "loss": 0.9441,
+      "step": 7115
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019999725407721946,
+      "loss": 0.7877,
+      "step": 7120
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001999970654087766,
+      "loss": 0.9829,
+      "step": 7125
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019999687047240382,
+      "loss": 0.9581,
+      "step": 7130
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0001999966692681133,
+      "loss": 0.9482,
+      "step": 7135
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019999646179591767,
+      "loss": 1.0197,
+      "step": 7140
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.0001999962480558299,
+      "loss": 0.8861,
+      "step": 7145
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019999602804786343,
+      "loss": 1.0572,
+      "step": 7150
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019999580177203208,
+      "loss": 0.8916,
+      "step": 7155
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019999556922834993,
+      "loss": 0.8698,
+      "step": 7160
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019999533041683166,
+      "loss": 0.9218,
+      "step": 7165
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019999508533749216,
+      "loss": 0.8811,
+      "step": 7170
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019999483399034683,
+      "loss": 1.0427,
+      "step": 7175
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019999457637541142,
+      "loss": 0.8641,
+      "step": 7180
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001999943124927021,
+      "loss": 0.9981,
+      "step": 7185
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019999404234223537,
+      "loss": 1.0612,
+      "step": 7190
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0001999937659240282,
+      "loss": 0.9003,
+      "step": 7195
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001999934832380979,
+      "loss": 1.0519,
+      "step": 7200
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019999319428446217,
+      "loss": 1.1604,
+      "step": 7205
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019999289906313914,
+      "loss": 1.0522,
+      "step": 7210
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019999259757414733,
+      "loss": 0.9722,
+      "step": 7215
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001999922898175056,
+      "loss": 1.0899,
+      "step": 7220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019999197579323327,
+      "loss": 0.9977,
+      "step": 7225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019999165550135003,
+      "loss": 0.9424,
+      "step": 7230
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0001999913289418759,
+      "loss": 0.8548,
+      "step": 7235
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019999099611483145,
+      "loss": 0.9436,
+      "step": 7240
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019999065702023744,
+      "loss": 1.0087,
+      "step": 7245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001999903116581152,
+      "loss": 1.0426,
+      "step": 7250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019998996002848634,
+      "loss": 0.9219,
+      "step": 7255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00019998960213137286,
+      "loss": 0.933,
+      "step": 7260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001999892379667973,
+      "loss": 1.0413,
+      "step": 7265
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.0001999888675347824,
+      "loss": 0.9412,
+      "step": 7270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001999884908353514,
+      "loss": 0.9745,
+      "step": 7275
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00019998810786852795,
+      "loss": 0.8132,
+      "step": 7280
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.000199987718634336,
+      "loss": 0.9773,
+      "step": 7285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00019998732313279994,
+      "loss": 0.8849,
+      "step": 7290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019998692136394465,
+      "loss": 0.9524,
+      "step": 7295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001999865133277952,
+      "loss": 1.0294,
+      "step": 7300
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019998609902437725,
+      "loss": 1.0705,
+      "step": 7305
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00019998567845371673,
+      "loss": 0.9558,
+      "step": 7310
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019998525161584,
+      "loss": 1.0696,
+      "step": 7315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019998481851077384,
+      "loss": 0.8552,
+      "step": 7320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019998437913854538,
+      "loss": 0.9012,
+      "step": 7325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019998393349918216,
+      "loss": 0.9747,
+      "step": 7330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0001999834815927121,
+      "loss": 0.9947,
+      "step": 7335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019998302341916357,
+      "loss": 0.9596,
+      "step": 7340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019998255897856523,
+      "loss": 0.9642,
+      "step": 7345
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019998208827094627,
+      "loss": 1.0481,
+      "step": 7350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001999816112963361,
+      "loss": 1.0289,
+      "step": 7355
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001999811280547647,
+      "loss": 1.0296,
+      "step": 7360
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0001999806385462623,
+      "loss": 0.9311,
+      "step": 7365
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001999801427708596,
+      "loss": 1.0396,
+      "step": 7370
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0001999796407285877,
+      "loss": 0.9284,
+      "step": 7375
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019997913241947806,
+      "loss": 1.107,
+      "step": 7380
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00019997861784356254,
+      "loss": 0.946,
+      "step": 7385
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019997809700087336,
+      "loss": 0.9533,
+      "step": 7390
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019997756989144317,
+      "loss": 0.9806,
+      "step": 7395
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001999770365153051,
+      "loss": 1.0817,
+      "step": 7400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001999764968724925,
+      "loss": 0.9468,
+      "step": 7405
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019997595096303919,
+      "loss": 0.9327,
+      "step": 7410
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019997539878697943,
+      "loss": 1.0226,
+      "step": 7415
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001999748403443478,
+      "loss": 1.0015,
+      "step": 7420
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001999742756351793,
+      "loss": 1.029,
+      "step": 7425
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019997370465950936,
+      "loss": 0.9649,
+      "step": 7430
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019997312741737376,
+      "loss": 1.0384,
+      "step": 7435
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019997254390880867,
+      "loss": 0.9221,
+      "step": 7440
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019997195413385064,
+      "loss": 1.0112,
+      "step": 7445
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019997135809253668,
+      "loss": 0.8739,
+      "step": 7450
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019997075578490417,
+      "loss": 1.0613,
+      "step": 7455
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019997014721099082,
+      "loss": 1.0139,
+      "step": 7460
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019996953237083474,
+      "loss": 1.0393,
+      "step": 7465
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019996891126447457,
+      "loss": 1.0109,
+      "step": 7470
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019996828389194914,
+      "loss": 0.901,
+      "step": 7475
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019996765025329785,
+      "loss": 0.9787,
+      "step": 7480
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001999670103485604,
+      "loss": 1.0034,
+      "step": 7485
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019996636417777687,
+      "loss": 1.0497,
+      "step": 7490
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001999657117409878,
+      "loss": 0.8797,
+      "step": 7495
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019996505303823402,
+      "loss": 1.0155,
+      "step": 7500
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019996438806955692,
+      "loss": 0.9911,
+      "step": 7505
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001999637168349981,
+      "loss": 1.0781,
+      "step": 7510
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019996303933459965,
+      "loss": 1.0164,
+      "step": 7515
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019996235556840405,
+      "loss": 0.9753,
+      "step": 7520
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019996166553645417,
+      "loss": 0.96,
+      "step": 7525
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019996096923879323,
+      "loss": 0.9442,
+      "step": 7530
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019996026667546492,
+      "loss": 0.8743,
+      "step": 7535
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019995955784651322,
+      "loss": 0.978,
+      "step": 7540
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001999588427519826,
+      "loss": 0.969,
+      "step": 7545
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019995812139191787,
+      "loss": 0.8663,
+      "step": 7550
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00019995739376636425,
+      "loss": 1.0276,
+      "step": 7555
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019995665987536737,
+      "loss": 1.1052,
+      "step": 7560
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001999559197189732,
+      "loss": 1.019,
+      "step": 7565
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001999551732972281,
+      "loss": 0.8811,
+      "step": 7570
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019995442061017892,
+      "loss": 0.8933,
+      "step": 7575
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019995366165787283,
+      "loss": 1.0845,
+      "step": 7580
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001999528964403574,
+      "loss": 0.8869,
+      "step": 7585
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019995212495768055,
+      "loss": 0.9254,
+      "step": 7590
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001999513472098907,
+      "loss": 0.938,
+      "step": 7595
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00019995056319703657,
+      "loss": 1.0029,
+      "step": 7600
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00019994977291916727,
+      "loss": 0.9692,
+      "step": 7605
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019994897637633244,
+      "loss": 1.0439,
+      "step": 7610
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019994817356858187,
+      "loss": 0.8371,
+      "step": 7615
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019994736449596598,
+      "loss": 0.9537,
+      "step": 7620
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019994654915853546,
+      "loss": 0.9284,
+      "step": 7625
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019994572755634138,
+      "loss": 1.0318,
+      "step": 7630
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019994489968943528,
+      "loss": 1.1444,
+      "step": 7635
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019994406555786905,
+      "loss": 0.8694,
+      "step": 7640
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019994322516169496,
+      "loss": 1.0167,
+      "step": 7645
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001999423785009657,
+      "loss": 1.1432,
+      "step": 7650
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001999415255757343,
+      "loss": 0.9968,
+      "step": 7655
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001999406663860543,
+      "loss": 0.9843,
+      "step": 7660
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019993980093197946,
+      "loss": 0.9883,
+      "step": 7665
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001999389292135641,
+      "loss": 0.8624,
+      "step": 7670
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019993805123086285,
+      "loss": 1.0464,
+      "step": 7675
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001999371669839307,
+      "loss": 0.8542,
+      "step": 7680
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001999362764728231,
+      "loss": 0.9581,
+      "step": 7685
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019993537969759592,
+      "loss": 0.935,
+      "step": 7690
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001999344766583053,
+      "loss": 0.8918,
+      "step": 7695
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019993356735500782,
+      "loss": 1.0286,
+      "step": 7700
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001999326517877606,
+      "loss": 1.0116,
+      "step": 7705
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001999317299566209,
+      "loss": 0.9422,
+      "step": 7710
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001999308018616466,
+      "loss": 0.9841,
+      "step": 7715
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019992986750289577,
+      "loss": 1.05,
+      "step": 7720
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019992892688042705,
+      "loss": 0.9811,
+      "step": 7725
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001999279799942994,
+      "loss": 1.144,
+      "step": 7730
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019992702684457216,
+      "loss": 0.8471,
+      "step": 7735
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00019992606743130507,
+      "loss": 0.8222,
+      "step": 7740
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019992510175455825,
+      "loss": 0.9594,
+      "step": 7745
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019992412981439226,
+      "loss": 0.9451,
+      "step": 7750
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.000199923151610868,
+      "loss": 0.8905,
+      "step": 7755
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019992216714404681,
+      "loss": 1.06,
+      "step": 7760
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00019992117641399036,
+      "loss": 0.9044,
+      "step": 7765
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001999201794207608,
+      "loss": 0.8762,
+      "step": 7770
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00019991917616442057,
+      "loss": 0.9525,
+      "step": 7775
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001999181666450326,
+      "loss": 1.1321,
+      "step": 7780
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001999171508626601,
+      "loss": 0.9884,
+      "step": 7785
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019991612881736684,
+      "loss": 0.9493,
+      "step": 7790
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001999151005092168,
+      "loss": 1.0176,
+      "step": 7795
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001999140659382745,
+      "loss": 0.9507,
+      "step": 7800
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019991302510460472,
+      "loss": 0.9906,
+      "step": 7805
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019991197800827275,
+      "loss": 0.9058,
+      "step": 7810
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001999109246493442,
+      "loss": 0.9245,
+      "step": 7815
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001999098650278851,
+      "loss": 1.0691,
+      "step": 7820
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019990879914396189,
+      "loss": 0.9908,
+      "step": 7825
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00019990772699764135,
+      "loss": 0.8805,
+      "step": 7830
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001999066485889907,
+      "loss": 0.9424,
+      "step": 7835
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019990556391807752,
+      "loss": 0.8939,
+      "step": 7840
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019990447298496986,
+      "loss": 1.0941,
+      "step": 7845
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.000199903375789736,
+      "loss": 1.064,
+      "step": 7850
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019990227233244477,
+      "loss": 1.1336,
+      "step": 7855
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00019990116261316536,
+      "loss": 0.8951,
+      "step": 7860
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019990004663196728,
+      "loss": 0.8861,
+      "step": 7865
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001998989243889205,
+      "loss": 0.9515,
+      "step": 7870
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019989779588409538,
+      "loss": 0.81,
+      "step": 7875
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019989666111756263,
+      "loss": 0.937,
+      "step": 7880
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019989552008939336,
+      "loss": 1.0635,
+      "step": 7885
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019989437279965917,
+      "loss": 1.0161,
+      "step": 7890
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019989321924843186,
+      "loss": 1.0589,
+      "step": 7895
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019989205943578387,
+      "loss": 0.9836,
+      "step": 7900
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001998908933617878,
+      "loss": 0.9613,
+      "step": 7905
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019988972102651675,
+      "loss": 1.0425,
+      "step": 7910
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019988854243004422,
+      "loss": 0.9164,
+      "step": 7915
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001998873575724441,
+      "loss": 0.9471,
+      "step": 7920
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019988616645379064,
+      "loss": 0.8745,
+      "step": 7925
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001998849690741585,
+      "loss": 0.8713,
+      "step": 7930
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019988376543362277,
+      "loss": 1.0146,
+      "step": 7935
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019988255553225886,
+      "loss": 1.0469,
+      "step": 7940
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019988133937014256,
+      "loss": 1.032,
+      "step": 7945
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001998801169473502,
+      "loss": 0.9113,
+      "step": 7950
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019987888826395836,
+      "loss": 1.0555,
+      "step": 7955
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00019987765332004403,
+      "loss": 1.0353,
+      "step": 7960
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019987641211568463,
+      "loss": 0.8749,
+      "step": 7965
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019987516465095798,
+      "loss": 1.0204,
+      "step": 7970
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019987391092594228,
+      "loss": 0.978,
+      "step": 7975
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019987265094071607,
+      "loss": 1.0604,
+      "step": 7980
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00019987138469535837,
+      "loss": 0.9035,
+      "step": 7985
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019987011218994854,
+      "loss": 1.1171,
+      "step": 7990
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019986883342456633,
+      "loss": 0.9754,
+      "step": 7995
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019986754839929188,
+      "loss": 1.0346,
+      "step": 8000
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019986625711420578,
+      "loss": 0.9792,
+      "step": 8005
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019986495956938894,
+      "loss": 1.1406,
+      "step": 8010
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019986365576492266,
+      "loss": 0.9641,
+      "step": 8015
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019986234570088876,
+      "loss": 1.0433,
+      "step": 8020
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001998610293773693,
+      "loss": 1.013,
+      "step": 8025
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001998597067944467,
+      "loss": 1.0205,
+      "step": 8030
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019985837795220404,
+      "loss": 0.8346,
+      "step": 8035
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001998570428507245,
+      "loss": 1.0299,
+      "step": 8040
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019985570149009177,
+      "loss": 0.8954,
+      "step": 8045
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019985435387038996,
+      "loss": 1.0256,
+      "step": 8050
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001998529999917035,
+      "loss": 1.1181,
+      "step": 8055
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001998516398541173,
+      "loss": 0.9315,
+      "step": 8060
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019985027345771657,
+      "loss": 1.0552,
+      "step": 8065
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.000199848900802587,
+      "loss": 1.0288,
+      "step": 8070
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001998475218888146,
+      "loss": 1.0476,
+      "step": 8075
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019984613671648579,
+      "loss": 1.0866,
+      "step": 8080
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019984474528568746,
+      "loss": 0.9443,
+      "step": 8085
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019984334759650676,
+      "loss": 0.9872,
+      "step": 8090
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001998419436490313,
+      "loss": 0.9306,
+      "step": 8095
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019984053344334916,
+      "loss": 1.1319,
+      "step": 8100
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019983911697954863,
+      "loss": 0.9468,
+      "step": 8105
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019983769425771855,
+      "loss": 1.0312,
+      "step": 8110
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019983626527794808,
+      "loss": 1.0196,
+      "step": 8115
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001998348300403268,
+      "loss": 0.8862,
+      "step": 8120
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019983338854494468,
+      "loss": 0.8766,
+      "step": 8125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019983194079189204,
+      "loss": 0.9119,
+      "step": 8130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001998304867812597,
+      "loss": 1.1187,
+      "step": 8135
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019982902651313872,
+      "loss": 0.9005,
+      "step": 8140
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019982755998762067,
+      "loss": 0.9234,
+      "step": 8145
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019982608720479745,
+      "loss": 1.0712,
+      "step": 8150
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019982460816476142,
+      "loss": 0.8635,
+      "step": 8155
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019982312286760524,
+      "loss": 1.1264,
+      "step": 8160
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019982163131342201,
+      "loss": 0.9491,
+      "step": 8165
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001998201335023053,
+      "loss": 0.956,
+      "step": 8170
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00019981862943434887,
+      "loss": 0.983,
+      "step": 8175
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00019981711910964714,
+      "loss": 0.9417,
+      "step": 8180
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019981560252829464,
+      "loss": 1.0731,
+      "step": 8185
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019981407969038651,
+      "loss": 0.9941,
+      "step": 8190
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001998125505960182,
+      "loss": 0.935,
+      "step": 8195
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019981101524528552,
+      "loss": 0.8414,
+      "step": 8200
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019980947363828472,
+      "loss": 1.0664,
+      "step": 8205
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019980792577511247,
+      "loss": 0.88,
+      "step": 8210
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019980637165586574,
+      "loss": 0.9858,
+      "step": 8215
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019980481128064196,
+      "loss": 0.9296,
+      "step": 8220
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00019980324464953896,
+      "loss": 1.0408,
+      "step": 8225
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001998016717626549,
+      "loss": 1.1235,
+      "step": 8230
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019980009262008844,
+      "loss": 0.8829,
+      "step": 8235
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019979850722193847,
+      "loss": 1.0015,
+      "step": 8240
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001997969155683044,
+      "loss": 0.8638,
+      "step": 8245
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.000199795317659286,
+      "loss": 0.8831,
+      "step": 8250
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00019979371349498345,
+      "loss": 0.8855,
+      "step": 8255
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019979210307549726,
+      "loss": 0.9473,
+      "step": 8260
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019979048640092844,
+      "loss": 0.919,
+      "step": 8265
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019978886347137825,
+      "loss": 0.9902,
+      "step": 8270
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019978723428694845,
+      "loss": 0.9017,
+      "step": 8275
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019978559884774116,
+      "loss": 0.8228,
+      "step": 8280
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001997839571538589,
+      "loss": 0.9621,
+      "step": 8285
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019978230920540452,
+      "loss": 0.9191,
+      "step": 8290
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019978065500248139,
+      "loss": 1.051,
+      "step": 8295
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019977899454519315,
+      "loss": 0.8845,
+      "step": 8300
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001997773278336439,
+      "loss": 1.0515,
+      "step": 8305
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0001997756548679381,
+      "loss": 0.9412,
+      "step": 8310
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019977397564818064,
+      "loss": 1.0884,
+      "step": 8315
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019977229017447673,
+      "loss": 1.1093,
+      "step": 8320
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019977059844693206,
+      "loss": 0.9655,
+      "step": 8325
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001997689004656526,
+      "loss": 0.9096,
+      "step": 8330
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001997671962307449,
+      "loss": 1.0204,
+      "step": 8335
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001997654857423157,
+      "loss": 0.909,
+      "step": 8340
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001997637690004722,
+      "loss": 0.966,
+      "step": 8345
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019976204600532206,
+      "loss": 1.0793,
+      "step": 8350
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00019976031675697322,
+      "loss": 0.8746,
+      "step": 8355
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019975858125553415,
+      "loss": 1.0465,
+      "step": 8360
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019975683950111358,
+      "loss": 0.8873,
+      "step": 8365
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001997550914938207,
+      "loss": 0.9766,
+      "step": 8370
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00019975333723376505,
+      "loss": 0.8231,
+      "step": 8375
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00019975157672105665,
+      "loss": 0.8231,
+      "step": 8380
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019974980995580578,
+      "loss": 0.9867,
+      "step": 8385
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019974803693812324,
+      "loss": 0.91,
+      "step": 8390
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019974625766812013,
+      "loss": 0.9998,
+      "step": 8395
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.000199744472145908,
+      "loss": 0.9511,
+      "step": 8400
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019974268037159873,
+      "loss": 1.0058,
+      "step": 8405
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019974088234530468,
+      "loss": 1.0336,
+      "step": 8410
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00019973907806713852,
+      "loss": 1.1555,
+      "step": 8415
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019973726753721335,
+      "loss": 0.99,
+      "step": 8420
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019973545075564268,
+      "loss": 0.9725,
+      "step": 8425
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019973362772254035,
+      "loss": 0.9521,
+      "step": 8430
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019973179843802064,
+      "loss": 0.9352,
+      "step": 8435
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019972996290219824,
+      "loss": 0.8785,
+      "step": 8440
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019972812111518817,
+      "loss": 0.9064,
+      "step": 8445
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001997262730771059,
+      "loss": 0.9741,
+      "step": 8450
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019972441878806721,
+      "loss": 1.0906,
+      "step": 8455
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019972255824818845,
+      "loss": 0.9678,
+      "step": 8460
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019972069145758609,
+      "loss": 1.0063,
+      "step": 8465
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019971881841637727,
+      "loss": 0.9492,
+      "step": 8470
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019971693912467932,
+      "loss": 1.1059,
+      "step": 8475
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019971505358261005,
+      "loss": 0.866,
+      "step": 8480
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019971316179028765,
+      "loss": 0.9676,
+      "step": 8485
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019971126374783074,
+      "loss": 0.8942,
+      "step": 8490
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00019970935945535823,
+      "loss": 0.955,
+      "step": 8495
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019970744891298952,
+      "loss": 1.0123,
+      "step": 8500
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019970553212084435,
+      "loss": 1.1463,
+      "step": 8505
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00019970360907904287,
+      "loss": 0.8798,
+      "step": 8510
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001997016797877056,
+      "loss": 0.9562,
+      "step": 8515
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019969974424695352,
+      "loss": 0.9939,
+      "step": 8520
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019969780245690792,
+      "loss": 0.9863,
+      "step": 8525
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00019969585441769052,
+      "loss": 0.9419,
+      "step": 8530
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001996939001294234,
+      "loss": 0.9845,
+      "step": 8535
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001996919395922291,
+      "loss": 0.9737,
+      "step": 8540
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019968997280623047,
+      "loss": 0.9785,
+      "step": 8545
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019968799977155083,
+      "loss": 0.963,
+      "step": 8550
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001996860204883138,
+      "loss": 0.8494,
+      "step": 8555
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019968403495664348,
+      "loss": 0.8966,
+      "step": 8560
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019968204317666436,
+      "loss": 0.8522,
+      "step": 8565
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001996800451485012,
+      "loss": 1.0312,
+      "step": 8570
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019967804087227928,
+      "loss": 1.0109,
+      "step": 8575
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019967603034812425,
+      "loss": 0.8581,
+      "step": 8580
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001996740135761621,
+      "loss": 0.9733,
+      "step": 8585
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019967199055651928,
+      "loss": 0.9642,
+      "step": 8590
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019966996128932256,
+      "loss": 0.9924,
+      "step": 8595
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00019966792577469917,
+      "loss": 0.7756,
+      "step": 8600
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019966588401277666,
+      "loss": 0.9274,
+      "step": 8605
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019966383600368307,
+      "loss": 1.006,
+      "step": 8610
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019966178174754667,
+      "loss": 0.9224,
+      "step": 8615
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00019965972124449634,
+      "loss": 0.8353,
+      "step": 8620
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019965765449466116,
+      "loss": 0.9543,
+      "step": 8625
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019965558149817069,
+      "loss": 0.9076,
+      "step": 8630
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019965350225515486,
+      "loss": 0.9186,
+      "step": 8635
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019965141676574403,
+      "loss": 1.0729,
+      "step": 8640
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019964932503006893,
+      "loss": 1.0694,
+      "step": 8645
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001996472270482606,
+      "loss": 0.9504,
+      "step": 8650
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001996451228204506,
+      "loss": 0.9101,
+      "step": 8655
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019964301234677082,
+      "loss": 1.0089,
+      "step": 8660
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019964089562735356,
+      "loss": 1.0573,
+      "step": 8665
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019963877266233147,
+      "loss": 0.9161,
+      "step": 8670
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001996366434518376,
+      "loss": 0.8975,
+      "step": 8675
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0001996345079960055,
+      "loss": 0.986,
+      "step": 8680
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019963236629496893,
+      "loss": 1.1205,
+      "step": 8685
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019963021834886217,
+      "loss": 0.9446,
+      "step": 8690
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019962806415781988,
+      "loss": 0.8427,
+      "step": 8695
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019962590372197701,
+      "loss": 1.0005,
+      "step": 8700
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019962373704146907,
+      "loss": 1.1333,
+      "step": 8705
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001996215641164318,
+      "loss": 1.1934,
+      "step": 8710
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019961938494700147,
+      "loss": 0.916,
+      "step": 8715
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019961719953331462,
+      "loss": 1.0266,
+      "step": 8720
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019961500787550823,
+      "loss": 1.1552,
+      "step": 8725
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0001996128099737197,
+      "loss": 1.0242,
+      "step": 8730
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019961060582808683,
+      "loss": 1.0109,
+      "step": 8735
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001996083954387477,
+      "loss": 1.0398,
+      "step": 8740
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019960617880584091,
+      "loss": 0.8758,
+      "step": 8745
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001996039559295054,
+      "loss": 1.0004,
+      "step": 8750
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019960172680988047,
+      "loss": 0.9889,
+      "step": 8755
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019959949144710587,
+      "loss": 0.8624,
+      "step": 8760
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019959724984132174,
+      "loss": 0.8611,
+      "step": 8765
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019959500199266854,
+      "loss": 0.9709,
+      "step": 8770
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019959274790128716,
+      "loss": 1.0855,
+      "step": 8775
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019959048756731896,
+      "loss": 1.0042,
+      "step": 8780
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00019958822099090556,
+      "loss": 0.9496,
+      "step": 8785
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019958594817218902,
+      "loss": 0.9307,
+      "step": 8790
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019958366911131188,
+      "loss": 0.9806,
+      "step": 8795
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001995813838084169,
+      "loss": 0.8629,
+      "step": 8800
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001995790922636474,
+      "loss": 1.0448,
+      "step": 8805
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019957679447714697,
+      "loss": 0.9303,
+      "step": 8810
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019957449044905964,
+      "loss": 1.0203,
+      "step": 8815
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019957218017952987,
+      "loss": 0.8014,
+      "step": 8820
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019956986366870242,
+      "loss": 1.0335,
+      "step": 8825
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00019956754091672254,
+      "loss": 1.1559,
+      "step": 8830
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001995652119237358,
+      "loss": 0.9023,
+      "step": 8835
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019956287668988814,
+      "loss": 0.9111,
+      "step": 8840
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019956053521532602,
+      "loss": 1.0862,
+      "step": 8845
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019955818750019613,
+      "loss": 1.0293,
+      "step": 8850
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019955583354464568,
+      "loss": 1.2116,
+      "step": 8855
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001995534733488222,
+      "loss": 0.877,
+      "step": 8860
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019955110691287362,
+      "loss": 0.9599,
+      "step": 8865
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019954873423694825,
+      "loss": 0.965,
+      "step": 8870
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019954635532119487,
+      "loss": 0.9688,
+      "step": 8875
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00019954397016576258,
+      "loss": 0.9937,
+      "step": 8880
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019954157877080086,
+      "loss": 0.9031,
+      "step": 8885
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001995391811364596,
+      "loss": 1.0922,
+      "step": 8890
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019953677726288914,
+      "loss": 1.0146,
+      "step": 8895
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019953436715024008,
+      "loss": 1.1008,
+      "step": 8900
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019953195079866354,
+      "loss": 1.0151,
+      "step": 8905
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019952952820831093,
+      "loss": 0.9264,
+      "step": 8910
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001995270993793342,
+      "loss": 0.9496,
+      "step": 8915
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001995246643118855,
+      "loss": 0.9531,
+      "step": 8920
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019952222300611747,
+      "loss": 0.9197,
+      "step": 8925
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019951977546218318,
+      "loss": 1.096,
+      "step": 8930
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019951732168023602,
+      "loss": 0.9807,
+      "step": 8935
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019951486166042978,
+      "loss": 0.9146,
+      "step": 8940
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001995123954029187,
+      "loss": 0.9749,
+      "step": 8945
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019950992290785732,
+      "loss": 0.9829,
+      "step": 8950
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019950744417540067,
+      "loss": 0.8879,
+      "step": 8955
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019950495920570408,
+      "loss": 0.8371,
+      "step": 8960
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019950246799892328,
+      "loss": 1.046,
+      "step": 8965
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001994999705552145,
+      "loss": 0.9569,
+      "step": 8970
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019949746687473422,
+      "loss": 0.9587,
+      "step": 8975
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019949495695763942,
+      "loss": 1.0237,
+      "step": 8980
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001994924408040874,
+      "loss": 1.0305,
+      "step": 8985
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019948991841423587,
+      "loss": 0.9636,
+      "step": 8990
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00019948738978824295,
+      "loss": 1.0234,
+      "step": 8995
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001994848549262671,
+      "loss": 0.9714,
+      "step": 9000
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001994823138284673,
+      "loss": 0.9048,
+      "step": 9005
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019947976649500274,
+      "loss": 0.996,
+      "step": 9010
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00019947721292603313,
+      "loss": 0.7694,
+      "step": 9015
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019947465312171846,
+      "loss": 0.8679,
+      "step": 9020
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019947208708221933,
+      "loss": 1.1444,
+      "step": 9025
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019946951480769642,
+      "loss": 1.1269,
+      "step": 9030
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019946693629831105,
+      "loss": 1.1152,
+      "step": 9035
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019946435155422486,
+      "loss": 0.9622,
+      "step": 9040
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00019946176057559982,
+      "loss": 0.9847,
+      "step": 9045
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00019945916336259833,
+      "loss": 0.8883,
+      "step": 9050
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019945655991538322,
+      "loss": 0.9141,
+      "step": 9055
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019945395023411767,
+      "loss": 1.0952,
+      "step": 9060
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019945133431896524,
+      "loss": 1.006,
+      "step": 9065
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001994487121700899,
+      "loss": 0.9484,
+      "step": 9070
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019944608378765603,
+      "loss": 0.8655,
+      "step": 9075
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001994434491718284,
+      "loss": 1.0319,
+      "step": 9080
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019944080832277208,
+      "loss": 1.0843,
+      "step": 9085
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019943816124065266,
+      "loss": 0.9651,
+      "step": 9090
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019943550792563604,
+      "loss": 0.9859,
+      "step": 9095
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001994328483778885,
+      "loss": 1.0562,
+      "step": 9100
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019943018259757685,
+      "loss": 1.0792,
+      "step": 9105
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019942751058486807,
+      "loss": 0.9782,
+      "step": 9110
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019942483233992967,
+      "loss": 0.8319,
+      "step": 9115
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019942214786292957,
+      "loss": 1.0149,
+      "step": 9120
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019941945715403598,
+      "loss": 0.9469,
+      "step": 9125
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019941676021341764,
+      "loss": 0.9073,
+      "step": 9130
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001994140570412435,
+      "loss": 1.0213,
+      "step": 9135
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019941134763768305,
+      "loss": 1.0976,
+      "step": 9140
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001994086320029061,
+      "loss": 0.9995,
+      "step": 9145
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001994059101370829,
+      "loss": 0.9779,
+      "step": 9150
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019940318204038406,
+      "loss": 0.931,
+      "step": 9155
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001994004477129805,
+      "loss": 0.9962,
+      "step": 9160
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019939770715504373,
+      "loss": 0.9804,
+      "step": 9165
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019939496036674542,
+      "loss": 1.0479,
+      "step": 9170
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019939220734825784,
+      "loss": 1.0682,
+      "step": 9175
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001993894480997535,
+      "loss": 0.8179,
+      "step": 9180
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001993866826214053,
+      "loss": 1.1365,
+      "step": 9185
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019938391091338668,
+      "loss": 1.0454,
+      "step": 9190
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019938113297587133,
+      "loss": 0.9802,
+      "step": 9195
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019937834880903342,
+      "loss": 0.9689,
+      "step": 9200
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019937555841304736,
+      "loss": 0.9684,
+      "step": 9205
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019937276178808814,
+      "loss": 0.9603,
+      "step": 9210
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019936995893433105,
+      "loss": 1.0888,
+      "step": 9215
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019936714985195177,
+      "loss": 0.991,
+      "step": 9220
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019936433454112635,
+      "loss": 0.9825,
+      "step": 9225
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019936151300203127,
+      "loss": 0.9383,
+      "step": 9230
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001993586852348434,
+      "loss": 0.9222,
+      "step": 9235
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019935585123973998,
+      "loss": 0.9562,
+      "step": 9240
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019935301101689864,
+      "loss": 1.1302,
+      "step": 9245
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001993501645664974,
+      "loss": 0.8784,
+      "step": 9250
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019934731188871473,
+      "loss": 1.0054,
+      "step": 9255
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001993444529837294,
+      "loss": 0.9292,
+      "step": 9260
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019934158785172058,
+      "loss": 0.8246,
+      "step": 9265
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019933871649286796,
+      "loss": 0.8338,
+      "step": 9270
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019933583890735138,
+      "loss": 0.8329,
+      "step": 9275
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001993329550953513,
+      "loss": 1.0388,
+      "step": 9280
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001993300650570485,
+      "loss": 0.952,
+      "step": 9285
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00019932716879262404,
+      "loss": 1.2071,
+      "step": 9290
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019932426630225956,
+      "loss": 0.9085,
+      "step": 9295
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019932135758613694,
+      "loss": 0.9346,
+      "step": 9300
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001993184426444385,
+      "loss": 0.9858,
+      "step": 9305
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019931552147734697,
+      "loss": 1.0675,
+      "step": 9310
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019931259408504545,
+      "loss": 1.0765,
+      "step": 9315
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001993096604677174,
+      "loss": 1.045,
+      "step": 9320
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019930672062554674,
+      "loss": 0.7794,
+      "step": 9325
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019930377455871771,
+      "loss": 0.8622,
+      "step": 9330
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019930082226741504,
+      "loss": 0.9834,
+      "step": 9335
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001992978637518237,
+      "loss": 0.9345,
+      "step": 9340
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019929489901212918,
+      "loss": 0.9969,
+      "step": 9345
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001992919280485173,
+      "loss": 1.1695,
+      "step": 9350
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001992889508611743,
+      "loss": 1.2468,
+      "step": 9355
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019928596745028677,
+      "loss": 1.0997,
+      "step": 9360
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001992829778160417,
+      "loss": 0.8832,
+      "step": 9365
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001992799819586265,
+      "loss": 1.0756,
+      "step": 9370
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.000199276979878229,
+      "loss": 0.8576,
+      "step": 9375
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001992739715750373,
+      "loss": 1.2125,
+      "step": 9380
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019927095704924005,
+      "loss": 0.8716,
+      "step": 9385
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019926793630102612,
+      "loss": 1.0045,
+      "step": 9390
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001992649093305849,
+      "loss": 0.9902,
+      "step": 9395
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019926187613810608,
+      "loss": 0.9739,
+      "step": 9400
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019925883672377982,
+      "loss": 0.9858,
+      "step": 9405
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019925579108779663,
+      "loss": 1.0679,
+      "step": 9410
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019925273923034742,
+      "loss": 0.9112,
+      "step": 9415
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019924968115162347,
+      "loss": 0.9754,
+      "step": 9420
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001992466168518165,
+      "loss": 0.8947,
+      "step": 9425
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019924354633111852,
+      "loss": 1.0981,
+      "step": 9430
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019924046958972206,
+      "loss": 0.8397,
+      "step": 9435
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00019923738662781988,
+      "loss": 0.7539,
+      "step": 9440
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019923429744560532,
+      "loss": 0.8619,
+      "step": 9445
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.000199231202043272,
+      "loss": 0.9305,
+      "step": 9450
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019922810042101387,
+      "loss": 0.8837,
+      "step": 9455
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019922499257902544,
+      "loss": 1.0377,
+      "step": 9460
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019922187851750144,
+      "loss": 1.0431,
+      "step": 9465
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001992187582366371,
+      "loss": 1.0584,
+      "step": 9470
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.000199215631736628,
+      "loss": 0.913,
+      "step": 9475
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001992124990176701,
+      "loss": 0.8658,
+      "step": 9480
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001992093600799598,
+      "loss": 1.0133,
+      "step": 9485
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019920621492369375,
+      "loss": 0.9382,
+      "step": 9490
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001992030635490692,
+      "loss": 0.9738,
+      "step": 9495
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019919990595628363,
+      "loss": 1.0165,
+      "step": 9500
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.000199196742145535,
+      "loss": 0.8768,
+      "step": 9505
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019919357211702158,
+      "loss": 0.8994,
+      "step": 9510
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019919039587094211,
+      "loss": 0.9562,
+      "step": 9515
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019918721340749564,
+      "loss": 0.9853,
+      "step": 9520
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019918402472688166,
+      "loss": 1.0501,
+      "step": 9525
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019918082982930008,
+      "loss": 1.0426,
+      "step": 9530
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019917762871495112,
+      "loss": 0.9506,
+      "step": 9535
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019917442138403543,
+      "loss": 0.9277,
+      "step": 9540
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.000199171207836754,
+      "loss": 1.0632,
+      "step": 9545
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019916798807330838,
+      "loss": 0.9718,
+      "step": 9550
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019916476209390035,
+      "loss": 0.9782,
+      "step": 9555
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019916152989873204,
+      "loss": 0.9661,
+      "step": 9560
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019915829148800613,
+      "loss": 0.8995,
+      "step": 9565
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00019915504686192557,
+      "loss": 0.9292,
+      "step": 9570
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019915179602069373,
+      "loss": 0.9884,
+      "step": 9575
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001991485389645144,
+      "loss": 0.9105,
+      "step": 9580
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019914527569359173,
+      "loss": 1.0626,
+      "step": 9585
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019914200620813025,
+      "loss": 1.0535,
+      "step": 9590
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019913873050833494,
+      "loss": 1.0736,
+      "step": 9595
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019913544859441107,
+      "loss": 0.8547,
+      "step": 9600
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019913216046656436,
+      "loss": 1.1305,
+      "step": 9605
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019912886612500095,
+      "loss": 1.0175,
+      "step": 9610
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019912556556992732,
+      "loss": 0.9938,
+      "step": 9615
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019912225880155032,
+      "loss": 0.8953,
+      "step": 9620
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001991189458200773,
+      "loss": 0.9436,
+      "step": 9625
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019911562662571581,
+      "loss": 1.0913,
+      "step": 9630
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019911230121867396,
+      "loss": 0.9126,
+      "step": 9635
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019910896959916024,
+      "loss": 1.0492,
+      "step": 9640
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001991056317673834,
+      "loss": 1.0146,
+      "step": 9645
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019910228772355268,
+      "loss": 0.9471,
+      "step": 9650
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019909893746787772,
+      "loss": 0.9236,
+      "step": 9655
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019909558100056847,
+      "loss": 0.8822,
+      "step": 9660
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019909221832183538,
+      "loss": 1.0455,
+      "step": 9665
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019908884943188915,
+      "loss": 1.0162,
+      "step": 9670
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.000199085474330941,
+      "loss": 0.9446,
+      "step": 9675
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019908209301920248,
+      "loss": 0.9998,
+      "step": 9680
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001990787054968855,
+      "loss": 0.977,
+      "step": 9685
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019907531176420245,
+      "loss": 0.9763,
+      "step": 9690
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019907191182136596,
+      "loss": 0.7974,
+      "step": 9695
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019906850566858928,
+      "loss": 0.9493,
+      "step": 9700
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001990650933060858,
+      "loss": 0.8047,
+      "step": 9705
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019906167473406945,
+      "loss": 0.9946,
+      "step": 9710
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019905824995275452,
+      "loss": 0.928,
+      "step": 9715
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019905481896235566,
+      "loss": 1.0442,
+      "step": 9720
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019905138176308791,
+      "loss": 0.8989,
+      "step": 9725
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019904793835516676,
+      "loss": 1.0771,
+      "step": 9730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.000199044488738808,
+      "loss": 1.0894,
+      "step": 9735
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001990410329142279,
+      "loss": 1.1339,
+      "step": 9740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019903757088164306,
+      "loss": 0.8538,
+      "step": 9745
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001990341026412705,
+      "loss": 1.1117,
+      "step": 9750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001990306281933276,
+      "loss": 0.9435,
+      "step": 9755
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019902714753803212,
+      "loss": 1.0701,
+      "step": 9760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019902366067560222,
+      "loss": 1.0426,
+      "step": 9765
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019902016760625654,
+      "loss": 1.1211,
+      "step": 9770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019901666833021397,
+      "loss": 0.9565,
+      "step": 9775
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019901316284769385,
+      "loss": 1.0333,
+      "step": 9780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001990096511589159,
+      "loss": 1.0048,
+      "step": 9785
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019900613326410027,
+      "loss": 1.0639,
+      "step": 9790
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019900260916346743,
+      "loss": 1.0796,
+      "step": 9795
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019899907885723834,
+      "loss": 0.9319,
+      "step": 9800
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001989955423456342,
+      "loss": 0.9127,
+      "step": 9805
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019899199962887672,
+      "loss": 0.9493,
+      "step": 9810
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019898845070718797,
+      "loss": 0.9051,
+      "step": 9815
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019898489558079039,
+      "loss": 0.9273,
+      "step": 9820
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019898133424990682,
+      "loss": 0.9571,
+      "step": 9825
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001989777667147605,
+      "loss": 0.9934,
+      "step": 9830
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019897419297557504,
+      "loss": 0.9298,
+      "step": 9835
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001989706130325744,
+      "loss": 0.9945,
+      "step": 9840
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019896702688598306,
+      "loss": 1.0029,
+      "step": 9845
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019896343453602576,
+      "loss": 0.9952,
+      "step": 9850
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019895983598292762,
+      "loss": 0.9123,
+      "step": 9855
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001989562312269143,
+      "loss": 0.9369,
+      "step": 9860
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001989526202682117,
+      "loss": 1.1773,
+      "step": 9865
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001989490031070462,
+      "loss": 0.9184,
+      "step": 9870
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019894537974364442,
+      "loss": 0.9259,
+      "step": 9875
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019894175017823358,
+      "loss": 0.9172,
+      "step": 9880
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019893811441104115,
+      "loss": 0.9146,
+      "step": 9885
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019893447244229503,
+      "loss": 1.0027,
+      "step": 9890
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019893082427222352,
+      "loss": 0.9167,
+      "step": 9895
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019892716990105528,
+      "loss": 0.9703,
+      "step": 9900
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001989235093290193,
+      "loss": 0.9928,
+      "step": 9905
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019891984255634513,
+      "loss": 0.9418,
+      "step": 9910
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019891616958326257,
+      "loss": 0.8257,
+      "step": 9915
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019891249041000184,
+      "loss": 0.8762,
+      "step": 9920
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019890880503679358,
+      "loss": 0.9428,
+      "step": 9925
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019890511346386873,
+      "loss": 1.0612,
+      "step": 9930
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019890141569145877,
+      "loss": 1.0748,
+      "step": 9935
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001988977117197954,
+      "loss": 1.0498,
+      "step": 9940
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019889400154911085,
+      "loss": 0.9022,
+      "step": 9945
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019889028517963762,
+      "loss": 0.9875,
+      "step": 9950
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019888656261160873,
+      "loss": 0.8687,
+      "step": 9955
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019888283384525743,
+      "loss": 0.9364,
+      "step": 9960
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019887909888081752,
+      "loss": 0.904,
+      "step": 9965
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019887535771852307,
+      "loss": 1.2971,
+      "step": 9970
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019887161035860859,
+      "loss": 0.9255,
+      "step": 9975
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019886785680130892,
+      "loss": 0.9778,
+      "step": 9980
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019886409704685944,
+      "loss": 0.9391,
+      "step": 9985
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019886033109549575,
+      "loss": 0.9114,
+      "step": 9990
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001988565589474539,
+      "loss": 1.144,
+      "step": 9995
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019885278060297038,
+      "loss": 1.0354,
+      "step": 10000
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019884899606228195,
+      "loss": 1.0957,
+      "step": 10005
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001988452053256259,
+      "loss": 1.0502,
+      "step": 10010
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019884140839323977,
+      "loss": 0.8835,
+      "step": 10015
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019883760526536161,
+      "loss": 0.9943,
+      "step": 10020
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001988337959422298,
+      "loss": 0.9613,
+      "step": 10025
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019882998042408307,
+      "loss": 0.9662,
+      "step": 10030
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019882615871116062,
+      "loss": 0.8121,
+      "step": 10035
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5,
+      "learning_rate": 0.000198822330803702,
+      "loss": 0.813,
+      "step": 10040
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00019881849670194712,
+      "loss": 0.9225,
+      "step": 10045
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019881465640613635,
+      "loss": 0.9777,
+      "step": 10050
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019881080991651033,
+      "loss": 0.9568,
+      "step": 10055
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019880695723331024,
+      "loss": 0.9948,
+      "step": 10060
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019880309835677753,
+      "loss": 1.3049,
+      "step": 10065
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019879923328715407,
+      "loss": 0.945,
+      "step": 10070
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001987953620246822,
+      "loss": 1.1085,
+      "step": 10075
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019879148456960447,
+      "loss": 1.068,
+      "step": 10080
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.000198787600922164,
+      "loss": 0.9367,
+      "step": 10085
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019878371108260416,
+      "loss": 1.0229,
+      "step": 10090
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019877981505116884,
+      "loss": 0.9013,
+      "step": 10095
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001987759128281022,
+      "loss": 1.0527,
+      "step": 10100
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019877200441364884,
+      "loss": 1.0512,
+      "step": 10105
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019876808980805375,
+      "loss": 0.9033,
+      "step": 10110
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001987641690115623,
+      "loss": 0.9431,
+      "step": 10115
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019876024202442028,
+      "loss": 1.0798,
+      "step": 10120
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001987563088468738,
+      "loss": 1.0832,
+      "step": 10125
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001987523694791694,
+      "loss": 1.0411,
+      "step": 10130
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.000198748423921554,
+      "loss": 0.9704,
+      "step": 10135
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019874447217427493,
+      "loss": 1.0958,
+      "step": 10140
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001987405142375799,
+      "loss": 1.013,
+      "step": 10145
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019873655011171698,
+      "loss": 0.7886,
+      "step": 10150
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019873257979693463,
+      "loss": 0.8691,
+      "step": 10155
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019872860329348173,
+      "loss": 0.9006,
+      "step": 10160
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019872462060160754,
+      "loss": 0.8719,
+      "step": 10165
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001987206317215617,
+      "loss": 0.8805,
+      "step": 10170
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019871663665359422,
+      "loss": 0.8455,
+      "step": 10175
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019871263539795555,
+      "loss": 1.0638,
+      "step": 10180
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019870862795489646,
+      "loss": 0.9571,
+      "step": 10185
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019870461432466814,
+      "loss": 0.9332,
+      "step": 10190
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019870059450752216,
+      "loss": 1.0385,
+      "step": 10195
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019869656850371055,
+      "loss": 1.0117,
+      "step": 10200
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019869253631348558,
+      "loss": 0.9248,
+      "step": 10205
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00019868849793710006,
+      "loss": 0.9542,
+      "step": 10210
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019868445337480707,
+      "loss": 0.9228,
+      "step": 10215
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019868040262686017,
+      "loss": 0.8607,
+      "step": 10220
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019867634569351324,
+      "loss": 0.9858,
+      "step": 10225
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019867228257502056,
+      "loss": 0.9653,
+      "step": 10230
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019866821327163689,
+      "loss": 0.9971,
+      "step": 10235
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019866413778361718,
+      "loss": 1.0423,
+      "step": 10240
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019866005611121694,
+      "loss": 1.0405,
+      "step": 10245
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019865596825469206,
+      "loss": 1.1111,
+      "step": 10250
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019865187421429868,
+      "loss": 0.8356,
+      "step": 10255
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019864777399029353,
+      "loss": 0.8764,
+      "step": 10260
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019864366758293352,
+      "loss": 0.9006,
+      "step": 10265
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001986395549924761,
+      "loss": 0.9196,
+      "step": 10270
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019863543621917898,
+      "loss": 1.0036,
+      "step": 10275
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019863131126330043,
+      "loss": 0.8909,
+      "step": 10280
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019862718012509897,
+      "loss": 1.0645,
+      "step": 10285
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019862304280483347,
+      "loss": 1.0257,
+      "step": 10290
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019861889930276338,
+      "loss": 0.9767,
+      "step": 10295
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019861474961914834,
+      "loss": 0.9614,
+      "step": 10300
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019861059375424848,
+      "loss": 0.9972,
+      "step": 10305
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019860643170832432,
+      "loss": 0.9258,
+      "step": 10310
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001986022634816367,
+      "loss": 1.049,
+      "step": 10315
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001985980890744469,
+      "loss": 1.1258,
+      "step": 10320
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001985939084870166,
+      "loss": 1.0743,
+      "step": 10325
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019858972171960782,
+      "loss": 0.882,
+      "step": 10330
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019858552877248298,
+      "loss": 1.1144,
+      "step": 10335
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019858132964590495,
+      "loss": 1.0556,
+      "step": 10340
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019857712434013687,
+      "loss": 1.0406,
+      "step": 10345
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019857291285544238,
+      "loss": 0.9308,
+      "step": 10350
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019856869519208544,
+      "loss": 0.9467,
+      "step": 10355
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019856447135033046,
+      "loss": 0.9573,
+      "step": 10360
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019856024133044214,
+      "loss": 1.1789,
+      "step": 10365
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001985560051326856,
+      "loss": 0.9767,
+      "step": 10370
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019855176275732647,
+      "loss": 0.9285,
+      "step": 10375
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001985475142046306,
+      "loss": 0.999,
+      "step": 10380
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019854325947486428,
+      "loss": 1.1369,
+      "step": 10385
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019853899856829424,
+      "loss": 0.9692,
+      "step": 10390
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019853473148518755,
+      "loss": 0.9785,
+      "step": 10395
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019853045822581166,
+      "loss": 1.0419,
+      "step": 10400
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019852617879043442,
+      "loss": 1.0083,
+      "step": 10405
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001985218931793241,
+      "loss": 1.0458,
+      "step": 10410
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019851760139274932,
+      "loss": 0.999,
+      "step": 10415
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019851330343097903,
+      "loss": 0.8451,
+      "step": 10420
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019850899929428276,
+      "loss": 1.2424,
+      "step": 10425
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019850468898293015,
+      "loss": 0.8905,
+      "step": 10430
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019850037249719149,
+      "loss": 0.9937,
+      "step": 10435
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001984960498373373,
+      "loss": 0.9661,
+      "step": 10440
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019849172100363851,
+      "loss": 1.0552,
+      "step": 10445
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019848738599636647,
+      "loss": 0.9607,
+      "step": 10450
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019848304481579293,
+      "loss": 0.8086,
+      "step": 10455
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019847869746218995,
+      "loss": 0.9596,
+      "step": 10460
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019847434393583006,
+      "loss": 0.9708,
+      "step": 10465
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019846998423698616,
+      "loss": 0.9375,
+      "step": 10470
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019846561836593148,
+      "loss": 1.0335,
+      "step": 10475
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019846124632293973,
+      "loss": 0.9171,
+      "step": 10480
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019845686810828487,
+      "loss": 0.9829,
+      "step": 10485
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019845248372224144,
+      "loss": 0.9931,
+      "step": 10490
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019844809316508418,
+      "loss": 0.9137,
+      "step": 10495
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019844369643708828,
+      "loss": 0.9316,
+      "step": 10500
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019843929353852944,
+      "loss": 0.9184,
+      "step": 10505
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001984348844696835,
+      "loss": 0.9606,
+      "step": 10510
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019843046923082692,
+      "loss": 0.9036,
+      "step": 10515
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019842604782223643,
+      "loss": 0.9305,
+      "step": 10520
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019842162024418918,
+      "loss": 0.8998,
+      "step": 10525
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019841718649696267,
+      "loss": 1.0292,
+      "step": 10530
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019841274658083483,
+      "loss": 1.0275,
+      "step": 10535
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019840830049608395,
+      "loss": 0.8424,
+      "step": 10540
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019840384824298867,
+      "loss": 0.9166,
+      "step": 10545
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019839938982182815,
+      "loss": 1.1204,
+      "step": 10550
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019839492523288183,
+      "loss": 0.8521,
+      "step": 10555
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001983904544764295,
+      "loss": 0.9154,
+      "step": 10560
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019838597755275143,
+      "loss": 1.0765,
+      "step": 10565
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019838149446212825,
+      "loss": 0.899,
+      "step": 10570
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019837700520484094,
+      "loss": 0.8723,
+      "step": 10575
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001983725097811709,
+      "loss": 1.09,
+      "step": 10580
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001983680081913999,
+      "loss": 1.0788,
+      "step": 10585
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019836350043581013,
+      "loss": 0.9981,
+      "step": 10590
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001983589865146841,
+      "loss": 0.9492,
+      "step": 10595
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019835446642830484,
+      "loss": 1.0889,
+      "step": 10600
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019834994017695556,
+      "loss": 1.0074,
+      "step": 10605
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019834540776092,
+      "loss": 1.0849,
+      "step": 10610
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001983408691804823,
+      "loss": 1.0903,
+      "step": 10615
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019833632443592693,
+      "loss": 1.02,
+      "step": 10620
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019833177352753873,
+      "loss": 0.9941,
+      "step": 10625
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.000198327216455603,
+      "loss": 1.003,
+      "step": 10630
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019832265322040533,
+      "loss": 0.9198,
+      "step": 10635
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019831808382223177,
+      "loss": 1.0734,
+      "step": 10640
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019831350826136877,
+      "loss": 1.1687,
+      "step": 10645
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019830892653810306,
+      "loss": 1.0696,
+      "step": 10650
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019830433865272192,
+      "loss": 1.0275,
+      "step": 10655
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019829974460551286,
+      "loss": 1.0318,
+      "step": 10660
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019829514439676383,
+      "loss": 0.8166,
+      "step": 10665
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019829053802676322,
+      "loss": 1.0499,
+      "step": 10670
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019828592549579974,
+      "loss": 1.0607,
+      "step": 10675
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001982813068041625,
+      "loss": 1.0365,
+      "step": 10680
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019827668195214105,
+      "loss": 0.9428,
+      "step": 10685
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019827205094002524,
+      "loss": 0.9886,
+      "step": 10690
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019826741376810533,
+      "loss": 1.0607,
+      "step": 10695
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.000198262770436672,
+      "loss": 1.0827,
+      "step": 10700
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019825812094601632,
+      "loss": 0.9896,
+      "step": 10705
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001982534652964297,
+      "loss": 0.9955,
+      "step": 10710
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.000198248803488204,
+      "loss": 0.8947,
+      "step": 10715
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019824413552163141,
+      "loss": 1.035,
+      "step": 10720
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001982394613970045,
+      "loss": 0.9582,
+      "step": 10725
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019823478111461625,
+      "loss": 1.1495,
+      "step": 10730
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019823009467476006,
+      "loss": 0.8319,
+      "step": 10735
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019822540207772965,
+      "loss": 0.9966,
+      "step": 10740
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019822070332381917,
+      "loss": 0.9219,
+      "step": 10745
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019821599841332314,
+      "loss": 0.9079,
+      "step": 10750
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001982112873465365,
+      "loss": 0.9619,
+      "step": 10755
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019820657012375445,
+      "loss": 0.9418,
+      "step": 10760
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019820184674527278,
+      "loss": 1.1247,
+      "step": 10765
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019819711721138751,
+      "loss": 1.0533,
+      "step": 10770
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001981923815223951,
+      "loss": 0.879,
+      "step": 10775
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001981876396785924,
+      "loss": 1.1022,
+      "step": 10780
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001981828916802766,
+      "loss": 0.9689,
+      "step": 10785
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001981781375277453,
+      "loss": 0.852,
+      "step": 10790
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019817337722129657,
+      "loss": 1.0256,
+      "step": 10795
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019816861076122873,
+      "loss": 0.9881,
+      "step": 10800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019816383814784055,
+      "loss": 0.9893,
+      "step": 10805
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001981590593814312,
+      "loss": 1.0052,
+      "step": 10810
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019815427446230022,
+      "loss": 1.0588,
+      "step": 10815
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019814948339074752,
+      "loss": 0.9212,
+      "step": 10820
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019814468616707344,
+      "loss": 1.0306,
+      "step": 10825
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019813988279157862,
+      "loss": 0.963,
+      "step": 10830
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001981350732645642,
+      "loss": 1.1285,
+      "step": 10835
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001981302575863316,
+      "loss": 0.8325,
+      "step": 10840
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019812543575718273,
+      "loss": 0.8524,
+      "step": 10845
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019812060777741976,
+      "loss": 0.9555,
+      "step": 10850
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019811577364734536,
+      "loss": 0.8622,
+      "step": 10855
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001981109333672625,
+      "loss": 0.9831,
+      "step": 10860
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019810608693747464,
+      "loss": 0.8993,
+      "step": 10865
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001981012343582855,
+      "loss": 0.905,
+      "step": 10870
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019809637562999925,
+      "loss": 1.0481,
+      "step": 10875
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019809151075292046,
+      "loss": 0.9381,
+      "step": 10880
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019808663972735406,
+      "loss": 0.9586,
+      "step": 10885
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019808176255360537,
+      "loss": 0.9594,
+      "step": 10890
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019807687923198007,
+      "loss": 1.0916,
+      "step": 10895
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019807198976278434,
+      "loss": 0.9862,
+      "step": 10900
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00019806709414632457,
+      "loss": 0.8475,
+      "step": 10905
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001980621923829076,
+      "loss": 1.0222,
+      "step": 10910
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019805728447284078,
+      "loss": 0.9598,
+      "step": 10915
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001980523704164317,
+      "loss": 0.9676,
+      "step": 10920
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019804745021398835,
+      "loss": 0.9503,
+      "step": 10925
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00019804252386581913,
+      "loss": 0.967,
+      "step": 10930
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001980375913722329,
+      "loss": 0.8711,
+      "step": 10935
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019803265273353877,
+      "loss": 0.8866,
+      "step": 10940
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001980277079500463,
+      "loss": 1.1107,
+      "step": 10945
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019802275702206546,
+      "loss": 1.0041,
+      "step": 10950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019801779994990657,
+      "loss": 0.9462,
+      "step": 10955
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019801283673388036,
+      "loss": 0.8799,
+      "step": 10960
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019800786737429792,
+      "loss": 1.0314,
+      "step": 10965
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019800289187147068,
+      "loss": 0.9071,
+      "step": 10970
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001979979102257106,
+      "loss": 1.0473,
+      "step": 10975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019799292243732994,
+      "loss": 0.8929,
+      "step": 10980
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019798792850664123,
+      "loss": 0.9483,
+      "step": 10985
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001979829284339576,
+      "loss": 1.109,
+      "step": 10990
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019797792221959242,
+      "loss": 0.9952,
+      "step": 10995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019797290986385945,
+      "loss": 0.9381,
+      "step": 11000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019796789136707296,
+      "loss": 1.0971,
+      "step": 11005
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001979628667295474,
+      "loss": 0.9669,
+      "step": 11010
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00019795783595159784,
+      "loss": 1.0967,
+      "step": 11015
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019795279903353955,
+      "loss": 0.9798,
+      "step": 11020
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019794775597568824,
+      "loss": 0.9328,
+      "step": 11025
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019794270677836004,
+      "loss": 0.8322,
+      "step": 11030
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019793765144187145,
+      "loss": 0.8911,
+      "step": 11035
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001979325899665393,
+      "loss": 0.9504,
+      "step": 11040
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019792752235268086,
+      "loss": 1.0508,
+      "step": 11045
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019792244860061383,
+      "loss": 0.9632,
+      "step": 11050
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019791736871065617,
+      "loss": 1.0252,
+      "step": 11055
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019791228268312635,
+      "loss": 0.9486,
+      "step": 11060
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001979071905183431,
+      "loss": 1.2584,
+      "step": 11065
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019790209221662566,
+      "loss": 0.8252,
+      "step": 11070
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019789698777829357,
+      "loss": 0.9679,
+      "step": 11075
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001978918772036668,
+      "loss": 1.0461,
+      "step": 11080
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001978867604930657,
+      "loss": 1.0895,
+      "step": 11085
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019788163764681093,
+      "loss": 1.0478,
+      "step": 11090
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019787650866522364,
+      "loss": 0.9994,
+      "step": 11095
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019787137354862532,
+      "loss": 1.0482,
+      "step": 11100
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019786623229733785,
+      "loss": 1.0325,
+      "step": 11105
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019786108491168347,
+      "loss": 0.9032,
+      "step": 11110
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019785593139198484,
+      "loss": 0.9816,
+      "step": 11115
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019785077173856496,
+      "loss": 0.9775,
+      "step": 11120
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00019784560595174732,
+      "loss": 0.9643,
+      "step": 11125
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019784043403185558,
+      "loss": 0.98,
+      "step": 11130
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019783525597921408,
+      "loss": 1.0203,
+      "step": 11135
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00019783007179414728,
+      "loss": 1.126,
+      "step": 11140
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019782488147698015,
+      "loss": 1.0501,
+      "step": 11145
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019781968502803805,
+      "loss": 0.8757,
+      "step": 11150
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019781448244764665,
+      "loss": 0.9667,
+      "step": 11155
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019780927373613217,
+      "loss": 0.8694,
+      "step": 11160
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019780405889382094,
+      "loss": 1.0174,
+      "step": 11165
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019779883792103996,
+      "loss": 1.0074,
+      "step": 11170
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001977936108181164,
+      "loss": 0.9958,
+      "step": 11175
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00019778837758537793,
+      "loss": 1.0806,
+      "step": 11180
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001977831382231526,
+      "loss": 1.0919,
+      "step": 11185
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001977778927317688,
+      "loss": 1.0725,
+      "step": 11190
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019777264111155534,
+      "loss": 1.0368,
+      "step": 11195
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019776738336284134,
+      "loss": 0.9784,
+      "step": 11200
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019776211948595646,
+      "loss": 1.0625,
+      "step": 11205
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019775684948123052,
+      "loss": 1.036,
+      "step": 11210
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00019775157334899396,
+      "loss": 0.8934,
+      "step": 11215
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019774629108957746,
+      "loss": 0.9584,
+      "step": 11220
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001977410027033121,
+      "loss": 1.0195,
+      "step": 11225
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019773570819052938,
+      "loss": 1.1691,
+      "step": 11230
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019773040755156115,
+      "loss": 0.9637,
+      "step": 11235
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019772510078673965,
+      "loss": 1.0448,
+      "step": 11240
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019771978789639758,
+      "loss": 0.9052,
+      "step": 11245
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019771446888086787,
+      "loss": 1.0487,
+      "step": 11250
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019770914374048397,
+      "loss": 1.0113,
+      "step": 11255
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019770381247557968,
+      "loss": 0.9715,
+      "step": 11260
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019769847508648911,
+      "loss": 0.8075,
+      "step": 11265
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001976931315735469,
+      "loss": 0.871,
+      "step": 11270
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019768778193708793,
+      "loss": 1.0295,
+      "step": 11275
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001976824261774475,
+      "loss": 1.0336,
+      "step": 11280
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001976770642949614,
+      "loss": 1.0173,
+      "step": 11285
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019767169628996558,
+      "loss": 0.9169,
+      "step": 11290
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019766632216279664,
+      "loss": 1.0464,
+      "step": 11295
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001976609419137914,
+      "loss": 1.0582,
+      "step": 11300
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019765555554328713,
+      "loss": 0.9371,
+      "step": 11305
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001976501630516214,
+      "loss": 0.9651,
+      "step": 11310
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019764476443913221,
+      "loss": 1.0393,
+      "step": 11315
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00019763935970615798,
+      "loss": 1.0131,
+      "step": 11320
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019763394885303747,
+      "loss": 1.0477,
+      "step": 11325
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019762853188010987,
+      "loss": 0.9775,
+      "step": 11330
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019762310878771472,
+      "loss": 1.0786,
+      "step": 11335
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019761767957619187,
+      "loss": 0.9287,
+      "step": 11340
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019761224424588174,
+      "loss": 1.0053,
+      "step": 11345
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019760680279712497,
+      "loss": 0.963,
+      "step": 11350
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019760135523026263,
+      "loss": 0.9512,
+      "step": 11355
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019759590154563617,
+      "loss": 0.987,
+      "step": 11360
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019759044174358744,
+      "loss": 0.9325,
+      "step": 11365
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001975849758244587,
+      "loss": 0.9854,
+      "step": 11370
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019757950378859252,
+      "loss": 1.0722,
+      "step": 11375
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019757402563633193,
+      "loss": 1.0457,
+      "step": 11380
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019756854136802025,
+      "loss": 0.8436,
+      "step": 11385
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019756305098400133,
+      "loss": 1.1211,
+      "step": 11390
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001975575544846192,
+      "loss": 1.0195,
+      "step": 11395
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019755205187021847,
+      "loss": 1.0107,
+      "step": 11400
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019754654314114404,
+      "loss": 0.9498,
+      "step": 11405
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019754102829774118,
+      "loss": 1.0059,
+      "step": 11410
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019753550734035558,
+      "loss": 0.8315,
+      "step": 11415
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019752998026933326,
+      "loss": 1.0347,
+      "step": 11420
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019752444708502073,
+      "loss": 0.9961,
+      "step": 11425
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001975189077877648,
+      "loss": 1.0512,
+      "step": 11430
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019751336237791263,
+      "loss": 1.0064,
+      "step": 11435
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001975078108558119,
+      "loss": 0.987,
+      "step": 11440
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001975022532218105,
+      "loss": 1.0732,
+      "step": 11445
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019749668947625682,
+      "loss": 0.9559,
+      "step": 11450
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001974911196194996,
+      "loss": 0.9522,
+      "step": 11455
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00019748554365188798,
+      "loss": 0.9707,
+      "step": 11460
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019747996157377148,
+      "loss": 0.8298,
+      "step": 11465
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019747437338549993,
+      "loss": 0.9397,
+      "step": 11470
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019746877908742363,
+      "loss": 0.8971,
+      "step": 11475
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019746317867989327,
+      "loss": 1.0147,
+      "step": 11480
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001974575721632599,
+      "loss": 0.9297,
+      "step": 11485
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019745195953787487,
+      "loss": 0.9276,
+      "step": 11490
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019744634080409005,
+      "loss": 0.8326,
+      "step": 11495
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019744071596225759,
+      "loss": 0.9892,
+      "step": 11500
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001974350850127301,
+      "loss": 1.118,
+      "step": 11505
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001974294479558605,
+      "loss": 0.8706,
+      "step": 11510
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019742380479200212,
+      "loss": 0.9876,
+      "step": 11515
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019741815552150874,
+      "loss": 1.0086,
+      "step": 11520
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001974125001447344,
+      "loss": 0.8989,
+      "step": 11525
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001974068386620336,
+      "loss": 0.7927,
+      "step": 11530
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001974011710737612,
+      "loss": 1.1665,
+      "step": 11535
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019739549738027252,
+      "loss": 0.921,
+      "step": 11540
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001973898175819231,
+      "loss": 1.1424,
+      "step": 11545
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.000197384131679069,
+      "loss": 1.1059,
+      "step": 11550
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001973784396720666,
+      "loss": 1.1901,
+      "step": 11555
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019737274156127272,
+      "loss": 0.9919,
+      "step": 11560
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001973670373470445,
+      "loss": 0.9388,
+      "step": 11565
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001973613270297395,
+      "loss": 1.0375,
+      "step": 11570
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019735561060971565,
+      "loss": 0.942,
+      "step": 11575
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001973498880873312,
+      "loss": 1.0162,
+      "step": 11580
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00019734415946294492,
+      "loss": 1.0539,
+      "step": 11585
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019733842473691583,
+      "loss": 1.2076,
+      "step": 11590
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019733268390960344,
+      "loss": 0.9424,
+      "step": 11595
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019732693698136758,
+      "loss": 0.8787,
+      "step": 11600
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019732118395256846,
+      "loss": 1.0662,
+      "step": 11605
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001973154248235667,
+      "loss": 1.0387,
+      "step": 11610
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019730965959472325,
+      "loss": 1.0715,
+      "step": 11615
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019730388826639953,
+      "loss": 0.9547,
+      "step": 11620
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019729811083895723,
+      "loss": 0.8558,
+      "step": 11625
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019729232731275858,
+      "loss": 0.8651,
+      "step": 11630
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019728653768816604,
+      "loss": 1.0039,
+      "step": 11635
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001972807419655425,
+      "loss": 0.8732,
+      "step": 11640
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001972749401452513,
+      "loss": 1.0182,
+      "step": 11645
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019726913222765601,
+      "loss": 0.9004,
+      "step": 11650
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019726331821312077,
+      "loss": 1.0023,
+      "step": 11655
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019725749810200994,
+      "loss": 0.8979,
+      "step": 11660
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019725167189468842,
+      "loss": 1.0709,
+      "step": 11665
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019724583959152128,
+      "loss": 1.1703,
+      "step": 11670
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001972400011928742,
+      "loss": 1.0387,
+      "step": 11675
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019723415669911307,
+      "loss": 1.0629,
+      "step": 11680
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001972283061106043,
+      "loss": 0.961,
+      "step": 11685
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019722244942771452,
+      "loss": 1.0338,
+      "step": 11690
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001972165866508109,
+      "loss": 0.9863,
+      "step": 11695
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001972107177802609,
+      "loss": 0.982,
+      "step": 11700
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001972048428164324,
+      "loss": 1.0622,
+      "step": 11705
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019719896175969369,
+      "loss": 0.9321,
+      "step": 11710
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019719307461041333,
+      "loss": 0.9296,
+      "step": 11715
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019718718136896033,
+      "loss": 0.9618,
+      "step": 11720
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019718128203570414,
+      "loss": 1.0673,
+      "step": 11725
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001971753766110145,
+      "loss": 0.816,
+      "step": 11730
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019716946509526156,
+      "loss": 0.9263,
+      "step": 11735
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019716354748881593,
+      "loss": 1.0468,
+      "step": 11740
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019715762379204845,
+      "loss": 0.867,
+      "step": 11745
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019715169400533048,
+      "loss": 1.0149,
+      "step": 11750
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001971457581290337,
+      "loss": 0.7832,
+      "step": 11755
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001971398161635301,
+      "loss": 1.0403,
+      "step": 11760
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00019713386810919222,
+      "loss": 1.0341,
+      "step": 11765
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019712791396639287,
+      "loss": 0.9868,
+      "step": 11770
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019712195373550526,
+      "loss": 1.0789,
+      "step": 11775
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019711598741690295,
+      "loss": 1.1115,
+      "step": 11780
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019711001501096,
+      "loss": 1.047,
+      "step": 11785
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019710403651805063,
+      "loss": 1.0612,
+      "step": 11790
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001970980519385497,
+      "loss": 1.044,
+      "step": 11795
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001970920612728323,
+      "loss": 1.0137,
+      "step": 11800
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001970860645212739,
+      "loss": 1.017,
+      "step": 11805
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001970800616842504,
+      "loss": 1.0174,
+      "step": 11810
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019707405276213807,
+      "loss": 1.0831,
+      "step": 11815
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019706803775531358,
+      "loss": 1.0993,
+      "step": 11820
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001970620166641539,
+      "loss": 1.0144,
+      "step": 11825
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00019705598948903649,
+      "loss": 0.8345,
+      "step": 11830
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019704995623033914,
+      "loss": 0.954,
+      "step": 11835
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019704391688843997,
+      "loss": 0.9742,
+      "step": 11840
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019703787146371757,
+      "loss": 0.858,
+      "step": 11845
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001970318199565509,
+      "loss": 0.9053,
+      "step": 11850
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001970257623673192,
+      "loss": 0.9909,
+      "step": 11855
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001970196986964022,
+      "loss": 1.0612,
+      "step": 11860
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019701362894418005,
+      "loss": 0.9934,
+      "step": 11865
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019700755311103315,
+      "loss": 0.9206,
+      "step": 11870
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001970014711973423,
+      "loss": 0.8881,
+      "step": 11875
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019699538320348876,
+      "loss": 0.99,
+      "step": 11880
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019698928912985415,
+      "loss": 1.1218,
+      "step": 11885
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019698318897682043,
+      "loss": 0.842,
+      "step": 11890
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00019697708274476995,
+      "loss": 0.9669,
+      "step": 11895
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001969709704340855,
+      "loss": 1.0681,
+      "step": 11900
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001969648520451502,
+      "loss": 0.9509,
+      "step": 11905
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001969587275783475,
+      "loss": 0.9891,
+      "step": 11910
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019695259703406134,
+      "loss": 0.9673,
+      "step": 11915
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019694646041267598,
+      "loss": 1.0024,
+      "step": 11920
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019694031771457608,
+      "loss": 1.1921,
+      "step": 11925
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019693416894014666,
+      "loss": 0.9211,
+      "step": 11930
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019692801408977313,
+      "loss": 0.9258,
+      "step": 11935
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019692185316384127,
+      "loss": 1.2574,
+      "step": 11940
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001969156861627373,
+      "loss": 1.0631,
+      "step": 11945
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001969095130868477,
+      "loss": 0.9337,
+      "step": 11950
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00019690333393655948,
+      "loss": 1.0891,
+      "step": 11955
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001968971487122599,
+      "loss": 0.8698,
+      "step": 11960
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019689095741433672,
+      "loss": 1.0847,
+      "step": 11965
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019688476004317795,
+      "loss": 1.0954,
+      "step": 11970
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019687855659917207,
+      "loss": 1.0362,
+      "step": 11975
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00019687234708270793,
+      "loss": 0.9058,
+      "step": 11980
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019686613149417478,
+      "loss": 0.9234,
+      "step": 11985
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019685990983396215,
+      "loss": 0.9797,
+      "step": 11990
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019685368210246008,
+      "loss": 0.9666,
+      "step": 11995
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019684744830005893,
+      "loss": 0.9774,
+      "step": 12000
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019684120842714938,
+      "loss": 1.1606,
+      "step": 12005
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001968349624841226,
+      "loss": 0.9442,
+      "step": 12010
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001968287104713701,
+      "loss": 1.0067,
+      "step": 12015
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019682245238928373,
+      "loss": 1.0373,
+      "step": 12020
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019681618823825581,
+      "loss": 1.0156,
+      "step": 12025
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019680991801867892,
+      "loss": 0.9226,
+      "step": 12030
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019680364173094609,
+      "loss": 0.9687,
+      "step": 12035
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019679735937545075,
+      "loss": 0.9303,
+      "step": 12040
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019679107095258672,
+      "loss": 0.9594,
+      "step": 12045
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001967847764627481,
+      "loss": 0.9388,
+      "step": 12050
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019677847590632947,
+      "loss": 0.8384,
+      "step": 12055
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019677216928372574,
+      "loss": 1.0197,
+      "step": 12060
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019676585659533224,
+      "loss": 0.9054,
+      "step": 12065
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00019675953784154465,
+      "loss": 0.8919,
+      "step": 12070
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.000196753213022759,
+      "loss": 0.9564,
+      "step": 12075
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00019674688213937178,
+      "loss": 0.9152,
+      "step": 12080
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019674054519177981,
+      "loss": 1.0695,
+      "step": 12085
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019673420218038027,
+      "loss": 1.0131,
+      "step": 12090
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019672785310557078,
+      "loss": 1.0804,
+      "step": 12095
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019672149796774927,
+      "loss": 0.9286,
+      "step": 12100
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019671513676731414,
+      "loss": 0.992,
+      "step": 12105
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001967087695046641,
+      "loss": 1.0514,
+      "step": 12110
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001967023961801982,
+      "loss": 1.1135,
+      "step": 12115
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.000196696016794316,
+      "loss": 0.8374,
+      "step": 12120
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019668963134741733,
+      "loss": 1.0395,
+      "step": 12125
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019668323983990246,
+      "loss": 0.776,
+      "step": 12130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.000196676842272172,
+      "loss": 0.9524,
+      "step": 12135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019667043864462694,
+      "loss": 0.8601,
+      "step": 12140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019666402895766871,
+      "loss": 0.98,
+      "step": 12145
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019665761321169907,
+      "loss": 0.921,
+      "step": 12150
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001966511914071201,
+      "loss": 0.9572,
+      "step": 12155
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019664476354433444,
+      "loss": 0.9071,
+      "step": 12160
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019663832962374487,
+      "loss": 0.9123,
+      "step": 12165
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019663188964575475,
+      "loss": 0.8975,
+      "step": 12170
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019662544361076774,
+      "loss": 0.9547,
+      "step": 12175
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001966189915191879,
+      "loss": 1.0201,
+      "step": 12180
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019661253337141963,
+      "loss": 0.985,
+      "step": 12185
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9765625,
+      "learning_rate": 0.00019660606916786772,
+      "loss": 0.9262,
+      "step": 12190
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019659959890893736,
+      "loss": 0.9385,
+      "step": 12195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019659312259503412,
+      "loss": 0.9243,
+      "step": 12200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019658664022656396,
+      "loss": 0.9716,
+      "step": 12205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019658015180393318,
+      "loss": 1.0187,
+      "step": 12210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001965736573275485,
+      "loss": 0.9634,
+      "step": 12215
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019656715679781698,
+      "loss": 0.9778,
+      "step": 12220
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001965606502151461,
+      "loss": 0.8875,
+      "step": 12225
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001965541375799437,
+      "loss": 0.9571,
+      "step": 12230
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.000196547618892618,
+      "loss": 0.9802,
+      "step": 12235
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019654109415357754,
+      "loss": 1.049,
+      "step": 12240
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001965345633632314,
+      "loss": 1.0749,
+      "step": 12245
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019652802652198886,
+      "loss": 0.8459,
+      "step": 12250
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019652148363025966,
+      "loss": 1.0139,
+      "step": 12255
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019651493468845398,
+      "loss": 0.8877,
+      "step": 12260
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019650837969698224,
+      "loss": 0.9687,
+      "step": 12265
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019650181865625536,
+      "loss": 0.9932,
+      "step": 12270
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001964952515666846,
+      "loss": 0.886,
+      "step": 12275
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019648867842868154,
+      "loss": 0.963,
+      "step": 12280
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019648209924265825,
+      "loss": 1.1426,
+      "step": 12285
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019647551400902704,
+      "loss": 0.9924,
+      "step": 12290
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019646892272820078,
+      "loss": 1.0424,
+      "step": 12295
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019646232540059257,
+      "loss": 1.1567,
+      "step": 12300
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019645572202661596,
+      "loss": 1.0748,
+      "step": 12305
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019644911260668482,
+      "loss": 0.9622,
+      "step": 12310
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019644249714121346,
+      "loss": 0.9486,
+      "step": 12315
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019643587563061653,
+      "loss": 0.9666,
+      "step": 12320
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001964292480753091,
+      "loss": 1.1216,
+      "step": 12325
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019642261447570656,
+      "loss": 1.001,
+      "step": 12330
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019641597483222476,
+      "loss": 0.9032,
+      "step": 12335
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001964093291452798,
+      "loss": 0.863,
+      "step": 12340
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019640267741528834,
+      "loss": 1.0108,
+      "step": 12345
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019639601964266726,
+      "loss": 1.112,
+      "step": 12350
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019638935582783385,
+      "loss": 0.9948,
+      "step": 12355
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019638268597120585,
+      "loss": 0.9937,
+      "step": 12360
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001963760100732013,
+      "loss": 1.0482,
+      "step": 12365
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019636932813423875,
+      "loss": 0.9995,
+      "step": 12370
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0001963626401547369,
+      "loss": 0.9609,
+      "step": 12375
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.000196355946135115,
+      "loss": 1.0666,
+      "step": 12380
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001963492460757927,
+      "loss": 0.8948,
+      "step": 12385
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001963425399771899,
+      "loss": 0.9701,
+      "step": 12390
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019633582783972698,
+      "loss": 0.9144,
+      "step": 12395
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019632910966382465,
+      "loss": 0.9507,
+      "step": 12400
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019632238544990403,
+      "loss": 1.0839,
+      "step": 12405
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019631565519838656,
+      "loss": 1.1661,
+      "step": 12410
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019630891890969415,
+      "loss": 1.0304,
+      "step": 12415
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.000196302176584249,
+      "loss": 0.8476,
+      "step": 12420
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019629542822247373,
+      "loss": 1.0206,
+      "step": 12425
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019628867382479138,
+      "loss": 0.9953,
+      "step": 12430
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001962819133916253,
+      "loss": 0.9312,
+      "step": 12435
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001962751469233992,
+      "loss": 1.1058,
+      "step": 12440
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019626837442053726,
+      "loss": 0.9904,
+      "step": 12445
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019626159588346392,
+      "loss": 0.8433,
+      "step": 12450
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019625481131260418,
+      "loss": 0.9796,
+      "step": 12455
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019624802070838325,
+      "loss": 0.9605,
+      "step": 12460
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001962412240712267,
+      "loss": 1.0074,
+      "step": 12465
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019623442140156066,
+      "loss": 0.9153,
+      "step": 12470
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001962276126998115,
+      "loss": 0.9167,
+      "step": 12475
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00019622079796640597,
+      "loss": 1.0444,
+      "step": 12480
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001962139772017712,
+      "loss": 1.0472,
+      "step": 12485
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001962071504063348,
+      "loss": 0.9135,
+      "step": 12490
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019620031758052465,
+      "loss": 1.1986,
+      "step": 12495
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.000196193478724769,
+      "loss": 0.8138,
+      "step": 12500
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019618663383949656,
+      "loss": 0.8546,
+      "step": 12505
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019617978292513634,
+      "loss": 1.0783,
+      "step": 12510
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019617292598211782,
+      "loss": 0.8674,
+      "step": 12515
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019616606301087078,
+      "loss": 0.9813,
+      "step": 12520
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00019615919401182535,
+      "loss": 0.9427,
+      "step": 12525
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019615231898541214,
+      "loss": 1.0899,
+      "step": 12530
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019614543793206206,
+      "loss": 1.0309,
+      "step": 12535
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019613855085220646,
+      "loss": 1.0088,
+      "step": 12540
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019613165774627696,
+      "loss": 0.92,
+      "step": 12545
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001961247586147057,
+      "loss": 0.8939,
+      "step": 12550
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019611785345792507,
+      "loss": 0.9448,
+      "step": 12555
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001961109422763679,
+      "loss": 0.9925,
+      "step": 12560
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019610402507046744,
+      "loss": 1.0034,
+      "step": 12565
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019609710184065722,
+      "loss": 0.9819,
+      "step": 12570
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001960901725873712,
+      "loss": 0.9709,
+      "step": 12575
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019608323731104375,
+      "loss": 0.8864,
+      "step": 12580
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019607629601210954,
+      "loss": 0.8654,
+      "step": 12585
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001960693486910037,
+      "loss": 0.9974,
+      "step": 12590
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019606239534816165,
+      "loss": 0.9497,
+      "step": 12595
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019605543598401925,
+      "loss": 1.0166,
+      "step": 12600
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001960484705990127,
+      "loss": 0.8977,
+      "step": 12605
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001960414991935787,
+      "loss": 1.0628,
+      "step": 12610
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00019603452176815406,
+      "loss": 1.0787,
+      "step": 12615
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019602753832317628,
+      "loss": 0.9722,
+      "step": 12620
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.000196020548859083,
+      "loss": 0.9783,
+      "step": 12625
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019601355337631237,
+      "loss": 0.9863,
+      "step": 12630
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019600655187530287,
+      "loss": 0.9108,
+      "step": 12635
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019599954435649337,
+      "loss": 0.925,
+      "step": 12640
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019599253082032308,
+      "loss": 0.9331,
+      "step": 12645
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00019598551126723162,
+      "loss": 1.043,
+      "step": 12650
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.000195978485697659,
+      "loss": 0.8841,
+      "step": 12655
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001959714541120456,
+      "loss": 0.9505,
+      "step": 12660
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019596441651083214,
+      "loss": 0.9653,
+      "step": 12665
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019595737289445977,
+      "loss": 0.9831,
+      "step": 12670
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019595032326336998,
+      "loss": 1.0024,
+      "step": 12675
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019594326761800462,
+      "loss": 1.0317,
+      "step": 12680
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019593620595880602,
+      "loss": 0.9532,
+      "step": 12685
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019592913828621672,
+      "loss": 1.076,
+      "step": 12690
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019592206460067983,
+      "loss": 0.862,
+      "step": 12695
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019591498490263866,
+      "loss": 0.9048,
+      "step": 12700
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.000195907899192537,
+      "loss": 0.9484,
+      "step": 12705
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.000195900807470819,
+      "loss": 0.9225,
+      "step": 12710
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019589370973792914,
+      "loss": 1.0204,
+      "step": 12715
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019588660599431234,
+      "loss": 0.9455,
+      "step": 12720
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001958794962404139,
+      "loss": 0.8872,
+      "step": 12725
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019587238047667942,
+      "loss": 0.9301,
+      "step": 12730
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019586525870355495,
+      "loss": 1.0429,
+      "step": 12735
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019585813092148688,
+      "loss": 1.094,
+      "step": 12740
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00019585099713092198,
+      "loss": 0.9153,
+      "step": 12745
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019584385733230743,
+      "loss": 0.9439,
+      "step": 12750
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019583671152609075,
+      "loss": 1.0327,
+      "step": 12755
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019582955971271981,
+      "loss": 1.1026,
+      "step": 12760
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019582240189264295,
+      "loss": 1.0614,
+      "step": 12765
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001958152380663088,
+      "loss": 0.9044,
+      "step": 12770
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019580806823416642,
+      "loss": 0.9334,
+      "step": 12775
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019580089239666518,
+      "loss": 0.9532,
+      "step": 12780
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019579371055425493,
+      "loss": 1.0215,
+      "step": 12785
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019578652270738577,
+      "loss": 0.9541,
+      "step": 12790
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001957793288565083,
+      "loss": 0.9159,
+      "step": 12795
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001957721290020734,
+      "loss": 1.0008,
+      "step": 12800
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019576492314453238,
+      "loss": 0.9768,
+      "step": 12805
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019575771128433692,
+      "loss": 0.8682,
+      "step": 12810
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019575049342193906,
+      "loss": 0.926,
+      "step": 12815
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019574326955779122,
+      "loss": 1.1345,
+      "step": 12820
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019573603969234618,
+      "loss": 0.8593,
+      "step": 12825
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019572880382605714,
+      "loss": 1.0199,
+      "step": 12830
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019572156195937765,
+      "loss": 1.0294,
+      "step": 12835
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019571431409276166,
+      "loss": 0.9291,
+      "step": 12840
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00019570706022666343,
+      "loss": 0.8817,
+      "step": 12845
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001956998003615377,
+      "loss": 0.9355,
+      "step": 12850
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019569253449783943,
+      "loss": 0.8385,
+      "step": 12855
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019568526263602417,
+      "loss": 1.161,
+      "step": 12860
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019567798477654767,
+      "loss": 1.0588,
+      "step": 12865
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019567070091986607,
+      "loss": 0.8892,
+      "step": 12870
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019566341106643602,
+      "loss": 0.9518,
+      "step": 12875
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001956561152167144,
+      "loss": 0.9279,
+      "step": 12880
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019564881337115852,
+      "loss": 0.9089,
+      "step": 12885
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001956415055302261,
+      "loss": 0.8285,
+      "step": 12890
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019563419169437517,
+      "loss": 1.0355,
+      "step": 12895
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019562687186406422,
+      "loss": 0.9374,
+      "step": 12900
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.000195619546039752,
+      "loss": 0.9838,
+      "step": 12905
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019561221422189775,
+      "loss": 1.0857,
+      "step": 12910
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019560487641096102,
+      "loss": 1.0689,
+      "step": 12915
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019559753260740172,
+      "loss": 0.9646,
+      "step": 12920
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019559018281168025,
+      "loss": 0.9258,
+      "step": 12925
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019558282702425724,
+      "loss": 1.0102,
+      "step": 12930
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019557546524559373,
+      "loss": 1.1096,
+      "step": 12935
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019556809747615125,
+      "loss": 0.8644,
+      "step": 12940
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019556072371639156,
+      "loss": 0.8821,
+      "step": 12945
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00019555334396677688,
+      "loss": 0.9682,
+      "step": 12950
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019554595822776977,
+      "loss": 1.192,
+      "step": 12955
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019553856649983317,
+      "loss": 1.0012,
+      "step": 12960
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019553116878343044,
+      "loss": 1.0151,
+      "step": 12965
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019552376507902524,
+      "loss": 0.904,
+      "step": 12970
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019551635538708162,
+      "loss": 1.1135,
+      "step": 12975
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001955089397080641,
+      "loss": 0.9702,
+      "step": 12980
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00019550151804243745,
+      "loss": 1.0448,
+      "step": 12985
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019549409039066687,
+      "loss": 1.125,
+      "step": 12990
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019548665675321795,
+      "loss": 0.9953,
+      "step": 12995
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019547921713055665,
+      "loss": 1.0935,
+      "step": 13000
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019547177152314928,
+      "loss": 0.9381,
+      "step": 13005
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019546431993146252,
+      "loss": 1.1262,
+      "step": 13010
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001954568623559635,
+      "loss": 0.9953,
+      "step": 13015
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001954493987971196,
+      "loss": 1.2148,
+      "step": 13020
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019544192925539865,
+      "loss": 1.1218,
+      "step": 13025
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001954344537312689,
+      "loss": 0.9282,
+      "step": 13030
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001954269722251989,
+      "loss": 1.0104,
+      "step": 13035
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001954194847376576,
+      "loss": 0.8732,
+      "step": 13040
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019541199126911433,
+      "loss": 1.0397,
+      "step": 13045
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019540449182003878,
+      "loss": 0.9709,
+      "step": 13050
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.000195396986390901,
+      "loss": 0.9555,
+      "step": 13055
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001953894749821715,
+      "loss": 0.9479,
+      "step": 13060
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019538195759432106,
+      "loss": 0.88,
+      "step": 13065
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019537443422782089,
+      "loss": 0.985,
+      "step": 13070
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019536690488314252,
+      "loss": 0.8963,
+      "step": 13075
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.000195359369560758,
+      "loss": 0.9127,
+      "step": 13080
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019535182826113954,
+      "loss": 1.0403,
+      "step": 13085
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019534428098475992,
+      "loss": 0.8949,
+      "step": 13090
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019533672773209217,
+      "loss": 0.8789,
+      "step": 13095
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019532916850360975,
+      "loss": 1.138,
+      "step": 13100
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001953216032997865,
+      "loss": 0.9711,
+      "step": 13105
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019531403212109653,
+      "loss": 1.0461,
+      "step": 13110
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001953064549680145,
+      "loss": 0.8697,
+      "step": 13115
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019529887184101537,
+      "loss": 0.8727,
+      "step": 13120
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019529128274057436,
+      "loss": 0.905,
+      "step": 13125
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019528368766716725,
+      "loss": 1.0066,
+      "step": 13130
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019527608662127009,
+      "loss": 0.9105,
+      "step": 13135
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019526847960335927,
+      "loss": 0.8907,
+      "step": 13140
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019526086661391168,
+      "loss": 1.1167,
+      "step": 13145
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019525324765340445,
+      "loss": 0.9489,
+      "step": 13150
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019524562272231518,
+      "loss": 1.1006,
+      "step": 13155
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019523799182112183,
+      "loss": 1.1097,
+      "step": 13160
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019523035495030267,
+      "loss": 0.9017,
+      "step": 13165
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001952227121103364,
+      "loss": 0.8943,
+      "step": 13170
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019521506330170206,
+      "loss": 1.0787,
+      "step": 13175
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019520740852487916,
+      "loss": 0.9864,
+      "step": 13180
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019519974778034742,
+      "loss": 1.0124,
+      "step": 13185
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019519208106858708,
+      "loss": 1.1291,
+      "step": 13190
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.0001951844083900787,
+      "loss": 1.0683,
+      "step": 13195
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001951767297453032,
+      "loss": 0.9281,
+      "step": 13200
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019516904513474187,
+      "loss": 1.0935,
+      "step": 13205
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019516135455887638,
+      "loss": 0.951,
+      "step": 13210
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019515365801818884,
+      "loss": 0.9578,
+      "step": 13215
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019514595551316162,
+      "loss": 1.0304,
+      "step": 13220
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019513824704427756,
+      "loss": 0.8505,
+      "step": 13225
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019513053261201982,
+      "loss": 0.9522,
+      "step": 13230
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019512281221687193,
+      "loss": 0.9674,
+      "step": 13235
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019511508585931786,
+      "loss": 1.0524,
+      "step": 13240
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019510735353984185,
+      "loss": 0.874,
+      "step": 13245
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019509961525892867,
+      "loss": 1.0424,
+      "step": 13250
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019509187101706324,
+      "loss": 0.7539,
+      "step": 13255
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019508412081473104,
+      "loss": 0.9879,
+      "step": 13260
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019507636465241785,
+      "loss": 1.2274,
+      "step": 13265
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001950686025306099,
+      "loss": 0.9844,
+      "step": 13270
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001950608344497936,
+      "loss": 1.0499,
+      "step": 13275
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019505306041045595,
+      "loss": 0.9718,
+      "step": 13280
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019504528041308422,
+      "loss": 1.0062,
+      "step": 13285
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019503749445816607,
+      "loss": 0.9244,
+      "step": 13290
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019502970254618954,
+      "loss": 0.875,
+      "step": 13295
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.625,
+      "learning_rate": 0.000195021904677643,
+      "loss": 0.9849,
+      "step": 13300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019501410085301526,
+      "loss": 0.9765,
+      "step": 13305
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001950062910727955,
+      "loss": 0.8658,
+      "step": 13310
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019499847533747315,
+      "loss": 0.8656,
+      "step": 13315
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019499065364753822,
+      "loss": 0.9422,
+      "step": 13320
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001949828260034809,
+      "loss": 1.0623,
+      "step": 13325
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001949749924057919,
+      "loss": 1.1,
+      "step": 13330
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001949671528549622,
+      "loss": 0.8578,
+      "step": 13335
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001949593073514832,
+      "loss": 0.9376,
+      "step": 13340
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019495145589584664,
+      "loss": 0.9657,
+      "step": 13345
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019494359848854474,
+      "loss": 1.0854,
+      "step": 13350
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019493573513006994,
+      "loss": 0.8838,
+      "step": 13355
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019492786582091508,
+      "loss": 0.9277,
+      "step": 13360
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019491999056157352,
+      "loss": 0.9805,
+      "step": 13365
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019491210935253887,
+      "loss": 1.0989,
+      "step": 13370
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019490422219430506,
+      "loss": 1.0407,
+      "step": 13375
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019489632908736658,
+      "loss": 0.9724,
+      "step": 13380
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019488843003221805,
+      "loss": 0.9384,
+      "step": 13385
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019488052502935469,
+      "loss": 1.0806,
+      "step": 13390
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019487261407927196,
+      "loss": 0.9685,
+      "step": 13395
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019486469718246573,
+      "loss": 0.936,
+      "step": 13400
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001948567743394322,
+      "loss": 0.9386,
+      "step": 13405
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019484884555066807,
+      "loss": 0.9172,
+      "step": 13410
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019484091081667027,
+      "loss": 1.1164,
+      "step": 13415
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019483297013793617,
+      "loss": 1.0413,
+      "step": 13420
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019482502351496345,
+      "loss": 1.0837,
+      "step": 13425
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001948170709482503,
+      "loss": 0.9828,
+      "step": 13430
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019480911243829514,
+      "loss": 1.0261,
+      "step": 13435
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019480114798559685,
+      "loss": 1.0415,
+      "step": 13440
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019479317759065463,
+      "loss": 0.8573,
+      "step": 13445
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019478520125396805,
+      "loss": 0.9587,
+      "step": 13450
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019477721897603715,
+      "loss": 1.146,
+      "step": 13455
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019476923075736218,
+      "loss": 0.9267,
+      "step": 13460
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001947612365984439,
+      "loss": 0.979,
+      "step": 13465
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019475323649978344,
+      "loss": 0.9616,
+      "step": 13470
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019474523046188216,
+      "loss": 0.9822,
+      "step": 13475
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019473721848524195,
+      "loss": 0.897,
+      "step": 13480
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019472920057036493,
+      "loss": 0.9629,
+      "step": 13485
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001947211767177538,
+      "loss": 1.0021,
+      "step": 13490
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019471314692791138,
+      "loss": 1.0638,
+      "step": 13495
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019470511120134106,
+      "loss": 0.9434,
+      "step": 13500
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00019469706953854652,
+      "loss": 1.125,
+      "step": 13505
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001946890219400318,
+      "loss": 0.7527,
+      "step": 13510
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019468096840630136,
+      "loss": 0.971,
+      "step": 13515
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019467290893785996,
+      "loss": 0.9305,
+      "step": 13520
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019466484353521284,
+      "loss": 0.9969,
+      "step": 13525
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019465677219886548,
+      "loss": 0.8256,
+      "step": 13530
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019464869492932384,
+      "loss": 1.0183,
+      "step": 13535
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001946406117270942,
+      "loss": 0.9527,
+      "step": 13540
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019463252259268327,
+      "loss": 0.9422,
+      "step": 13545
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.000194624427526598,
+      "loss": 1.123,
+      "step": 13550
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019461632652934587,
+      "loss": 0.9995,
+      "step": 13555
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019460821960143462,
+      "loss": 0.9578,
+      "step": 13560
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00019460010674337243,
+      "loss": 0.9716,
+      "step": 13565
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00019459198795566783,
+      "loss": 1.0666,
+      "step": 13570
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019458386323882966,
+      "loss": 0.9313,
+      "step": 13575
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001945757325933672,
+      "loss": 0.8971,
+      "step": 13580
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019456759601979014,
+      "loss": 0.9537,
+      "step": 13585
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00019455945351860848,
+      "loss": 0.9548,
+      "step": 13590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019455130509033253,
+      "loss": 1.0053,
+      "step": 13595
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019454315073547313,
+      "loss": 1.0101,
+      "step": 13600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019453499045454136,
+      "loss": 0.9598,
+      "step": 13605
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019452682424804868,
+      "loss": 0.8437,
+      "step": 13610
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019451865211650703,
+      "loss": 0.8778,
+      "step": 13615
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019451047406042864,
+      "loss": 0.9147,
+      "step": 13620
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019450229008032607,
+      "loss": 1.0566,
+      "step": 13625
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019449410017671236,
+      "loss": 1.0564,
+      "step": 13630
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001944859043501008,
+      "loss": 1.0702,
+      "step": 13635
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019447770260100513,
+      "loss": 0.8292,
+      "step": 13640
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001944694949299395,
+      "loss": 0.9116,
+      "step": 13645
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001944612813374183,
+      "loss": 0.8663,
+      "step": 13650
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019445306182395644,
+      "loss": 1.0415,
+      "step": 13655
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019444483639006904,
+      "loss": 0.8293,
+      "step": 13660
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001944366050362718,
+      "loss": 0.9584,
+      "step": 13665
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.80859375,
+      "learning_rate": 0.00019442836776308056,
+      "loss": 1.0026,
+      "step": 13670
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019442012457101168,
+      "loss": 0.8867,
+      "step": 13675
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019441187546058187,
+      "loss": 0.9783,
+      "step": 13680
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019440362043230818,
+      "loss": 0.871,
+      "step": 13685
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019439535948670805,
+      "loss": 0.9344,
+      "step": 13690
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019438709262429926,
+      "loss": 1.0194,
+      "step": 13695
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00019437881984560003,
+      "loss": 0.9557,
+      "step": 13700
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019437054115112887,
+      "loss": 0.9423,
+      "step": 13705
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019436225654140472,
+      "loss": 1.0194,
+      "step": 13710
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019435396601694686,
+      "loss": 1.0522,
+      "step": 13715
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019434566957827492,
+      "loss": 0.9833,
+      "step": 13720
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.000194337367225909,
+      "loss": 0.8905,
+      "step": 13725
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019432905896036944,
+      "loss": 1.0386,
+      "step": 13730
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.000194320744782177,
+      "loss": 1.0965,
+      "step": 13735
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019431242469185293,
+      "loss": 0.9553,
+      "step": 13740
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019430409868991864,
+      "loss": 0.9008,
+      "step": 13745
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.000194295766776896,
+      "loss": 0.9997,
+      "step": 13750
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019428742895330733,
+      "loss": 0.9672,
+      "step": 13755
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019427908521967523,
+      "loss": 0.9669,
+      "step": 13760
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001942707355765227,
+      "loss": 1.1381,
+      "step": 13765
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001942623800243731,
+      "loss": 1.1075,
+      "step": 13770
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019425401856375016,
+      "loss": 1.0386,
+      "step": 13775
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.000194245651195178,
+      "loss": 0.9591,
+      "step": 13780
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001942372779191811,
+      "loss": 0.9495,
+      "step": 13785
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019422889873628424,
+      "loss": 0.8961,
+      "step": 13790
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019422051364701272,
+      "loss": 0.9591,
+      "step": 13795
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001942121226518921,
+      "loss": 0.887,
+      "step": 13800
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019420372575144833,
+      "loss": 1.0454,
+      "step": 13805
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019419532294620773,
+      "loss": 0.9947,
+      "step": 13810
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019418691423669704,
+      "loss": 0.9633,
+      "step": 13815
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019417849962344327,
+      "loss": 1.0141,
+      "step": 13820
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019417007910697387,
+      "loss": 0.9378,
+      "step": 13825
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019416165268781668,
+      "loss": 0.9427,
+      "step": 13830
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019415322036649986,
+      "loss": 0.8609,
+      "step": 13835
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019414478214355193,
+      "loss": 0.8503,
+      "step": 13840
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019413633801950187,
+      "loss": 1.0469,
+      "step": 13845
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001941278879948789,
+      "loss": 1.0054,
+      "step": 13850
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019411943207021275,
+      "loss": 1.0992,
+      "step": 13855
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001941109702460334,
+      "loss": 1.0576,
+      "step": 13860
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001941025025228712,
+      "loss": 0.9899,
+      "step": 13865
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019409402890125697,
+      "loss": 1.023,
+      "step": 13870
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001940855493817219,
+      "loss": 1.0024,
+      "step": 13875
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001940770639647974,
+      "loss": 0.9097,
+      "step": 13880
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001940685726510154,
+      "loss": 0.9621,
+      "step": 13885
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019406007544090813,
+      "loss": 0.9476,
+      "step": 13890
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001940515723350082,
+      "loss": 0.957,
+      "step": 13895
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019404306333384864,
+      "loss": 1.1225,
+      "step": 13900
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001940345484379627,
+      "loss": 0.9353,
+      "step": 13905
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001940260276478842,
+      "loss": 0.9559,
+      "step": 13910
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019401750096414722,
+      "loss": 1.0461,
+      "step": 13915
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019400896838728618,
+      "loss": 1.0111,
+      "step": 13920
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019400042991783592,
+      "loss": 0.9265,
+      "step": 13925
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00019399188555633168,
+      "loss": 1.1099,
+      "step": 13930
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.000193983335303309,
+      "loss": 1.0102,
+      "step": 13935
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001939747791593038,
+      "loss": 0.9348,
+      "step": 13940
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019396621712485245,
+      "loss": 0.846,
+      "step": 13945
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019395764920049157,
+      "loss": 0.994,
+      "step": 13950
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019394907538675823,
+      "loss": 1.009,
+      "step": 13955
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019394049568418985,
+      "loss": 0.9882,
+      "step": 13960
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001939319100933242,
+      "loss": 0.983,
+      "step": 13965
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001939233186146995,
+      "loss": 0.9918,
+      "step": 13970
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019391472124885418,
+      "loss": 0.9375,
+      "step": 13975
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019390611799632716,
+      "loss": 0.9159,
+      "step": 13980
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019389750885765773,
+      "loss": 1.0269,
+      "step": 13985
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019388889383338548,
+      "loss": 0.9451,
+      "step": 13990
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019388027292405045,
+      "loss": 1.1171,
+      "step": 13995
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.000193871646130193,
+      "loss": 0.9076,
+      "step": 14000
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019386301345235385,
+      "loss": 0.9822,
+      "step": 14005
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001938543748910741,
+      "loss": 1.1278,
+      "step": 14010
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.0001938457304468953,
+      "loss": 1.0708,
+      "step": 14015
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001938370801203592,
+      "loss": 0.9955,
+      "step": 14020
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.000193828423912008,
+      "loss": 0.8778,
+      "step": 14025
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001938197618223844,
+      "loss": 0.9365,
+      "step": 14030
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019381109385203124,
+      "loss": 0.9629,
+      "step": 14035
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001938024200014919,
+      "loss": 0.9758,
+      "step": 14040
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00019379374027131,
+      "loss": 0.9081,
+      "step": 14045
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00019378505466202968,
+      "loss": 1.1199,
+      "step": 14050
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001937763631741953,
+      "loss": 0.9182,
+      "step": 14055
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001937676658083517,
+      "loss": 0.9741,
+      "step": 14060
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019375896256504399,
+      "loss": 0.9724,
+      "step": 14065
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019375025344481772,
+      "loss": 1.0913,
+      "step": 14070
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001937415384482188,
+      "loss": 1.0246,
+      "step": 14075
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019373281757579348,
+      "loss": 1.2072,
+      "step": 14080
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001937240908280884,
+      "loss": 1.1545,
+      "step": 14085
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019371535820565056,
+      "loss": 1.0993,
+      "step": 14090
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019370661970902737,
+      "loss": 1.0449,
+      "step": 14095
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001936978753387665,
+      "loss": 1.0995,
+      "step": 14100
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001936891250954161,
+      "loss": 0.9807,
+      "step": 14105
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019368036897952464,
+      "loss": 1.0008,
+      "step": 14110
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019367160699164097,
+      "loss": 0.99,
+      "step": 14115
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019366283913231427,
+      "loss": 1.0677,
+      "step": 14120
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.00019365406540209414,
+      "loss": 1.0974,
+      "step": 14125
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019364528580153055,
+      "loss": 0.7966,
+      "step": 14130
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019363650033117375,
+      "loss": 0.8748,
+      "step": 14135
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019362770899157452,
+      "loss": 0.9802,
+      "step": 14140
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019361891178328383,
+      "loss": 1.0155,
+      "step": 14145
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019361010870685313,
+      "loss": 1.0516,
+      "step": 14150
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001936012997628342,
+      "loss": 0.9472,
+      "step": 14155
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001935924849517792,
+      "loss": 0.8923,
+      "step": 14160
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019358366427424064,
+      "loss": 0.9306,
+      "step": 14165
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019357483773077143,
+      "loss": 0.9439,
+      "step": 14170
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001935660053219248,
+      "loss": 0.9135,
+      "step": 14175
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019355716704825442,
+      "loss": 0.9764,
+      "step": 14180
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019354832291031425,
+      "loss": 0.8676,
+      "step": 14185
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001935394729086587,
+      "loss": 1.0203,
+      "step": 14190
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019353061704384237,
+      "loss": 0.8991,
+      "step": 14195
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019352175531642049,
+      "loss": 0.933,
+      "step": 14200
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019351288772694847,
+      "loss": 0.9262,
+      "step": 14205
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019350401427598214,
+      "loss": 0.9976,
+      "step": 14210
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019349513496407772,
+      "loss": 0.8633,
+      "step": 14215
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019348624979179173,
+      "loss": 1.0477,
+      "step": 14220
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019347735875968115,
+      "loss": 1.0083,
+      "step": 14225
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019346846186830326,
+      "loss": 0.9553,
+      "step": 14230
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019345955911821572,
+      "loss": 0.9713,
+      "step": 14235
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019345065050997657,
+      "loss": 0.9801,
+      "step": 14240
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001934417360441442,
+      "loss": 1.0721,
+      "step": 14245
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019343281572127742,
+      "loss": 1.081,
+      "step": 14250
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019342388954193536,
+      "loss": 0.8543,
+      "step": 14255
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019341495750667748,
+      "loss": 0.878,
+      "step": 14260
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019340601961606365,
+      "loss": 1.0285,
+      "step": 14265
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019339707587065415,
+      "loss": 0.9378,
+      "step": 14270
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019338812627100958,
+      "loss": 1.1077,
+      "step": 14275
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001933791708176909,
+      "loss": 1.0227,
+      "step": 14280
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019337020951125942,
+      "loss": 1.2616,
+      "step": 14285
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019336124235227686,
+      "loss": 0.9189,
+      "step": 14290
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001933522693413053,
+      "loss": 0.9835,
+      "step": 14295
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001933432904789072,
+      "loss": 1.0314,
+      "step": 14300
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019333430576564534,
+      "loss": 1.0991,
+      "step": 14305
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019332531520208286,
+      "loss": 1.0224,
+      "step": 14310
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019331631878878337,
+      "loss": 0.9544,
+      "step": 14315
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019330731652631073,
+      "loss": 0.9824,
+      "step": 14320
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019329830841522923,
+      "loss": 1.0508,
+      "step": 14325
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019328929445610351,
+      "loss": 0.8179,
+      "step": 14330
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00019328027464949853,
+      "loss": 1.1035,
+      "step": 14335
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019327124899597972,
+      "loss": 0.9759,
+      "step": 14340
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019326221749611282,
+      "loss": 1.1146,
+      "step": 14345
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019325318015046388,
+      "loss": 0.9521,
+      "step": 14350
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001932441369595994,
+      "loss": 0.9631,
+      "step": 14355
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001932350879240862,
+      "loss": 0.879,
+      "step": 14360
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019322603304449155,
+      "loss": 0.8715,
+      "step": 14365
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2421875,
+      "learning_rate": 0.00019321697232138296,
+      "loss": 0.9968,
+      "step": 14370
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019320790575532836,
+      "loss": 0.8983,
+      "step": 14375
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001931988333468961,
+      "loss": 0.9373,
+      "step": 14380
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001931897550966548,
+      "loss": 0.9609,
+      "step": 14385
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001931806710051735,
+      "loss": 0.8751,
+      "step": 14390
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019317158107302166,
+      "loss": 0.9652,
+      "step": 14395
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.000193162485300769,
+      "loss": 0.9237,
+      "step": 14400
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019315338368898564,
+      "loss": 0.932,
+      "step": 14405
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001931442762382421,
+      "loss": 0.9976,
+      "step": 14410
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019313516294910924,
+      "loss": 0.9154,
+      "step": 14415
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001931260438221583,
+      "loss": 1.094,
+      "step": 14420
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00019311691885796087,
+      "loss": 1.0531,
+      "step": 14425
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019310778805708888,
+      "loss": 1.0205,
+      "step": 14430
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019309865142011474,
+      "loss": 1.0082,
+      "step": 14435
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001930895089476111,
+      "loss": 1.0266,
+      "step": 14440
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.000193080360640151,
+      "loss": 0.9064,
+      "step": 14445
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001930712064983079,
+      "loss": 0.8931,
+      "step": 14450
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019306204652265558,
+      "loss": 0.907,
+      "step": 14455
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019305288071376817,
+      "loss": 0.9252,
+      "step": 14460
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.85546875,
+      "learning_rate": 0.0001930437090722202,
+      "loss": 0.946,
+      "step": 14465
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019303453159858665,
+      "loss": 1.1208,
+      "step": 14470
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019302534829344266,
+      "loss": 0.9683,
+      "step": 14475
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001930161591573639,
+      "loss": 0.8724,
+      "step": 14480
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001930069641909263,
+      "loss": 1.0984,
+      "step": 14485
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019299776339470632,
+      "loss": 0.9404,
+      "step": 14490
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001929885567692806,
+      "loss": 0.9694,
+      "step": 14495
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019297934431522623,
+      "loss": 1.0963,
+      "step": 14500
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019297012603312067,
+      "loss": 0.9967,
+      "step": 14505
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001929609019235417,
+      "loss": 0.8824,
+      "step": 14510
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019295167198706757,
+      "loss": 0.9929,
+      "step": 14515
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019294243622427674,
+      "loss": 1.0092,
+      "step": 14520
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019293319463574817,
+      "loss": 1.0737,
+      "step": 14525
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001929239472220611,
+      "loss": 0.932,
+      "step": 14530
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019291469398379524,
+      "loss": 0.906,
+      "step": 14535
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.046875,
+      "learning_rate": 0.00019290543492153045,
+      "loss": 1.0696,
+      "step": 14540
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001928961700358473,
+      "loss": 0.8614,
+      "step": 14545
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019288689932732634,
+      "loss": 1.0545,
+      "step": 14550
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00019287762279654872,
+      "loss": 0.9849,
+      "step": 14555
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019286834044409597,
+      "loss": 0.9447,
+      "step": 14560
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019285905227054985,
+      "loss": 0.9369,
+      "step": 14565
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019284975827649258,
+      "loss": 0.9404,
+      "step": 14570
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019284045846250673,
+      "loss": 0.899,
+      "step": 14575
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001928311528291752,
+      "loss": 1.0397,
+      "step": 14580
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001928218413770813,
+      "loss": 1.0044,
+      "step": 14585
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019281252410680862,
+      "loss": 0.9415,
+      "step": 14590
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019280320101894128,
+      "loss": 0.9335,
+      "step": 14595
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019279387211406358,
+      "loss": 0.8984,
+      "step": 14600
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019278453739276027,
+      "loss": 1.1,
+      "step": 14605
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001927751968556165,
+      "loss": 1.0613,
+      "step": 14610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019276585050321775,
+      "loss": 1.1709,
+      "step": 14615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019275649833614983,
+      "loss": 0.8661,
+      "step": 14620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019274714035499895,
+      "loss": 0.8943,
+      "step": 14625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019273777656035168,
+      "loss": 0.9973,
+      "step": 14630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019272840695279495,
+      "loss": 0.9958,
+      "step": 14635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019271903153291606,
+      "loss": 0.953,
+      "step": 14640
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1484375,
+      "learning_rate": 0.0001927096503013027,
+      "loss": 0.8613,
+      "step": 14645
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019270026325854284,
+      "loss": 0.8887,
+      "step": 14650
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019269087040522495,
+      "loss": 0.9941,
+      "step": 14655
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00019268147174193766,
+      "loss": 0.8934,
+      "step": 14660
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019267206726927025,
+      "loss": 0.959,
+      "step": 14665
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019266265698781205,
+      "loss": 0.9928,
+      "step": 14670
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019265324089815303,
+      "loss": 0.8441,
+      "step": 14675
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019264381900088333,
+      "loss": 0.9877,
+      "step": 14680
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001926343912965935,
+      "loss": 0.8847,
+      "step": 14685
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019262495778587458,
+      "loss": 1.1266,
+      "step": 14690
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019261551846931778,
+      "loss": 1.1062,
+      "step": 14695
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019260607334751483,
+      "loss": 1.1259,
+      "step": 14700
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001925966224210577,
+      "loss": 0.9801,
+      "step": 14705
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019258716569053885,
+      "loss": 0.9701,
+      "step": 14710
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019257770315655098,
+      "loss": 1.0589,
+      "step": 14715
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019256823481968727,
+      "loss": 0.9431,
+      "step": 14720
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019255876068054116,
+      "loss": 0.8557,
+      "step": 14725
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001925492807397065,
+      "loss": 1.1344,
+      "step": 14730
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019253979499777755,
+      "loss": 0.9122,
+      "step": 14735
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019253030345534882,
+      "loss": 1.0495,
+      "step": 14740
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001925208061130153,
+      "loss": 0.97,
+      "step": 14745
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019251130297137227,
+      "loss": 0.9526,
+      "step": 14750
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019250179403101542,
+      "loss": 1.0612,
+      "step": 14755
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019249227929254078,
+      "loss": 0.9857,
+      "step": 14760
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019248275875654474,
+      "loss": 0.9431,
+      "step": 14765
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00019247323242362402,
+      "loss": 0.9341,
+      "step": 14770
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001924637002943758,
+      "loss": 0.9805,
+      "step": 14775
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019245416236939752,
+      "loss": 1.0725,
+      "step": 14780
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019244461864928707,
+      "loss": 1.0164,
+      "step": 14785
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019243506913464261,
+      "loss": 1.2681,
+      "step": 14790
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019242551382606275,
+      "loss": 1.1246,
+      "step": 14795
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019241595272414643,
+      "loss": 1.091,
+      "step": 14800
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019240638582949292,
+      "loss": 0.8939,
+      "step": 14805
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00019239681314270194,
+      "loss": 1.2496,
+      "step": 14810
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019238723466437346,
+      "loss": 0.9049,
+      "step": 14815
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001923776503951079,
+      "loss": 0.8829,
+      "step": 14820
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019236806033550599,
+      "loss": 0.901,
+      "step": 14825
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019235846448616885,
+      "loss": 1.0361,
+      "step": 14830
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019234886284769798,
+      "loss": 0.9696,
+      "step": 14835
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019233925542069523,
+      "loss": 0.9323,
+      "step": 14840
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019232964220576275,
+      "loss": 0.9367,
+      "step": 14845
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019232002320350317,
+      "loss": 1.1198,
+      "step": 14850
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019231039841451937,
+      "loss": 0.9123,
+      "step": 14855
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019230076783941468,
+      "loss": 0.8982,
+      "step": 14860
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00019229113147879273,
+      "loss": 0.9017,
+      "step": 14865
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019228148933325755,
+      "loss": 0.9551,
+      "step": 14870
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001922718414034135,
+      "loss": 0.8955,
+      "step": 14875
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001922621876898654,
+      "loss": 0.9203,
+      "step": 14880
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019225252819321826,
+      "loss": 0.9607,
+      "step": 14885
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001922428629140776,
+      "loss": 0.8713,
+      "step": 14890
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001922331918530492,
+      "loss": 0.9595,
+      "step": 14895
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019222351501073934,
+      "loss": 0.9485,
+      "step": 14900
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019221383238775448,
+      "loss": 0.9874,
+      "step": 14905
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019220414398470162,
+      "loss": 0.9396,
+      "step": 14910
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019219444980218798,
+      "loss": 0.9336,
+      "step": 14915
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00019218474984082122,
+      "loss": 0.9949,
+      "step": 14920
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019217504410120936,
+      "loss": 1.0587,
+      "step": 14925
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019216533258396078,
+      "loss": 1.0473,
+      "step": 14930
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019215561528968414,
+      "loss": 0.9306,
+      "step": 14935
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019214589221898862,
+      "loss": 0.8904,
+      "step": 14940
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019213616337248362,
+      "loss": 0.9354,
+      "step": 14945
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019212642875077895,
+      "loss": 0.9908,
+      "step": 14950
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001921166883544848,
+      "loss": 1.0877,
+      "step": 14955
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019210694218421175,
+      "loss": 1.0058,
+      "step": 14960
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019209719024057063,
+      "loss": 0.8071,
+      "step": 14965
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019208743252417275,
+      "loss": 0.9298,
+      "step": 14970
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001920776690356297,
+      "loss": 0.853,
+      "step": 14975
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001920678997755535,
+      "loss": 1.0874,
+      "step": 14980
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019205812474455648,
+      "loss": 1.062,
+      "step": 14985
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019204834394325135,
+      "loss": 1.0068,
+      "step": 14990
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00019203855737225122,
+      "loss": 0.8639,
+      "step": 14995
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00019202876503216946,
+      "loss": 1.007,
+      "step": 15000
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001920189669236199,
+      "loss": 1.0242,
+      "step": 15005
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019200916304721672,
+      "loss": 1.0116,
+      "step": 15010
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019199935340357438,
+      "loss": 1.0486,
+      "step": 15015
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001919895379933078,
+      "loss": 1.1622,
+      "step": 15020
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001919797168170322,
+      "loss": 1.0576,
+      "step": 15025
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001919698898753632,
+      "loss": 0.9873,
+      "step": 15030
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019196005716891676,
+      "loss": 1.0213,
+      "step": 15035
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019195021869830922,
+      "loss": 1.0241,
+      "step": 15040
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019194037446415723,
+      "loss": 0.8688,
+      "step": 15045
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019193052446707785,
+      "loss": 0.8734,
+      "step": 15050
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019192066870768853,
+      "loss": 0.8426,
+      "step": 15055
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019191080718660697,
+      "loss": 0.949,
+      "step": 15060
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019190093990445134,
+      "loss": 0.9425,
+      "step": 15065
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019189106686184014,
+      "loss": 0.8756,
+      "step": 15070
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019188118805939222,
+      "loss": 0.9335,
+      "step": 15075
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001918713034977268,
+      "loss": 0.9527,
+      "step": 15080
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00019186141317746342,
+      "loss": 0.9751,
+      "step": 15085
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019185151709922205,
+      "loss": 0.9198,
+      "step": 15090
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019184161526362298,
+      "loss": 0.8982,
+      "step": 15095
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00019183170767128686,
+      "loss": 0.9553,
+      "step": 15100
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00019182179432283473,
+      "loss": 0.9782,
+      "step": 15105
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00019181187521888796,
+      "loss": 0.8864,
+      "step": 15110
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00019180195036006825,
+      "loss": 0.9765,
+      "step": 15115
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00019179201974699775,
+      "loss": 0.9345,
+      "step": 15120
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019178208338029894,
+      "loss": 1.0356,
+      "step": 15125
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019177214126059458,
+      "loss": 1.0129,
+      "step": 15130
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001917621933885079,
+      "loss": 0.9623,
+      "step": 15135
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00019175223976466242,
+      "loss": 0.8772,
+      "step": 15140
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019174228038968205,
+      "loss": 1.0285,
+      "step": 15145
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0001917323152641911,
+      "loss": 1.0615,
+      "step": 15150
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019172234438881412,
+      "loss": 0.952,
+      "step": 15155
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00019171236776417612,
+      "loss": 1.0457,
+      "step": 15160
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001917023853909025,
+      "loss": 1.1488,
+      "step": 15165
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019169239726961887,
+      "loss": 0.9128,
+      "step": 15170
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001916824034009514,
+      "loss": 1.0244,
+      "step": 15175
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00019167240378552644,
+      "loss": 0.9686,
+      "step": 15180
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001916623984239708,
+      "loss": 0.9202,
+      "step": 15185
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019165238731691164,
+      "loss": 0.8916,
+      "step": 15190
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019164237046497646,
+      "loss": 0.8949,
+      "step": 15195
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00019163234786879314,
+      "loss": 0.962,
+      "step": 15200
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019162231952898987,
+      "loss": 0.85,
+      "step": 15205
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001916122854461953,
+      "loss": 0.9933,
+      "step": 15210
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019160224562103832,
+      "loss": 0.9851,
+      "step": 15215
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019159220005414825,
+      "loss": 0.9937,
+      "step": 15220
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019158214874615475,
+      "loss": 0.9975,
+      "step": 15225
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001915720916976879,
+      "loss": 1.0316,
+      "step": 15230
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9140625,
+      "learning_rate": 0.00019156202890937803,
+      "loss": 1.1671,
+      "step": 15235
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00019155196038185592,
+      "loss": 1.0737,
+      "step": 15240
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019154188611575265,
+      "loss": 1.0276,
+      "step": 15245
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001915318061116997,
+      "loss": 0.9565,
+      "step": 15250
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019152172037032892,
+      "loss": 1.0703,
+      "step": 15255
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019151162889227246,
+      "loss": 0.8682,
+      "step": 15260
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019150153167816288,
+      "loss": 1.1064,
+      "step": 15265
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001914914287286331,
+      "loss": 0.9682,
+      "step": 15270
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00019148132004431632,
+      "loss": 0.9411,
+      "step": 15275
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019147120562584624,
+      "loss": 0.8581,
+      "step": 15280
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001914610854738568,
+      "loss": 0.9584,
+      "step": 15285
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019145095958898235,
+      "loss": 0.8806,
+      "step": 15290
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019144082797185763,
+      "loss": 1.026,
+      "step": 15295
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019143069062311767,
+      "loss": 0.9719,
+      "step": 15300
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001914205475433979,
+      "loss": 0.891,
+      "step": 15305
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00019141039873333404,
+      "loss": 0.9281,
+      "step": 15310
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019140024419356233,
+      "loss": 1.0256,
+      "step": 15315
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019139008392471917,
+      "loss": 1.1613,
+      "step": 15320
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001913799179274415,
+      "loss": 0.969,
+      "step": 15325
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019136974620236652,
+      "loss": 0.8954,
+      "step": 15330
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00019135956875013176,
+      "loss": 1.04,
+      "step": 15335
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019134938557137515,
+      "loss": 1.0477,
+      "step": 15340
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019133919666673507,
+      "loss": 0.9627,
+      "step": 15345
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019132900203685006,
+      "loss": 0.8964,
+      "step": 15350
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019131880168235922,
+      "loss": 0.9475,
+      "step": 15355
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001913085956039019,
+      "loss": 1.0046,
+      "step": 15360
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001912983838021178,
+      "loss": 0.8238,
+      "step": 15365
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019128816627764702,
+      "loss": 0.8787,
+      "step": 15370
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019127794303113002,
+      "loss": 0.9763,
+      "step": 15375
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019126771406320758,
+      "loss": 0.8963,
+      "step": 15380
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001912574793745209,
+      "loss": 0.9862,
+      "step": 15385
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019124723896571147,
+      "loss": 1.0727,
+      "step": 15390
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019123699283742117,
+      "loss": 0.9808,
+      "step": 15395
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00019122674099029225,
+      "loss": 1.028,
+      "step": 15400
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019121648342496731,
+      "loss": 0.9697,
+      "step": 15405
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019120622014208932,
+      "loss": 0.8534,
+      "step": 15410
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00019119595114230154,
+      "loss": 1.0241,
+      "step": 15415
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001911856764262477,
+      "loss": 1.088,
+      "step": 15420
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019117539599457182,
+      "loss": 1.0458,
+      "step": 15425
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019116510984791825,
+      "loss": 0.9081,
+      "step": 15430
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001911548179869318,
+      "loss": 0.9618,
+      "step": 15435
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001911445204122575,
+      "loss": 0.9968,
+      "step": 15440
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001911342171245409,
+      "loss": 0.9958,
+      "step": 15445
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019112390812442773,
+      "loss": 0.984,
+      "step": 15450
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019111359341256426,
+      "loss": 1.0099,
+      "step": 15455
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019110327298959697,
+      "loss": 0.879,
+      "step": 15460
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019109294685617275,
+      "loss": 1.2221,
+      "step": 15465
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001910826150129389,
+      "loss": 0.8186,
+      "step": 15470
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019107227746054302,
+      "loss": 0.8354,
+      "step": 15475
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00019106193419963304,
+      "loss": 1.0459,
+      "step": 15480
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019105158523085734,
+      "loss": 1.156,
+      "step": 15485
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019104123055486454,
+      "loss": 0.9457,
+      "step": 15490
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019103087017230375,
+      "loss": 0.8631,
+      "step": 15495
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00019102050408382432,
+      "loss": 1.0076,
+      "step": 15500
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019101013229007606,
+      "loss": 0.8997,
+      "step": 15505
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019099975479170903,
+      "loss": 1.0633,
+      "step": 15510
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019098937158937374,
+      "loss": 0.8956,
+      "step": 15515
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.000190978982683721,
+      "loss": 1.0573,
+      "step": 15520
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.000190968588075402,
+      "loss": 1.0133,
+      "step": 15525
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019095818776506836,
+      "loss": 0.9309,
+      "step": 15530
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00019094778175337185,
+      "loss": 1.1443,
+      "step": 15535
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00019093737004096485,
+      "loss": 1.0728,
+      "step": 15540
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019092695262849988,
+      "loss": 0.9081,
+      "step": 15545
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00019091652951662997,
+      "loss": 0.9224,
+      "step": 15550
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00019090610070600844,
+      "loss": 0.9895,
+      "step": 15555
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019089566619728902,
+      "loss": 0.8279,
+      "step": 15560
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001908852259911257,
+      "loss": 0.8762,
+      "step": 15565
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001908747800881729,
+      "loss": 0.9472,
+      "step": 15570
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001908643284890854,
+      "loss": 0.9305,
+      "step": 15575
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001908538711945183,
+      "loss": 0.9279,
+      "step": 15580
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00019084340820512706,
+      "loss": 1.154,
+      "step": 15585
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00019083293952156755,
+      "loss": 1.019,
+      "step": 15590
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019082246514449594,
+      "loss": 1.0037,
+      "step": 15595
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00019081198507456878,
+      "loss": 0.8323,
+      "step": 15600
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019080149931244297,
+      "loss": 0.9139,
+      "step": 15605
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019079100785877577,
+      "loss": 0.7942,
+      "step": 15610
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00019078051071422478,
+      "loss": 0.9099,
+      "step": 15615
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.000190770007879448,
+      "loss": 1.0347,
+      "step": 15620
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00019075949935510374,
+      "loss": 0.9075,
+      "step": 15625
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00019074898514185072,
+      "loss": 0.8973,
+      "step": 15630
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00019073846524034793,
+      "loss": 1.0093,
+      "step": 15635
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001907279396512548,
+      "loss": 1.025,
+      "step": 15640
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001907174083752311,
+      "loss": 1.1046,
+      "step": 15645
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00019070687141293689,
+      "loss": 0.807,
+      "step": 15650
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00019069632876503269,
+      "loss": 0.9024,
+      "step": 15655
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019068578043217934,
+      "loss": 0.8765,
+      "step": 15660
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00019067522641503794,
+      "loss": 0.976,
+      "step": 15665
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001906646667142701,
+      "loss": 1.0139,
+      "step": 15670
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00019065410133053766,
+      "loss": 0.979,
+      "step": 15675
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00019064353026450296,
+      "loss": 1.1048,
+      "step": 15680
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00019063295351682852,
+      "loss": 0.975,
+      "step": 15685
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00019062237108817732,
+      "loss": 0.9718,
+      "step": 15690
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00019061178297921272,
+      "loss": 0.962,
+      "step": 15695
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00019060118919059834,
+      "loss": 1.074,
+      "step": 15700
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019059058972299825,
+      "loss": 1.0041,
+      "step": 15705
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019057998457707682,
+      "loss": 0.9185,
+      "step": 15710
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019056937375349877,
+      "loss": 0.9434,
+      "step": 15715
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00019055875725292927,
+      "loss": 0.7916,
+      "step": 15720
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00019054813507603368,
+      "loss": 0.9599,
+      "step": 15725
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001905375072234779,
+      "loss": 0.8582,
+      "step": 15730
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00019052687369592802,
+      "loss": 0.8423,
+      "step": 15735
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00019051623449405062,
+      "loss": 0.9381,
+      "step": 15740
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00019050558961851254,
+      "loss": 0.8004,
+      "step": 15745
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019049493906998102,
+      "loss": 0.9623,
+      "step": 15750
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019048428284912364,
+      "loss": 1.1378,
+      "step": 15755
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001904736209566084,
+      "loss": 1.0624,
+      "step": 15760
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00019046295339310353,
+      "loss": 0.8383,
+      "step": 15765
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019045228015927772,
+      "loss": 0.9157,
+      "step": 15770
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019044160125579994,
+      "loss": 0.9132,
+      "step": 15775
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00019043091668333965,
+      "loss": 1.1574,
+      "step": 15780
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019042022644256648,
+      "loss": 1.047,
+      "step": 15785
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00019040953053415055,
+      "loss": 0.919,
+      "step": 15790
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00019039882895876224,
+      "loss": 0.9386,
+      "step": 15795
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019038812171707242,
+      "loss": 0.9384,
+      "step": 15800
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00019037740880975217,
+      "loss": 1.1885,
+      "step": 15805
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.000190366690237473,
+      "loss": 0.9901,
+      "step": 15810
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019035596600090675,
+      "loss": 0.9708,
+      "step": 15815
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001903452361007257,
+      "loss": 1.0002,
+      "step": 15820
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001903345005376023,
+      "loss": 1.0308,
+      "step": 15825
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019032375931220954,
+      "loss": 0.9574,
+      "step": 15830
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001903130124252207,
+      "loss": 0.972,
+      "step": 15835
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00019030225987730934,
+      "loss": 1.0213,
+      "step": 15840
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001902915016691495,
+      "loss": 1.0051,
+      "step": 15845
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019028073780141553,
+      "loss": 1.1086,
+      "step": 15850
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019026996827478207,
+      "loss": 0.8926,
+      "step": 15855
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001902591930899242,
+      "loss": 1.1078,
+      "step": 15860
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00019024841224751728,
+      "loss": 0.9868,
+      "step": 15865
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00019023762574823714,
+      "loss": 1.1652,
+      "step": 15870
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001902268335927598,
+      "loss": 0.849,
+      "step": 15875
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019021603578176183,
+      "loss": 1.0021,
+      "step": 15880
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00019020523231591993,
+      "loss": 0.8994,
+      "step": 15885
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00019019442319591137,
+      "loss": 1.0758,
+      "step": 15890
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00019018360842241362,
+      "loss": 0.9965,
+      "step": 15895
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001901727879961046,
+      "loss": 0.8621,
+      "step": 15900
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00019016196191766255,
+      "loss": 1.0158,
+      "step": 15905
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.000190151130187766,
+      "loss": 1.0215,
+      "step": 15910
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00019014029280709397,
+      "loss": 0.9089,
+      "step": 15915
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001901294497763257,
+      "loss": 0.8874,
+      "step": 15920
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00019011860109614088,
+      "loss": 0.908,
+      "step": 15925
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00019010774676721947,
+      "loss": 0.9505,
+      "step": 15930
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001900968867902419,
+      "loss": 0.8472,
+      "step": 15935
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001900860211658889,
+      "loss": 0.9841,
+      "step": 15940
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00019007514989484144,
+      "loss": 1.1091,
+      "step": 15945
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.000190064272977781,
+      "loss": 0.9715,
+      "step": 15950
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00019005339041538937,
+      "loss": 1.041,
+      "step": 15955
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00019004250220834866,
+      "loss": 0.941,
+      "step": 15960
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00019003160835734135,
+      "loss": 1.0787,
+      "step": 15965
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001900207088630503,
+      "loss": 0.8941,
+      "step": 15970
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001900098037261587,
+      "loss": 1.229,
+      "step": 15975
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001899988929473501,
+      "loss": 1.0056,
+      "step": 15980
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018998797652730837,
+      "loss": 0.9708,
+      "step": 15985
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018997705446671778,
+      "loss": 0.8722,
+      "step": 15990
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018996612676626295,
+      "loss": 1.0668,
+      "step": 15995
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00018995519342662883,
+      "loss": 1.0248,
+      "step": 16000
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018994425444850076,
+      "loss": 0.9662,
+      "step": 16005
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018993330983256435,
+      "loss": 0.9792,
+      "step": 16010
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018992235957950566,
+      "loss": 1.0946,
+      "step": 16015
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018991140369001107,
+      "loss": 0.8866,
+      "step": 16020
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001899004421647673,
+      "loss": 0.9569,
+      "step": 16025
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018988947500446138,
+      "loss": 1.0165,
+      "step": 16030
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018987850220978086,
+      "loss": 0.9958,
+      "step": 16035
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018986752378141343,
+      "loss": 1.1076,
+      "step": 16040
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001898565397200472,
+      "loss": 0.9612,
+      "step": 16045
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001898455500263708,
+      "loss": 0.9617,
+      "step": 16050
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.000189834554701073,
+      "loss": 1.0497,
+      "step": 16055
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018982355374484296,
+      "loss": 0.9259,
+      "step": 16060
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001898125471583703,
+      "loss": 0.955,
+      "step": 16065
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018980153494234486,
+      "loss": 0.8977,
+      "step": 16070
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.000189790517097457,
+      "loss": 0.8006,
+      "step": 16075
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001897794936243972,
+      "loss": 1.0832,
+      "step": 16080
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00018976846452385652,
+      "loss": 1.0089,
+      "step": 16085
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018975742979652623,
+      "loss": 0.9711,
+      "step": 16090
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018974638944309802,
+      "loss": 0.8898,
+      "step": 16095
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001897353434642639,
+      "loss": 0.8701,
+      "step": 16100
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001897242918607163,
+      "loss": 0.9969,
+      "step": 16105
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018971323463314784,
+      "loss": 0.9325,
+      "step": 16110
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018970217178225168,
+      "loss": 0.9804,
+      "step": 16115
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001896911033087212,
+      "loss": 0.8709,
+      "step": 16120
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018968002921325027,
+      "loss": 1.0715,
+      "step": 16125
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018966894949653293,
+      "loss": 1.1704,
+      "step": 16130
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001896578641592637,
+      "loss": 0.8445,
+      "step": 16135
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018964677320213748,
+      "loss": 0.8603,
+      "step": 16140
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001896356766258494,
+      "loss": 0.9586,
+      "step": 16145
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018962457443109503,
+      "loss": 0.9594,
+      "step": 16150
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018961346661857021,
+      "loss": 0.9027,
+      "step": 16155
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00018960235318897132,
+      "loss": 0.9021,
+      "step": 16160
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00018959123414299484,
+      "loss": 1.0627,
+      "step": 16165
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018958010948133777,
+      "loss": 1.0311,
+      "step": 16170
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018956897920469743,
+      "loss": 0.9209,
+      "step": 16175
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001895578433137714,
+      "loss": 0.9624,
+      "step": 16180
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018954670180925783,
+      "loss": 1.0426,
+      "step": 16185
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.000189535554691855,
+      "loss": 0.9737,
+      "step": 16190
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001895244019622616,
+      "loss": 1.0148,
+      "step": 16195
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018951324362117674,
+      "loss": 0.9849,
+      "step": 16200
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001895020796692998,
+      "loss": 0.985,
+      "step": 16205
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001894909101073306,
+      "loss": 0.7954,
+      "step": 16210
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001894797349359692,
+      "loss": 0.9957,
+      "step": 16215
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018946855415591615,
+      "loss": 1.074,
+      "step": 16220
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001894573677678722,
+      "loss": 0.9898,
+      "step": 16225
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018944617577253855,
+      "loss": 0.8775,
+      "step": 16230
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018943497817061676,
+      "loss": 1.0057,
+      "step": 16235
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018942377496280867,
+      "loss": 0.8942,
+      "step": 16240
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001894125661498165,
+      "loss": 1.0112,
+      "step": 16245
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00018940135173234286,
+      "loss": 0.7383,
+      "step": 16250
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018939013171109068,
+      "loss": 0.965,
+      "step": 16255
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018937890608676324,
+      "loss": 0.8931,
+      "step": 16260
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00018936767486006417,
+      "loss": 0.9322,
+      "step": 16265
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018935643803169746,
+      "loss": 1.1366,
+      "step": 16270
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018934519560236744,
+      "loss": 0.8809,
+      "step": 16275
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001893339475727788,
+      "loss": 0.9132,
+      "step": 16280
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001893226939436366,
+      "loss": 0.8929,
+      "step": 16285
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018931143471564622,
+      "loss": 0.919,
+      "step": 16290
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018930016988951334,
+      "loss": 0.8347,
+      "step": 16295
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018928889946594416,
+      "loss": 0.8911,
+      "step": 16300
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018927762344564503,
+      "loss": 1.0486,
+      "step": 16305
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018926634182932282,
+      "loss": 0.9989,
+      "step": 16310
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018925505461768466,
+      "loss": 0.949,
+      "step": 16315
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018924376181143798,
+      "loss": 1.0124,
+      "step": 16320
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018923246341129066,
+      "loss": 0.9844,
+      "step": 16325
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018922115941795092,
+      "loss": 0.9543,
+      "step": 16330
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001892098498321273,
+      "loss": 1.0143,
+      "step": 16335
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001891985346545287,
+      "loss": 0.997,
+      "step": 16340
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018918721388586435,
+      "loss": 1.0011,
+      "step": 16345
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018917588752684385,
+      "loss": 0.9518,
+      "step": 16350
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018916455557817718,
+      "loss": 0.9175,
+      "step": 16355
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001891532180405746,
+      "loss": 0.9749,
+      "step": 16360
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00018914187491474678,
+      "loss": 0.9813,
+      "step": 16365
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001891305262014047,
+      "loss": 0.9044,
+      "step": 16370
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018911917190125976,
+      "loss": 1.0183,
+      "step": 16375
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001891078120150236,
+      "loss": 1.0059,
+      "step": 16380
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001890964465434083,
+      "loss": 0.9395,
+      "step": 16385
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00018908507548712626,
+      "loss": 0.9566,
+      "step": 16390
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018907369884689024,
+      "loss": 0.9328,
+      "step": 16395
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00018906231662341332,
+      "loss": 1.021,
+      "step": 16400
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018905092881740899,
+      "loss": 0.9493,
+      "step": 16405
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018903953542959097,
+      "loss": 0.9246,
+      "step": 16410
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001890281364606735,
+      "loss": 1.0011,
+      "step": 16415
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018901673191137102,
+      "loss": 0.9221,
+      "step": 16420
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001890053217823984,
+      "loss": 0.9183,
+      "step": 16425
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018899390607447086,
+      "loss": 1.0647,
+      "step": 16430
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001889824847883039,
+      "loss": 1.0372,
+      "step": 16435
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001889710579246135,
+      "loss": 0.9747,
+      "step": 16440
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018895962548411583,
+      "loss": 0.9561,
+      "step": 16445
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001889481874675275,
+      "loss": 0.9192,
+      "step": 16450
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001889367438755655,
+      "loss": 0.9859,
+      "step": 16455
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018892529470894713,
+      "loss": 1.0249,
+      "step": 16460
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018891383996838998,
+      "loss": 0.9751,
+      "step": 16465
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018890237965461207,
+      "loss": 0.8279,
+      "step": 16470
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018889091376833177,
+      "loss": 1.1476,
+      "step": 16475
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018887944231026774,
+      "loss": 0.9666,
+      "step": 16480
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018886796528113907,
+      "loss": 0.9175,
+      "step": 16485
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018885648268166509,
+      "loss": 0.8682,
+      "step": 16490
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018884499451256556,
+      "loss": 0.8868,
+      "step": 16495
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00018883350077456064,
+      "loss": 0.9363,
+      "step": 16500
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018882200146837067,
+      "loss": 0.9751,
+      "step": 16505
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018881049659471652,
+      "loss": 1.0708,
+      "step": 16510
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018879898615431931,
+      "loss": 1.0396,
+      "step": 16515
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018878747014790048,
+      "loss": 0.7451,
+      "step": 16520
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018877594857618193,
+      "loss": 0.9338,
+      "step": 16525
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001887644214398858,
+      "loss": 1.0115,
+      "step": 16530
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00018875288873973465,
+      "loss": 0.8998,
+      "step": 16535
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018874135047645133,
+      "loss": 1.0624,
+      "step": 16540
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001887298066507591,
+      "loss": 0.9733,
+      "step": 16545
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018871825726338157,
+      "loss": 0.9463,
+      "step": 16550
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018870670231504257,
+      "loss": 1.0434,
+      "step": 16555
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001886951418064665,
+      "loss": 0.9086,
+      "step": 16560
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00018868357573837788,
+      "loss": 1.0508,
+      "step": 16565
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018867200411150177,
+      "loss": 1.0226,
+      "step": 16570
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018866042692656344,
+      "loss": 0.8664,
+      "step": 16575
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018864884418428855,
+      "loss": 0.908,
+      "step": 16580
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018863725588540316,
+      "loss": 0.9478,
+      "step": 16585
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018862566203063365,
+      "loss": 1.0568,
+      "step": 16590
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001886140626207067,
+      "loss": 1.0956,
+      "step": 16595
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018860245765634935,
+      "loss": 0.9001,
+      "step": 16600
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018859084713828908,
+      "loss": 1.122,
+      "step": 16605
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018857923106725357,
+      "loss": 0.9613,
+      "step": 16610
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.000188567609443971,
+      "loss": 1.0178,
+      "step": 16615
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001885559822691698,
+      "loss": 1.1274,
+      "step": 16620
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018854434954357875,
+      "loss": 0.9865,
+      "step": 16625
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018853271126792706,
+      "loss": 0.9709,
+      "step": 16630
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001885210674429442,
+      "loss": 0.9059,
+      "step": 16635
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018850941806936004,
+      "loss": 0.7779,
+      "step": 16640
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001884977631479047,
+      "loss": 0.9693,
+      "step": 16645
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018848610267930877,
+      "loss": 0.9123,
+      "step": 16650
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018847443666430316,
+      "loss": 1.0178,
+      "step": 16655
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001884627651036191,
+      "loss": 0.8877,
+      "step": 16660
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001884510879979882,
+      "loss": 1.0474,
+      "step": 16665
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018843940534814233,
+      "loss": 0.9245,
+      "step": 16670
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018842771715481382,
+      "loss": 0.9171,
+      "step": 16675
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00018841602341873527,
+      "loss": 0.8864,
+      "step": 16680
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018840432414063974,
+      "loss": 0.922,
+      "step": 16685
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018839261932126044,
+      "loss": 0.8619,
+      "step": 16690
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018838090896133107,
+      "loss": 0.9576,
+      "step": 16695
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018836919306158568,
+      "loss": 1.0071,
+      "step": 16700
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018835747162275864,
+      "loss": 1.0803,
+      "step": 16705
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018834574464558464,
+      "loss": 0.8814,
+      "step": 16710
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018833401213079877,
+      "loss": 0.938,
+      "step": 16715
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018832227407913638,
+      "loss": 1.0113,
+      "step": 16720
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018831053049133328,
+      "loss": 0.9646,
+      "step": 16725
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018829878136812558,
+      "loss": 1.1012,
+      "step": 16730
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018828702671024966,
+      "loss": 1.0641,
+      "step": 16735
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018827526651844236,
+      "loss": 0.8878,
+      "step": 16740
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018826350079344085,
+      "loss": 0.9392,
+      "step": 16745
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018825172953598252,
+      "loss": 0.9466,
+      "step": 16750
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018823995274680532,
+      "loss": 0.9176,
+      "step": 16755
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001882281704266474,
+      "loss": 0.8922,
+      "step": 16760
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018821638257624724,
+      "loss": 0.95,
+      "step": 16765
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018820458919634376,
+      "loss": 0.9833,
+      "step": 16770
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018819279028767618,
+      "loss": 0.9822,
+      "step": 16775
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018818098585098405,
+      "loss": 1.0689,
+      "step": 16780
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018816917588700732,
+      "loss": 0.8965,
+      "step": 16785
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001881573603964862,
+      "loss": 0.9886,
+      "step": 16790
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018814553938016135,
+      "loss": 1.0189,
+      "step": 16795
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018813371283877368,
+      "loss": 1.0355,
+      "step": 16800
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001881218807730645,
+      "loss": 0.9595,
+      "step": 16805
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018811004318377552,
+      "loss": 0.9699,
+      "step": 16810
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018809820007164863,
+      "loss": 0.9936,
+      "step": 16815
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018808635143742627,
+      "loss": 0.9886,
+      "step": 16820
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018807449728185108,
+      "loss": 0.9061,
+      "step": 16825
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018806263760566606,
+      "loss": 0.9264,
+      "step": 16830
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018805077240961464,
+      "loss": 1.0603,
+      "step": 16835
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001880389016944405,
+      "loss": 0.8612,
+      "step": 16840
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001880270254608878,
+      "loss": 0.983,
+      "step": 16845
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001880151437097008,
+      "loss": 0.9251,
+      "step": 16850
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018800325644162443,
+      "loss": 1.1613,
+      "step": 16855
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018799136365740368,
+      "loss": 0.9665,
+      "step": 16860
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018797946535778403,
+      "loss": 1.066,
+      "step": 16865
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018796756154351133,
+      "loss": 0.9028,
+      "step": 16870
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018795565221533167,
+      "loss": 0.9527,
+      "step": 16875
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018794373737399152,
+      "loss": 1.0282,
+      "step": 16880
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001879318170202378,
+      "loss": 0.8708,
+      "step": 16885
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018791989115481762,
+      "loss": 0.9773,
+      "step": 16890
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018790795977847852,
+      "loss": 0.9057,
+      "step": 16895
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018789602289196838,
+      "loss": 1.0618,
+      "step": 16900
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001878840804960355,
+      "loss": 0.8619,
+      "step": 16905
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018787213259142827,
+      "loss": 0.9153,
+      "step": 16910
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001878601791788957,
+      "loss": 1.0297,
+      "step": 16915
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018784822025918706,
+      "loss": 0.9308,
+      "step": 16920
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018783625583305194,
+      "loss": 1.0167,
+      "step": 16925
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00018782428590124025,
+      "loss": 0.984,
+      "step": 16930
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001878123104645023,
+      "loss": 0.9254,
+      "step": 16935
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001878003295235887,
+      "loss": 1.0324,
+      "step": 16940
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018778834307925046,
+      "loss": 1.0301,
+      "step": 16945
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00018777635113223888,
+      "loss": 0.8193,
+      "step": 16950
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018776435368330567,
+      "loss": 1.0801,
+      "step": 16955
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001877523507332028,
+      "loss": 0.9553,
+      "step": 16960
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018774034228268265,
+      "loss": 0.8491,
+      "step": 16965
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018772832833249792,
+      "loss": 0.9802,
+      "step": 16970
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018771630888340165,
+      "loss": 0.9014,
+      "step": 16975
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018770428393614724,
+      "loss": 0.9266,
+      "step": 16980
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018769225349148842,
+      "loss": 1.0798,
+      "step": 16985
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018768021755017928,
+      "loss": 0.9488,
+      "step": 16990
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018766817611297424,
+      "loss": 0.9043,
+      "step": 16995
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001876561291806281,
+      "loss": 1.104,
+      "step": 17000
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018764407675389593,
+      "loss": 1.1461,
+      "step": 17005
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018763201883353323,
+      "loss": 1.035,
+      "step": 17010
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018761995542029576,
+      "loss": 0.8472,
+      "step": 17015
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018760788651493973,
+      "loss": 0.9361,
+      "step": 17020
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018759581211822156,
+      "loss": 1.0741,
+      "step": 17025
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001875837322308981,
+      "loss": 1.028,
+      "step": 17030
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018757164685372663,
+      "loss": 1.028,
+      "step": 17035
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018755955598746455,
+      "loss": 1.055,
+      "step": 17040
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018754745963286978,
+      "loss": 0.8644,
+      "step": 17045
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018753535779070056,
+      "loss": 0.9841,
+      "step": 17050
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018752325046171537,
+      "loss": 0.908,
+      "step": 17055
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001875111376466732,
+      "loss": 1.0285,
+      "step": 17060
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018749901934633325,
+      "loss": 0.9899,
+      "step": 17065
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001874868955614551,
+      "loss": 1.1264,
+      "step": 17070
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001874747662927987,
+      "loss": 1.0565,
+      "step": 17075
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018746263154112433,
+      "loss": 0.9554,
+      "step": 17080
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001874504913071926,
+      "loss": 1.0431,
+      "step": 17085
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001874383455917645,
+      "loss": 0.9707,
+      "step": 17090
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018742619439560126,
+      "loss": 1.0394,
+      "step": 17095
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018741403771946463,
+      "loss": 0.9724,
+      "step": 17100
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018740187556411653,
+      "loss": 1.0729,
+      "step": 17105
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001873897079303194,
+      "loss": 0.9667,
+      "step": 17110
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018737753481883575,
+      "loss": 1.1016,
+      "step": 17115
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018736535623042877,
+      "loss": 0.9476,
+      "step": 17120
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018735317216586172,
+      "loss": 1.0506,
+      "step": 17125
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018734098262589835,
+      "loss": 0.9753,
+      "step": 17130
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018732878761130274,
+      "loss": 0.9667,
+      "step": 17135
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00018731658712283929,
+      "loss": 0.9733,
+      "step": 17140
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018730438116127266,
+      "loss": 0.9274,
+      "step": 17145
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.000187292169727368,
+      "loss": 1.0296,
+      "step": 17150
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018727995282189074,
+      "loss": 0.7776,
+      "step": 17155
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00018726773044560664,
+      "loss": 0.8341,
+      "step": 17160
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001872555025992818,
+      "loss": 0.7679,
+      "step": 17165
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018724326928368267,
+      "loss": 0.9864,
+      "step": 17170
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018723103049957606,
+      "loss": 1.0016,
+      "step": 17175
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018721878624772912,
+      "loss": 0.958,
+      "step": 17180
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018720653652890934,
+      "loss": 0.8523,
+      "step": 17185
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001871942813438845,
+      "loss": 0.8542,
+      "step": 17190
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018718202069342282,
+      "loss": 0.8759,
+      "step": 17195
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018716975457829275,
+      "loss": 0.9566,
+      "step": 17200
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018715748299926322,
+      "loss": 1.0393,
+      "step": 17205
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001871452059571034,
+      "loss": 1.0312,
+      "step": 17210
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0001871329234525828,
+      "loss": 1.0642,
+      "step": 17215
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018712063548647132,
+      "loss": 1.0684,
+      "step": 17220
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001871083420595392,
+      "loss": 0.9215,
+      "step": 17225
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018709604317255699,
+      "loss": 0.9932,
+      "step": 17230
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001870837388262956,
+      "loss": 0.8693,
+      "step": 17235
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018707142902152626,
+      "loss": 0.9005,
+      "step": 17240
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018705911375902062,
+      "loss": 1.029,
+      "step": 17245
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018704679303955054,
+      "loss": 1.0987,
+      "step": 17250
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018703446686388838,
+      "loss": 0.9246,
+      "step": 17255
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00018702213523280668,
+      "loss": 0.9038,
+      "step": 17260
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018700979814707843,
+      "loss": 1.0012,
+      "step": 17265
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00018699745560747696,
+      "loss": 0.8978,
+      "step": 17270
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018698510761477587,
+      "loss": 0.9811,
+      "step": 17275
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018697275416974915,
+      "loss": 0.9977,
+      "step": 17280
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018696039527317117,
+      "loss": 0.9776,
+      "step": 17285
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018694803092581655,
+      "loss": 1.0117,
+      "step": 17290
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018693566112846038,
+      "loss": 0.9689,
+      "step": 17295
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001869232858818779,
+      "loss": 0.8599,
+      "step": 17300
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001869109051868449,
+      "loss": 0.9914,
+      "step": 17305
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018689851904413738,
+      "loss": 1.0214,
+      "step": 17310
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018688612745453172,
+      "loss": 0.9126,
+      "step": 17315
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001868737304188046,
+      "loss": 0.972,
+      "step": 17320
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018686132793773315,
+      "loss": 0.9786,
+      "step": 17325
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018684892001209473,
+      "loss": 0.8262,
+      "step": 17330
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018683650664266707,
+      "loss": 0.9764,
+      "step": 17335
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018682408783022832,
+      "loss": 1.0312,
+      "step": 17340
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001868116635755568,
+      "loss": 0.8887,
+      "step": 17345
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001867992338794314,
+      "loss": 0.9327,
+      "step": 17350
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018678679874263113,
+      "loss": 1.0066,
+      "step": 17355
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018677435816593548,
+      "loss": 1.0451,
+      "step": 17360
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00018676191215012423,
+      "loss": 0.8877,
+      "step": 17365
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001867494606959775,
+      "loss": 0.951,
+      "step": 17370
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018673700380427582,
+      "loss": 1.0618,
+      "step": 17375
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018672454147579992,
+      "loss": 1.0022,
+      "step": 17380
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018671207371133097,
+      "loss": 0.85,
+      "step": 17385
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018669960051165052,
+      "loss": 0.8242,
+      "step": 17390
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018668712187754034,
+      "loss": 0.9412,
+      "step": 17395
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018667463780978265,
+      "loss": 0.9668,
+      "step": 17400
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00018666214830915997,
+      "loss": 0.9269,
+      "step": 17405
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001866496533764551,
+      "loss": 1.0226,
+      "step": 17410
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018663715301245128,
+      "loss": 0.9235,
+      "step": 17415
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018662464721793203,
+      "loss": 1.1844,
+      "step": 17420
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018661213599368125,
+      "loss": 1.0601,
+      "step": 17425
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018659961934048313,
+      "loss": 1.0451,
+      "step": 17430
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018658709725912225,
+      "loss": 0.9832,
+      "step": 17435
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001865745697503835,
+      "loss": 1.0443,
+      "step": 17440
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001865620368150521,
+      "loss": 0.8875,
+      "step": 17445
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018654949845391366,
+      "loss": 0.8692,
+      "step": 17450
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001865369546677541,
+      "loss": 1.0209,
+      "step": 17455
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00018652440545735964,
+      "loss": 0.7376,
+      "step": 17460
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001865118508235169,
+      "loss": 0.8804,
+      "step": 17465
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018649929076701283,
+      "loss": 1.0329,
+      "step": 17470
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001864867252886347,
+      "loss": 0.9949,
+      "step": 17475
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018647415438917015,
+      "loss": 0.9137,
+      "step": 17480
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018646157806940708,
+      "loss": 0.9676,
+      "step": 17485
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018644899633013388,
+      "loss": 1.0912,
+      "step": 17490
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018643640917213907,
+      "loss": 0.9975,
+      "step": 17495
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018642381659621173,
+      "loss": 0.932,
+      "step": 17500
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018641121860314114,
+      "loss": 1.0452,
+      "step": 17505
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00018639861519371693,
+      "loss": 0.9322,
+      "step": 17510
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018638600636872914,
+      "loss": 0.8564,
+      "step": 17515
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018637339212896806,
+      "loss": 0.9579,
+      "step": 17520
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001863607724752244,
+      "loss": 0.9324,
+      "step": 17525
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001863481474082892,
+      "loss": 0.9017,
+      "step": 17530
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018633551692895373,
+      "loss": 0.9417,
+      "step": 17535
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8203125,
+      "learning_rate": 0.00018632288103800974,
+      "loss": 0.8638,
+      "step": 17540
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00018631023973624927,
+      "loss": 0.9747,
+      "step": 17545
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018629759302446467,
+      "loss": 0.8661,
+      "step": 17550
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018628494090344863,
+      "loss": 1.106,
+      "step": 17555
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018627228337399427,
+      "loss": 1.0067,
+      "step": 17560
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001862596204368949,
+      "loss": 1.1062,
+      "step": 17565
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018624695209294427,
+      "loss": 0.8607,
+      "step": 17570
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001862342783429365,
+      "loss": 1.0601,
+      "step": 17575
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001862215991876659,
+      "loss": 0.8794,
+      "step": 17580
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018620891462792729,
+      "loss": 0.9462,
+      "step": 17585
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018619622466451574,
+      "loss": 1.0127,
+      "step": 17590
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00018618352929822662,
+      "loss": 1.0018,
+      "step": 17595
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018617082852985577,
+      "loss": 1.0369,
+      "step": 17600
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018615812236019922,
+      "loss": 0.9432,
+      "step": 17605
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018614541079005345,
+      "loss": 0.8503,
+      "step": 17610
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018613269382021522,
+      "loss": 0.9858,
+      "step": 17615
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018611997145148165,
+      "loss": 0.9059,
+      "step": 17620
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018610724368465015,
+      "loss": 1.0486,
+      "step": 17625
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018609451052051858,
+      "loss": 0.9702,
+      "step": 17630
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018608177195988504,
+      "loss": 1.2176,
+      "step": 17635
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018606902800354796,
+      "loss": 1.0093,
+      "step": 17640
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001860562786523062,
+      "loss": 0.9306,
+      "step": 17645
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001860435239069589,
+      "loss": 1.1033,
+      "step": 17650
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018603076376830555,
+      "loss": 0.9792,
+      "step": 17655
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001860179982371459,
+      "loss": 0.9309,
+      "step": 17660
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00018600522731428016,
+      "loss": 1.0316,
+      "step": 17665
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018599245100050883,
+      "loss": 1.0418,
+      "step": 17670
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018597966929663273,
+      "loss": 1.02,
+      "step": 17675
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018596688220345303,
+      "loss": 0.9424,
+      "step": 17680
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018595408972177127,
+      "loss": 1.0338,
+      "step": 17685
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018594129185238925,
+      "loss": 0.9645,
+      "step": 17690
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018592848859610918,
+      "loss": 0.9934,
+      "step": 17695
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001859156799537336,
+      "loss": 1.063,
+      "step": 17700
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001859028659260653,
+      "loss": 0.9458,
+      "step": 17705
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018589004651390758,
+      "loss": 0.9579,
+      "step": 17710
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.0001858772217180639,
+      "loss": 1.0698,
+      "step": 17715
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001858643915393382,
+      "loss": 0.9331,
+      "step": 17720
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001858515559785346,
+      "loss": 0.8905,
+      "step": 17725
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001858387150364577,
+      "loss": 0.9107,
+      "step": 17730
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018582586871391236,
+      "loss": 1.063,
+      "step": 17735
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018581301701170387,
+      "loss": 1.0574,
+      "step": 17740
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001858001599306377,
+      "loss": 1.01,
+      "step": 17745
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001857872974715198,
+      "loss": 0.9275,
+      "step": 17750
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001857744296351564,
+      "loss": 0.9735,
+      "step": 17755
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018576155642235407,
+      "loss": 1.0272,
+      "step": 17760
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001857486778339197,
+      "loss": 1.0457,
+      "step": 17765
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018573579387066053,
+      "loss": 0.9561,
+      "step": 17770
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018572290453338416,
+      "loss": 0.9612,
+      "step": 17775
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001857100098228985,
+      "loss": 0.8423,
+      "step": 17780
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018569710974001183,
+      "loss": 1.0384,
+      "step": 17785
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018568420428553272,
+      "loss": 1.1032,
+      "step": 17790
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018567129346027007,
+      "loss": 1.1377,
+      "step": 17795
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018565837726503318,
+      "loss": 0.9317,
+      "step": 17800
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018564545570063168,
+      "loss": 0.9791,
+      "step": 17805
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018563252876787546,
+      "loss": 0.9285,
+      "step": 17810
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018561959646757478,
+      "loss": 0.8951,
+      "step": 17815
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018560665880054032,
+      "loss": 0.9693,
+      "step": 17820
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018559371576758296,
+      "loss": 0.8104,
+      "step": 17825
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018558076736951404,
+      "loss": 0.9264,
+      "step": 17830
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00018556781360714512,
+      "loss": 1.1434,
+      "step": 17835
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001855548544812882,
+      "loss": 0.9131,
+      "step": 17840
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001855418899927556,
+      "loss": 1.0054,
+      "step": 17845
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018552892014235986,
+      "loss": 0.9281,
+      "step": 17850
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.625,
+      "learning_rate": 0.000185515944930914,
+      "loss": 1.033,
+      "step": 17855
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018550296435923135,
+      "loss": 0.9832,
+      "step": 17860
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018548997842812552,
+      "loss": 1.0676,
+      "step": 17865
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018547698713841047,
+      "loss": 0.847,
+      "step": 17870
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018546399049090048,
+      "loss": 0.8618,
+      "step": 17875
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018545098848641025,
+      "loss": 0.9026,
+      "step": 17880
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018543798112575474,
+      "loss": 1.1252,
+      "step": 17885
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001854249684097493,
+      "loss": 0.8721,
+      "step": 17890
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001854119503392095,
+      "loss": 0.9944,
+      "step": 17895
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001853989269149514,
+      "loss": 0.9431,
+      "step": 17900
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001853858981377913,
+      "loss": 1.072,
+      "step": 17905
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018537286400854583,
+      "loss": 0.9626,
+      "step": 17910
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018535982452803204,
+      "loss": 0.9894,
+      "step": 17915
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018534677969706724,
+      "loss": 0.9762,
+      "step": 17920
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018533372951646908,
+      "loss": 0.9588,
+      "step": 17925
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018532067398705556,
+      "loss": 0.828,
+      "step": 17930
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018530761310964504,
+      "loss": 1.1494,
+      "step": 17935
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00018529454688505614,
+      "loss": 1.1224,
+      "step": 17940
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018528147531410793,
+      "loss": 0.931,
+      "step": 17945
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018526839839761968,
+      "loss": 0.9902,
+      "step": 17950
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018525531613641113,
+      "loss": 0.9354,
+      "step": 17955
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001852422285313023,
+      "loss": 0.9686,
+      "step": 17960
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018522913558311345,
+      "loss": 0.9746,
+      "step": 17965
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018521603729266536,
+      "loss": 0.9197,
+      "step": 17970
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018520293366077896,
+      "loss": 1.0712,
+      "step": 17975
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018518982468827567,
+      "loss": 0.9947,
+      "step": 17980
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018517671037597712,
+      "loss": 0.8395,
+      "step": 17985
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018516359072470536,
+      "loss": 0.9635,
+      "step": 17990
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018515046573528275,
+      "loss": 0.9441,
+      "step": 17995
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018513733540853196,
+      "loss": 0.9905,
+      "step": 18000
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018512419974527604,
+      "loss": 0.9265,
+      "step": 18005
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018511105874633832,
+      "loss": 0.9096,
+      "step": 18010
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018509791241254253,
+      "loss": 0.9061,
+      "step": 18015
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018508476074471263,
+      "loss": 1.0156,
+      "step": 18020
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018507160374367306,
+      "loss": 1.0067,
+      "step": 18025
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018505844141024843,
+      "loss": 0.9401,
+      "step": 18030
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018504527374526387,
+      "loss": 1.0436,
+      "step": 18035
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018503210074954468,
+      "loss": 1.0336,
+      "step": 18040
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018501892242391653,
+      "loss": 0.9606,
+      "step": 18045
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018500573876920555,
+      "loss": 0.9777,
+      "step": 18050
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.000184992549786238,
+      "loss": 0.9084,
+      "step": 18055
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018497935547584068,
+      "loss": 1.0103,
+      "step": 18060
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018496615583884056,
+      "loss": 0.9744,
+      "step": 18065
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.000184952950876065,
+      "loss": 0.9016,
+      "step": 18070
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018493974058834178,
+      "loss": 0.8987,
+      "step": 18075
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018492652497649882,
+      "loss": 0.8471,
+      "step": 18080
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.00018491330404136458,
+      "loss": 1.0032,
+      "step": 18085
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018490007778376776,
+      "loss": 0.8814,
+      "step": 18090
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018488684620453731,
+      "loss": 0.9529,
+      "step": 18095
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018487360930450272,
+      "loss": 1.0356,
+      "step": 18100
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001848603670844936,
+      "loss": 1.0452,
+      "step": 18105
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00018484711954534002,
+      "loss": 0.723,
+      "step": 18110
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001848338666878724,
+      "loss": 1.0693,
+      "step": 18115
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018482060851292132,
+      "loss": 0.9527,
+      "step": 18120
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018480734502131796,
+      "loss": 0.9394,
+      "step": 18125
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018479407621389363,
+      "loss": 0.8132,
+      "step": 18130
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018478080209148,
+      "loss": 1.0694,
+      "step": 18135
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018476752265490914,
+      "loss": 0.9333,
+      "step": 18140
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018475423790501342,
+      "loss": 1.0483,
+      "step": 18145
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018474094784262554,
+      "loss": 0.9722,
+      "step": 18150
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018472765246857855,
+      "loss": 1.1281,
+      "step": 18155
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001847143517837058,
+      "loss": 1.0055,
+      "step": 18160
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018470104578884099,
+      "loss": 0.8147,
+      "step": 18165
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018468773448481818,
+      "loss": 1.1144,
+      "step": 18170
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001846744178724717,
+      "loss": 0.9699,
+      "step": 18175
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001846610959526363,
+      "loss": 1.0632,
+      "step": 18180
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018464776872614697,
+      "loss": 0.8508,
+      "step": 18185
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001846344361938391,
+      "loss": 1.0169,
+      "step": 18190
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018462109835654838,
+      "loss": 1.0482,
+      "step": 18195
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018460775521511082,
+      "loss": 0.9038,
+      "step": 18200
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001845944067703628,
+      "loss": 0.9428,
+      "step": 18205
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018458105302314104,
+      "loss": 0.8876,
+      "step": 18210
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018456769397428254,
+      "loss": 0.8324,
+      "step": 18215
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018455432962462466,
+      "loss": 1.1117,
+      "step": 18220
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001845409599750051,
+      "loss": 0.87,
+      "step": 18225
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001845275850262619,
+      "loss": 0.8693,
+      "step": 18230
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018451420477923338,
+      "loss": 1.0202,
+      "step": 18235
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00018450081923475828,
+      "loss": 1.0089,
+      "step": 18240
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018448742839367557,
+      "loss": 1.0084,
+      "step": 18245
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018447403225682464,
+      "loss": 0.797,
+      "step": 18250
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018446063082504512,
+      "loss": 1.0342,
+      "step": 18255
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001844472240991771,
+      "loss": 1.0519,
+      "step": 18260
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001844338120800609,
+      "loss": 0.9145,
+      "step": 18265
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001844203947685372,
+      "loss": 0.8796,
+      "step": 18270
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.000184406972165447,
+      "loss": 0.8601,
+      "step": 18275
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018439354427163162,
+      "loss": 0.9921,
+      "step": 18280
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00018438011108793282,
+      "loss": 0.91,
+      "step": 18285
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018436667261519254,
+      "loss": 1.024,
+      "step": 18290
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018435322885425312,
+      "loss": 0.9442,
+      "step": 18295
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00018433977980595727,
+      "loss": 0.9701,
+      "step": 18300
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018432632547114795,
+      "loss": 1.0441,
+      "step": 18305
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018431286585066851,
+      "loss": 0.8904,
+      "step": 18310
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001842994009453626,
+      "loss": 0.913,
+      "step": 18315
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018428593075607425,
+      "loss": 0.8808,
+      "step": 18320
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018427245528364778,
+      "loss": 0.9081,
+      "step": 18325
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018425897452892782,
+      "loss": 0.9171,
+      "step": 18330
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018424548849275935,
+      "loss": 0.8542,
+      "step": 18335
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018423199717598776,
+      "loss": 0.9848,
+      "step": 18340
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018421850057945863,
+      "loss": 0.8988,
+      "step": 18345
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018420499870401796,
+      "loss": 0.9853,
+      "step": 18350
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018419149155051207,
+      "loss": 0.9578,
+      "step": 18355
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001841779791197876,
+      "loss": 0.8141,
+      "step": 18360
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018416446141269156,
+      "loss": 0.8497,
+      "step": 18365
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001841509384300712,
+      "loss": 0.9357,
+      "step": 18370
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001841374101727742,
+      "loss": 1.0044,
+      "step": 18375
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018412387664164847,
+      "loss": 0.9814,
+      "step": 18380
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018411033783754234,
+      "loss": 0.9315,
+      "step": 18385
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00018409679376130445,
+      "loss": 0.9981,
+      "step": 18390
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001840832444137838,
+      "loss": 1.0124,
+      "step": 18395
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018406968979582956,
+      "loss": 0.9862,
+      "step": 18400
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018405612990829147,
+      "loss": 0.9314,
+      "step": 18405
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018404256475201938,
+      "loss": 0.9507,
+      "step": 18410
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018402899432786365,
+      "loss": 0.9749,
+      "step": 18415
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018401541863667485,
+      "loss": 0.9224,
+      "step": 18420
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018400183767930387,
+      "loss": 1.1241,
+      "step": 18425
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018398825145660212,
+      "loss": 0.8608,
+      "step": 18430
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00018397465996942107,
+      "loss": 1.0383,
+      "step": 18435
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018396106321861267,
+      "loss": 0.915,
+      "step": 18440
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018394746120502922,
+      "loss": 0.95,
+      "step": 18445
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001839338539295233,
+      "loss": 0.9485,
+      "step": 18450
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018392024139294785,
+      "loss": 0.8437,
+      "step": 18455
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018390662359615603,
+      "loss": 1.0354,
+      "step": 18460
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018389300054000155,
+      "loss": 1.0424,
+      "step": 18465
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018387937222533825,
+      "loss": 1.1084,
+      "step": 18470
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001838657386530203,
+      "loss": 0.909,
+      "step": 18475
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001838520998239024,
+      "loss": 1.0199,
+      "step": 18480
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018383845573883932,
+      "loss": 0.8707,
+      "step": 18485
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001838248063986864,
+      "loss": 1.0918,
+      "step": 18490
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018381115180429912,
+      "loss": 0.9805,
+      "step": 18495
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018379749195653343,
+      "loss": 0.9804,
+      "step": 18500
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018378382685624547,
+      "loss": 0.9463,
+      "step": 18505
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018377015650429182,
+      "loss": 1.065,
+      "step": 18510
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018375648090152938,
+      "loss": 0.9703,
+      "step": 18515
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018374280004881531,
+      "loss": 0.8268,
+      "step": 18520
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018372911394700717,
+      "loss": 1.0015,
+      "step": 18525
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001837154225969628,
+      "loss": 0.8507,
+      "step": 18530
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018370172599954041,
+      "loss": 1.0905,
+      "step": 18535
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001836880241555985,
+      "loss": 0.9969,
+      "step": 18540
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001836743170659959,
+      "loss": 0.9468,
+      "step": 18545
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018366060473159183,
+      "loss": 0.9021,
+      "step": 18550
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001836468871532458,
+      "loss": 0.8949,
+      "step": 18555
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00018363316433181757,
+      "loss": 1.0498,
+      "step": 18560
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018361943626816736,
+      "loss": 1.1729,
+      "step": 18565
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018360570296315566,
+      "loss": 1.0258,
+      "step": 18570
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018359196441764328,
+      "loss": 0.9827,
+      "step": 18575
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018357822063249136,
+      "loss": 0.9098,
+      "step": 18580
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001835644716085614,
+      "loss": 0.9919,
+      "step": 18585
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018355071734671517,
+      "loss": 1.004,
+      "step": 18590
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001835369578478148,
+      "loss": 0.9617,
+      "step": 18595
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001835231931127228,
+      "loss": 0.9761,
+      "step": 18600
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00018350942314230195,
+      "loss": 0.8244,
+      "step": 18605
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018349564793741533,
+      "loss": 0.97,
+      "step": 18610
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018348186749892639,
+      "loss": 1.1368,
+      "step": 18615
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001834680818276989,
+      "loss": 1.0208,
+      "step": 18620
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018345429092459704,
+      "loss": 0.8885,
+      "step": 18625
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018344049479048513,
+      "loss": 1.0854,
+      "step": 18630
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.000183426693426228,
+      "loss": 1.1566,
+      "step": 18635
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001834128868326907,
+      "loss": 0.9691,
+      "step": 18640
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018339907501073867,
+      "loss": 0.968,
+      "step": 18645
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001833852579612376,
+      "loss": 0.9445,
+      "step": 18650
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00018337143568505362,
+      "loss": 1.0196,
+      "step": 18655
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018335760818305309,
+      "loss": 0.914,
+      "step": 18660
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018334377545610274,
+      "loss": 0.9988,
+      "step": 18665
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018332993750506962,
+      "loss": 0.7872,
+      "step": 18670
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00018331609433082114,
+      "loss": 0.9388,
+      "step": 18675
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018330224593422496,
+      "loss": 0.9146,
+      "step": 18680
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018328839231614911,
+      "loss": 0.8824,
+      "step": 18685
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018327453347746203,
+      "loss": 1.1295,
+      "step": 18690
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018326066941903228,
+      "loss": 1.0299,
+      "step": 18695
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.000183246800141729,
+      "loss": 0.9898,
+      "step": 18700
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018323292564642146,
+      "loss": 1.019,
+      "step": 18705
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001832190459339793,
+      "loss": 0.89,
+      "step": 18710
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001832051610052726,
+      "loss": 0.9524,
+      "step": 18715
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018319127086117168,
+      "loss": 0.9742,
+      "step": 18720
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018317737550254713,
+      "loss": 1.1546,
+      "step": 18725
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018316347493026994,
+      "loss": 0.9396,
+      "step": 18730
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00018314956914521142,
+      "loss": 0.8559,
+      "step": 18735
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001831356581482432,
+      "loss": 1.0151,
+      "step": 18740
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001831217419402373,
+      "loss": 0.8159,
+      "step": 18745
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001831078205220659,
+      "loss": 1.0798,
+      "step": 18750
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018309389389460168,
+      "loss": 0.9936,
+      "step": 18755
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00018307996205871755,
+      "loss": 0.984,
+      "step": 18760
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018306602501528673,
+      "loss": 0.8814,
+      "step": 18765
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00018305208276518293,
+      "loss": 0.9234,
+      "step": 18770
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018303813530927995,
+      "loss": 0.908,
+      "step": 18775
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018302418264845208,
+      "loss": 0.8875,
+      "step": 18780
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018301022478357391,
+      "loss": 0.9008,
+      "step": 18785
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001829962617155203,
+      "loss": 1.0306,
+      "step": 18790
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018298229344516646,
+      "loss": 0.9696,
+      "step": 18795
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018296831997338797,
+      "loss": 0.9603,
+      "step": 18800
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001829543413010607,
+      "loss": 0.8585,
+      "step": 18805
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001829403574290608,
+      "loss": 0.8206,
+      "step": 18810
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001829263683582649,
+      "loss": 0.9683,
+      "step": 18815
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018291237408954976,
+      "loss": 0.9284,
+      "step": 18820
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00018289837462379257,
+      "loss": 1.0969,
+      "step": 18825
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018288436996187084,
+      "loss": 1.0336,
+      "step": 18830
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018287036010466244,
+      "loss": 0.9165,
+      "step": 18835
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018285634505304545,
+      "loss": 0.9036,
+      "step": 18840
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018284232480789841,
+      "loss": 0.8956,
+      "step": 18845
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018282829937010009,
+      "loss": 0.9551,
+      "step": 18850
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00018281426874052961,
+      "loss": 0.9882,
+      "step": 18855
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018280023292006648,
+      "loss": 1.0658,
+      "step": 18860
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018278619190959045,
+      "loss": 0.8386,
+      "step": 18865
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018277214570998161,
+      "loss": 0.9481,
+      "step": 18870
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00018275809432212041,
+      "loss": 1.0821,
+      "step": 18875
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001827440377468876,
+      "loss": 0.9736,
+      "step": 18880
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018272997598516431,
+      "loss": 0.9497,
+      "step": 18885
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018271590903783184,
+      "loss": 0.9594,
+      "step": 18890
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018270183690577202,
+      "loss": 0.9803,
+      "step": 18895
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018268775958986687,
+      "loss": 0.9781,
+      "step": 18900
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018267367709099878,
+      "loss": 1.1337,
+      "step": 18905
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018265958941005044,
+      "loss": 0.8523,
+      "step": 18910
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018264549654790487,
+      "loss": 0.8817,
+      "step": 18915
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018263139850544551,
+      "loss": 1.0435,
+      "step": 18920
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018261729528355595,
+      "loss": 1.0755,
+      "step": 18925
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001826031868831202,
+      "loss": 1.0781,
+      "step": 18930
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018258907330502265,
+      "loss": 1.0118,
+      "step": 18935
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018257495455014798,
+      "loss": 0.8649,
+      "step": 18940
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00018256083061938104,
+      "loss": 0.984,
+      "step": 18945
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018254670151360722,
+      "loss": 1.0077,
+      "step": 18950
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018253256723371216,
+      "loss": 1.1296,
+      "step": 18955
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018251842778058177,
+      "loss": 0.965,
+      "step": 18960
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018250428315510234,
+      "loss": 0.9761,
+      "step": 18965
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018249013335816048,
+      "loss": 0.9319,
+      "step": 18970
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001824759783906431,
+      "loss": 1.0028,
+      "step": 18975
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001824618182534375,
+      "loss": 1.0927,
+      "step": 18980
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018244765294743117,
+      "loss": 0.8849,
+      "step": 18985
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018243348247351212,
+      "loss": 0.924,
+      "step": 18990
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018241930683256846,
+      "loss": 0.892,
+      "step": 18995
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018240512602548875,
+      "loss": 1.0804,
+      "step": 19000
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018239094005316193,
+      "loss": 0.8454,
+      "step": 19005
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018237674891647716,
+      "loss": 0.9701,
+      "step": 19010
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018236255261632392,
+      "loss": 0.9946,
+      "step": 19015
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001823483511535921,
+      "loss": 1.0154,
+      "step": 19020
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018233414452917184,
+      "loss": 0.9028,
+      "step": 19025
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018231993274395362,
+      "loss": 0.8365,
+      "step": 19030
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018230571579882826,
+      "loss": 0.7945,
+      "step": 19035
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001822914936946869,
+      "loss": 1.0,
+      "step": 19040
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.000182277266432421,
+      "loss": 1.0195,
+      "step": 19045
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018226303401292233,
+      "loss": 1.0113,
+      "step": 19050
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018224879643708299,
+      "loss": 0.9697,
+      "step": 19055
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00018223455370579544,
+      "loss": 1.0527,
+      "step": 19060
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018222030581995237,
+      "loss": 1.1755,
+      "step": 19065
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018220605278044692,
+      "loss": 0.8671,
+      "step": 19070
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018219179458817247,
+      "loss": 0.9569,
+      "step": 19075
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018217753124402268,
+      "loss": 0.9445,
+      "step": 19080
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018216326274889165,
+      "loss": 1.0961,
+      "step": 19085
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018214898910367375,
+      "loss": 0.9009,
+      "step": 19090
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.82421875,
+      "learning_rate": 0.00018213471030926367,
+      "loss": 1.0007,
+      "step": 19095
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018212042636655637,
+      "loss": 1.0147,
+      "step": 19100
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00018210613727644723,
+      "loss": 0.8774,
+      "step": 19105
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001820918430398319,
+      "loss": 0.9651,
+      "step": 19110
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018207754365760637,
+      "loss": 1.0168,
+      "step": 19115
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00018206323913066687,
+      "loss": 1.0255,
+      "step": 19120
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018204892945991014,
+      "loss": 1.0751,
+      "step": 19125
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018203461464623302,
+      "loss": 0.9231,
+      "step": 19130
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00018202029469053285,
+      "loss": 1.022,
+      "step": 19135
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00018200596959370722,
+      "loss": 1.0119,
+      "step": 19140
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018199163935665396,
+      "loss": 0.9225,
+      "step": 19145
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018197730398027142,
+      "loss": 1.0322,
+      "step": 19150
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018196296346545805,
+      "loss": 1.0106,
+      "step": 19155
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018194861781311282,
+      "loss": 0.9608,
+      "step": 19160
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018193426702413487,
+      "loss": 0.9761,
+      "step": 19165
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018191991109942377,
+      "loss": 0.8546,
+      "step": 19170
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001819055500398793,
+      "loss": 0.9388,
+      "step": 19175
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018189118384640172,
+      "loss": 0.909,
+      "step": 19180
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001818768125198915,
+      "loss": 1.0421,
+      "step": 19185
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018186243606124934,
+      "loss": 0.8726,
+      "step": 19190
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001818480544713765,
+      "loss": 0.9523,
+      "step": 19195
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018183366775117437,
+      "loss": 0.9786,
+      "step": 19200
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018181927590154475,
+      "loss": 0.8752,
+      "step": 19205
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00018180487892338973,
+      "loss": 0.7925,
+      "step": 19210
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018179047681761174,
+      "loss": 0.957,
+      "step": 19215
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001817760695851135,
+      "loss": 1.0573,
+      "step": 19220
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001817616572267981,
+      "loss": 0.8925,
+      "step": 19225
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001817472397435689,
+      "loss": 0.8296,
+      "step": 19230
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001817328171363296,
+      "loss": 0.9467,
+      "step": 19235
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018171838940598425,
+      "loss": 0.8865,
+      "step": 19240
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018170395655343717,
+      "loss": 0.9796,
+      "step": 19245
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018168951857959305,
+      "loss": 0.8189,
+      "step": 19250
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00018167507548535685,
+      "loss": 0.9377,
+      "step": 19255
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018166062727163393,
+      "loss": 1.1658,
+      "step": 19260
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00018164617393932986,
+      "loss": 0.9539,
+      "step": 19265
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018163171548935062,
+      "loss": 1.0121,
+      "step": 19270
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018161725192260254,
+      "loss": 0.9584,
+      "step": 19275
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001816027832399921,
+      "loss": 0.9901,
+      "step": 19280
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018158830944242627,
+      "loss": 0.8591,
+      "step": 19285
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001815738305308123,
+      "loss": 0.9036,
+      "step": 19290
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001815593465060577,
+      "loss": 1.0187,
+      "step": 19295
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001815448573690704,
+      "loss": 0.856,
+      "step": 19300
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00018153036312075854,
+      "loss": 0.9173,
+      "step": 19305
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00018151586376203072,
+      "loss": 0.9322,
+      "step": 19310
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018150135929379565,
+      "loss": 0.9743,
+      "step": 19315
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001814868497169626,
+      "loss": 0.9384,
+      "step": 19320
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.000181472335032441,
+      "loss": 0.9545,
+      "step": 19325
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018145781524114068,
+      "loss": 0.9326,
+      "step": 19330
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018144329034397167,
+      "loss": 0.7758,
+      "step": 19335
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001814287603418445,
+      "loss": 0.9259,
+      "step": 19340
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018141422523566987,
+      "loss": 0.7652,
+      "step": 19345
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018139968502635888,
+      "loss": 1.0755,
+      "step": 19350
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018138513971482296,
+      "loss": 0.9437,
+      "step": 19355
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018137058930197376,
+      "loss": 0.9806,
+      "step": 19360
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018135603378872337,
+      "loss": 0.9889,
+      "step": 19365
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001813414731759841,
+      "loss": 0.8439,
+      "step": 19370
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00018132690746466867,
+      "loss": 1.0547,
+      "step": 19375
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018131233665569005,
+      "loss": 0.7771,
+      "step": 19380
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018129776074996156,
+      "loss": 1.1168,
+      "step": 19385
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018128317974839685,
+      "loss": 0.9792,
+      "step": 19390
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018126859365190986,
+      "loss": 0.9773,
+      "step": 19395
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00018125400246141486,
+      "loss": 1.0232,
+      "step": 19400
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018123940617782643,
+      "loss": 1.1173,
+      "step": 19405
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001812248048020595,
+      "loss": 1.0651,
+      "step": 19410
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001812101983350293,
+      "loss": 1.0051,
+      "step": 19415
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001811955867776514,
+      "loss": 0.995,
+      "step": 19420
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001811809701308416,
+      "loss": 0.9651,
+      "step": 19425
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018116634839551618,
+      "loss": 1.0279,
+      "step": 19430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018115172157259158,
+      "loss": 1.0307,
+      "step": 19435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018113708966298466,
+      "loss": 0.8745,
+      "step": 19440
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018112245266761255,
+      "loss": 0.8352,
+      "step": 19445
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001811078105873927,
+      "loss": 0.9852,
+      "step": 19450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001810931634232429,
+      "loss": 0.9463,
+      "step": 19455
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00018107851117608127,
+      "loss": 0.965,
+      "step": 19460
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00018106385384682625,
+      "loss": 0.9346,
+      "step": 19465
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00018104919143639654,
+      "loss": 1.0714,
+      "step": 19470
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00018103452394571117,
+      "loss": 0.9186,
+      "step": 19475
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00018101985137568955,
+      "loss": 1.0589,
+      "step": 19480
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018100517372725142,
+      "loss": 0.9316,
+      "step": 19485
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001809904910013167,
+      "loss": 0.9155,
+      "step": 19490
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.796875,
+      "learning_rate": 0.00018097580319880577,
+      "loss": 0.9344,
+      "step": 19495
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018096111032063928,
+      "loss": 1.1705,
+      "step": 19500
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018094641236773818,
+      "loss": 0.9314,
+      "step": 19505
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018093170934102378,
+      "loss": 0.8923,
+      "step": 19510
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018091700124141764,
+      "loss": 1.0893,
+      "step": 19515
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001809022880698417,
+      "loss": 0.9808,
+      "step": 19520
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00018088756982721825,
+      "loss": 0.9501,
+      "step": 19525
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00018087284651446977,
+      "loss": 0.9246,
+      "step": 19530
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018085811813251917,
+      "loss": 0.9049,
+      "step": 19535
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0001808433846822896,
+      "loss": 1.1336,
+      "step": 19540
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00018082864616470468,
+      "loss": 1.003,
+      "step": 19545
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00018081390258068808,
+      "loss": 1.0711,
+      "step": 19550
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018079915393116405,
+      "loss": 1.0544,
+      "step": 19555
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00018078440021705708,
+      "loss": 1.1562,
+      "step": 19560
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00018076964143929188,
+      "loss": 1.1116,
+      "step": 19565
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018075487759879353,
+      "loss": 0.9089,
+      "step": 19570
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001807401086964875,
+      "loss": 0.9719,
+      "step": 19575
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00018072533473329952,
+      "loss": 1.1497,
+      "step": 19580
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001807105557101556,
+      "loss": 0.9319,
+      "step": 19585
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001806957716279821,
+      "loss": 0.8869,
+      "step": 19590
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018068098248770576,
+      "loss": 0.9751,
+      "step": 19595
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00018066618829025354,
+      "loss": 0.8765,
+      "step": 19600
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001806513890365528,
+      "loss": 1.1602,
+      "step": 19605
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001806365847275311,
+      "loss": 0.9074,
+      "step": 19610
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018062177536411645,
+      "loss": 0.909,
+      "step": 19615
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00018060696094723708,
+      "loss": 0.8727,
+      "step": 19620
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00018059214147782163,
+      "loss": 1.0834,
+      "step": 19625
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018057731695679893,
+      "loss": 0.9149,
+      "step": 19630
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00018056248738509826,
+      "loss": 1.0154,
+      "step": 19635
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001805476527636491,
+      "loss": 1.0771,
+      "step": 19640
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018053281309338135,
+      "loss": 1.0866,
+      "step": 19645
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018051796837522516,
+      "loss": 0.9696,
+      "step": 19650
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.000180503118610111,
+      "loss": 0.7969,
+      "step": 19655
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018048826379896967,
+      "loss": 0.9516,
+      "step": 19660
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00018047340394273232,
+      "loss": 0.9688,
+      "step": 19665
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00018045853904233034,
+      "loss": 1.0596,
+      "step": 19670
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00018044366909869552,
+      "loss": 0.9945,
+      "step": 19675
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00018042879411275987,
+      "loss": 1.033,
+      "step": 19680
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00018041391408545586,
+      "loss": 0.9772,
+      "step": 19685
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018039902901771608,
+      "loss": 1.0647,
+      "step": 19690
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00018038413891047358,
+      "loss": 0.9532,
+      "step": 19695
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00018036924376466174,
+      "loss": 0.801,
+      "step": 19700
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018035434358121418,
+      "loss": 0.9809,
+      "step": 19705
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00018033943836106482,
+      "loss": 1.0384,
+      "step": 19710
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00018032452810514798,
+      "loss": 0.9302,
+      "step": 19715
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001803096128143982,
+      "loss": 1.0039,
+      "step": 19720
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00018029469248975047,
+      "loss": 1.1232,
+      "step": 19725
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00018027976713213994,
+      "loss": 0.9102,
+      "step": 19730
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001802648367425022,
+      "loss": 1.1901,
+      "step": 19735
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00018024990132177305,
+      "loss": 0.8535,
+      "step": 19740
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00018023496087088872,
+      "loss": 0.9464,
+      "step": 19745
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018022001539078563,
+      "loss": 1.0605,
+      "step": 19750
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00018020506488240065,
+      "loss": 0.892,
+      "step": 19755
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00018019010934667082,
+      "loss": 0.8213,
+      "step": 19760
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00018017514878453363,
+      "loss": 0.9154,
+      "step": 19765
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001801601831969268,
+      "loss": 1.0438,
+      "step": 19770
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00018014521258478839,
+      "loss": 0.9831,
+      "step": 19775
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00018013023694905678,
+      "loss": 1.0661,
+      "step": 19780
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00018011525629067063,
+      "loss": 0.8326,
+      "step": 19785
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.000180100270610569,
+      "loss": 1.0475,
+      "step": 19790
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018008527990969118,
+      "loss": 0.9246,
+      "step": 19795
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001800702841889768,
+      "loss": 1.0556,
+      "step": 19800
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00018005528344936582,
+      "loss": 1.0302,
+      "step": 19805
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001800402776917985,
+      "loss": 1.0522,
+      "step": 19810
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001800252669172154,
+      "loss": 0.9603,
+      "step": 19815
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00018001025112655743,
+      "loss": 1.0379,
+      "step": 19820
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001799952303207658,
+      "loss": 0.9,
+      "step": 19825
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017998020450078203,
+      "loss": 0.9419,
+      "step": 19830
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017996517366754798,
+      "loss": 0.8759,
+      "step": 19835
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00017995013782200574,
+      "loss": 1.1232,
+      "step": 19840
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001799350969650978,
+      "loss": 0.8869,
+      "step": 19845
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017992005109776694,
+      "loss": 0.9544,
+      "step": 19850
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001799050002209563,
+      "loss": 0.8654,
+      "step": 19855
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001798899443356092,
+      "loss": 0.879,
+      "step": 19860
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017987488344266942,
+      "loss": 1.0349,
+      "step": 19865
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017985981754308096,
+      "loss": 0.9102,
+      "step": 19870
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001798447466377882,
+      "loss": 0.8759,
+      "step": 19875
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017982967072773578,
+      "loss": 1.0562,
+      "step": 19880
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017981458981386868,
+      "loss": 0.9362,
+      "step": 19885
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017979950389713218,
+      "loss": 0.9396,
+      "step": 19890
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001797844129784719,
+      "loss": 1.1549,
+      "step": 19895
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00017976931705883376,
+      "loss": 1.0361,
+      "step": 19900
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017975421613916395,
+      "loss": 0.9686,
+      "step": 19905
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017973911022040905,
+      "loss": 0.9509,
+      "step": 19910
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00017972399930351593,
+      "loss": 1.1096,
+      "step": 19915
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00017970888338943172,
+      "loss": 1.0179,
+      "step": 19920
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.79296875,
+      "learning_rate": 0.00017969376247910392,
+      "loss": 1.0277,
+      "step": 19925
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001796786365734803,
+      "loss": 0.891,
+      "step": 19930
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017966350567350902,
+      "loss": 0.9891,
+      "step": 19935
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017964836978013845,
+      "loss": 1.0103,
+      "step": 19940
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001796332288943174,
+      "loss": 1.0574,
+      "step": 19945
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00017961808301699482,
+      "loss": 0.9753,
+      "step": 19950
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001796029321491201,
+      "loss": 1.0732,
+      "step": 19955
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017958777629164301,
+      "loss": 0.888,
+      "step": 19960
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017957261544551342,
+      "loss": 0.9886,
+      "step": 19965
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00017955744961168163,
+      "loss": 0.9075,
+      "step": 19970
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017954227879109834,
+      "loss": 0.9966,
+      "step": 19975
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00017952710298471442,
+      "loss": 1.0853,
+      "step": 19980
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001795119221934811,
+      "loss": 0.9075,
+      "step": 19985
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00017949673641834993,
+      "loss": 0.956,
+      "step": 19990
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001794815456602728,
+      "loss": 0.9857,
+      "step": 19995
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017946634992020187,
+      "loss": 1.0116,
+      "step": 20000
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017945114919908962,
+      "loss": 1.0116,
+      "step": 20005
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00017943594349788882,
+      "loss": 1.0758,
+      "step": 20010
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017942073281755264,
+      "loss": 0.9998,
+      "step": 20015
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00017940551715903448,
+      "loss": 1.0343,
+      "step": 20020
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017939029652328805,
+      "loss": 1.0741,
+      "step": 20025
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00017937507091126743,
+      "loss": 1.0459,
+      "step": 20030
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00017935984032392695,
+      "loss": 0.9559,
+      "step": 20035
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001793446047622213,
+      "loss": 0.8793,
+      "step": 20040
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017932936422710547,
+      "loss": 0.888,
+      "step": 20045
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017931411871953474,
+      "loss": 0.8765,
+      "step": 20050
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001792988682404647,
+      "loss": 1.0438,
+      "step": 20055
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017928361279085132,
+      "loss": 1.1002,
+      "step": 20060
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017926835237165074,
+      "loss": 1.0579,
+      "step": 20065
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001792530869838196,
+      "loss": 1.0192,
+      "step": 20070
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017923781662831467,
+      "loss": 1.0652,
+      "step": 20075
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017922254130609317,
+      "loss": 0.8573,
+      "step": 20080
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017920726101811255,
+      "loss": 0.848,
+      "step": 20085
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001791919757653306,
+      "loss": 0.8645,
+      "step": 20090
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017917668554870544,
+      "loss": 1.0636,
+      "step": 20095
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017916139036919544,
+      "loss": 0.9071,
+      "step": 20100
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001791460902277593,
+      "loss": 0.9749,
+      "step": 20105
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017913078512535611,
+      "loss": 0.9368,
+      "step": 20110
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001791154750629452,
+      "loss": 0.8993,
+      "step": 20115
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001791001600414862,
+      "loss": 0.8609,
+      "step": 20120
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0001790848400619391,
+      "loss": 1.0746,
+      "step": 20125
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017906951512526413,
+      "loss": 1.0527,
+      "step": 20130
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001790541852324219,
+      "loss": 0.9294,
+      "step": 20135
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017903885038437331,
+      "loss": 0.8542,
+      "step": 20140
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017902351058207957,
+      "loss": 0.9395,
+      "step": 20145
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00017900816582650214,
+      "loss": 1.0316,
+      "step": 20150
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00017899281611860295,
+      "loss": 0.8784,
+      "step": 20155
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017897746145934408,
+      "loss": 0.9608,
+      "step": 20160
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017896210184968793,
+      "loss": 1.044,
+      "step": 20165
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017894673729059734,
+      "loss": 0.8963,
+      "step": 20170
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017893136778303532,
+      "loss": 1.0971,
+      "step": 20175
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001789159933279653,
+      "loss": 0.9522,
+      "step": 20180
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017890061392635093,
+      "loss": 0.8735,
+      "step": 20185
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00017888522957915624,
+      "loss": 0.9023,
+      "step": 20190
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017886984028734547,
+      "loss": 0.8952,
+      "step": 20195
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001788544460518833,
+      "loss": 0.8191,
+      "step": 20200
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00017883904687373466,
+      "loss": 0.9671,
+      "step": 20205
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017882364275386477,
+      "loss": 1.1059,
+      "step": 20210
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00017880823369323918,
+      "loss": 0.9204,
+      "step": 20215
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00017879281969282373,
+      "loss": 0.9978,
+      "step": 20220
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017877740075358461,
+      "loss": 1.0445,
+      "step": 20225
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00017876197687648833,
+      "loss": 1.1439,
+      "step": 20230
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001787465480625016,
+      "loss": 0.9755,
+      "step": 20235
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017873111431259154,
+      "loss": 0.947,
+      "step": 20240
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00017871567562772559,
+      "loss": 0.9988,
+      "step": 20245
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017870023200887143,
+      "loss": 1.0291,
+      "step": 20250
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001786847834569971,
+      "loss": 0.8948,
+      "step": 20255
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017866932997307093,
+      "loss": 0.9763,
+      "step": 20260
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017865387155806156,
+      "loss": 0.9449,
+      "step": 20265
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017863840821293793,
+      "loss": 0.9888,
+      "step": 20270
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017862293993866935,
+      "loss": 0.9415,
+      "step": 20275
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00017860746673622533,
+      "loss": 1.0488,
+      "step": 20280
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00017859198860657575,
+      "loss": 0.9914,
+      "step": 20285
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001785765055506908,
+      "loss": 0.9855,
+      "step": 20290
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017856101756954105,
+      "loss": 1.1754,
+      "step": 20295
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017854552466409722,
+      "loss": 0.9697,
+      "step": 20300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00017853002683533044,
+      "loss": 0.9202,
+      "step": 20305
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017851452408421215,
+      "loss": 0.981,
+      "step": 20310
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00017849901641171404,
+      "loss": 0.9659,
+      "step": 20315
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017848350381880824,
+      "loss": 1.0127,
+      "step": 20320
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017846798630646697,
+      "loss": 0.8861,
+      "step": 20325
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00017845246387566296,
+      "loss": 0.9737,
+      "step": 20330
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017843693652736922,
+      "loss": 0.9233,
+      "step": 20335
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001784214042625589,
+      "loss": 0.8343,
+      "step": 20340
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017840586708220567,
+      "loss": 0.9862,
+      "step": 20345
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001783903249872834,
+      "loss": 0.9632,
+      "step": 20350
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017837477797876629,
+      "loss": 1.0639,
+      "step": 20355
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017835922605762884,
+      "loss": 0.9325,
+      "step": 20360
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001783436692248458,
+      "loss": 0.8678,
+      "step": 20365
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00017832810748139242,
+      "loss": 1.0376,
+      "step": 20370
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017831254082824403,
+      "loss": 0.9049,
+      "step": 20375
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017829696926637638,
+      "loss": 0.9619,
+      "step": 20380
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017828139279676551,
+      "loss": 0.9297,
+      "step": 20385
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017826581142038782,
+      "loss": 1.0391,
+      "step": 20390
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001782502251382199,
+      "loss": 1.0339,
+      "step": 20395
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017823463395123877,
+      "loss": 1.0506,
+      "step": 20400
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017821903786042166,
+      "loss": 1.0093,
+      "step": 20405
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001782034368667462,
+      "loss": 0.8458,
+      "step": 20410
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00017818783097119022,
+      "loss": 1.141,
+      "step": 20415
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00017817222017473198,
+      "loss": 0.891,
+      "step": 20420
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017815660447834995,
+      "loss": 0.838,
+      "step": 20425
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001781409838830229,
+      "loss": 0.9979,
+      "step": 20430
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00017812535838973004,
+      "loss": 1.0084,
+      "step": 20435
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001781097279994507,
+      "loss": 0.9987,
+      "step": 20440
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017809409271316465,
+      "loss": 1.0072,
+      "step": 20445
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00017807845253185194,
+      "loss": 0.9746,
+      "step": 20450
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00017806280745649288,
+      "loss": 1.0014,
+      "step": 20455
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017804715748806818,
+      "loss": 1.0265,
+      "step": 20460
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017803150262755873,
+      "loss": 0.874,
+      "step": 20465
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017801584287594583,
+      "loss": 0.8944,
+      "step": 20470
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00017800017823421102,
+      "loss": 0.9781,
+      "step": 20475
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00017798450870333625,
+      "loss": 0.9153,
+      "step": 20480
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017796883428430364,
+      "loss": 1.0358,
+      "step": 20485
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017795315497809571,
+      "loss": 1.0432,
+      "step": 20490
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00017793747078569523,
+      "loss": 0.943,
+      "step": 20495
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001779217817080853,
+      "loss": 0.8886,
+      "step": 20500
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017790608774624937,
+      "loss": 0.8394,
+      "step": 20505
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017789038890117113,
+      "loss": 0.9902,
+      "step": 20510
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017787468517383462,
+      "loss": 0.9307,
+      "step": 20515
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00017785897656522412,
+      "loss": 0.833,
+      "step": 20520
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001778432630763243,
+      "loss": 0.8931,
+      "step": 20525
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00017782754470812014,
+      "loss": 0.8259,
+      "step": 20530
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001778118214615968,
+      "loss": 0.9197,
+      "step": 20535
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00017779609333773988,
+      "loss": 0.7937,
+      "step": 20540
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001777803603375352,
+      "loss": 0.959,
+      "step": 20545
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5,
+      "learning_rate": 0.000177764622461969,
+      "loss": 0.7378,
+      "step": 20550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00017774887971202765,
+      "loss": 0.9992,
+      "step": 20555
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00017773313208869799,
+      "loss": 1.0006,
+      "step": 20560
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00017771737959296706,
+      "loss": 0.8735,
+      "step": 20565
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001777016222258223,
+      "loss": 0.8964,
+      "step": 20570
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017768585998825134,
+      "loss": 1.0407,
+      "step": 20575
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017767009288124222,
+      "loss": 0.9184,
+      "step": 20580
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001776543209057832,
+      "loss": 0.9547,
+      "step": 20585
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001776385440628629,
+      "loss": 0.994,
+      "step": 20590
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017762276235347026,
+      "loss": 0.8365,
+      "step": 20595
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00017760697577859447,
+      "loss": 0.9293,
+      "step": 20600
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.000177591184339225,
+      "loss": 1.171,
+      "step": 20605
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001775753880363518,
+      "loss": 1.011,
+      "step": 20610
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017755958687096487,
+      "loss": 0.9621,
+      "step": 20615
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017754378084405473,
+      "loss": 1.1206,
+      "step": 20620
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001775279699566121,
+      "loss": 0.9734,
+      "step": 20625
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017751215420962798,
+      "loss": 0.9819,
+      "step": 20630
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017749633360409378,
+      "loss": 1.086,
+      "step": 20635
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00017748050814100113,
+      "loss": 1.0491,
+      "step": 20640
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.000177464677821342,
+      "loss": 0.869,
+      "step": 20645
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017744884264610865,
+      "loss": 1.0623,
+      "step": 20650
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017743300261629358,
+      "loss": 0.8689,
+      "step": 20655
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00017741715773288976,
+      "loss": 1.1893,
+      "step": 20660
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001774013079968903,
+      "loss": 1.0839,
+      "step": 20665
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001773854534092887,
+      "loss": 0.8849,
+      "step": 20670
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017736959397107879,
+      "loss": 1.0321,
+      "step": 20675
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00017735372968325454,
+      "loss": 0.9963,
+      "step": 20680
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017733786054681047,
+      "loss": 1.074,
+      "step": 20685
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017732198656274123,
+      "loss": 1.1176,
+      "step": 20690
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017730610773204175,
+      "loss": 0.9677,
+      "step": 20695
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017729022405570745,
+      "loss": 1.0464,
+      "step": 20700
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017727433553473384,
+      "loss": 1.0147,
+      "step": 20705
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001772584421701169,
+      "loss": 0.9867,
+      "step": 20710
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001772425439628528,
+      "loss": 0.8194,
+      "step": 20715
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017722664091393808,
+      "loss": 0.9378,
+      "step": 20720
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017721073302436955,
+      "loss": 0.9764,
+      "step": 20725
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00017719482029514435,
+      "loss": 0.9781,
+      "step": 20730
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00017717890272725986,
+      "loss": 0.9302,
+      "step": 20735
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001771629803217139,
+      "loss": 0.9262,
+      "step": 20740
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001771470530795044,
+      "loss": 0.8817,
+      "step": 20745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017713112100162977,
+      "loss": 0.9105,
+      "step": 20750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00017711518408908866,
+      "loss": 0.8789,
+      "step": 20755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00017709924234287993,
+      "loss": 1.0325,
+      "step": 20760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017708329576400294,
+      "loss": 1.0813,
+      "step": 20765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017706734435345714,
+      "loss": 0.8091,
+      "step": 20770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017705138811224241,
+      "loss": 0.9214,
+      "step": 20775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017703542704135895,
+      "loss": 1.1351,
+      "step": 20780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001770194611418072,
+      "loss": 0.9709,
+      "step": 20785
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001770034904145879,
+      "loss": 1.0779,
+      "step": 20790
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001769875148607021,
+      "loss": 1.0441,
+      "step": 20795
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001769715344811512,
+      "loss": 1.1271,
+      "step": 20800
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017695554927693686,
+      "loss": 0.7982,
+      "step": 20805
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00017693955924906102,
+      "loss": 0.9824,
+      "step": 20810
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.000176923564398526,
+      "loss": 1.0395,
+      "step": 20815
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00017690756472633437,
+      "loss": 0.8858,
+      "step": 20820
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017689156023348898,
+      "loss": 1.0001,
+      "step": 20825
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.000176875550920993,
+      "loss": 0.9744,
+      "step": 20830
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00017685953678984998,
+      "loss": 0.8278,
+      "step": 20835
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001768435178410636,
+      "loss": 0.8934,
+      "step": 20840
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017682749407563805,
+      "loss": 0.958,
+      "step": 20845
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00017681146549457764,
+      "loss": 1.0769,
+      "step": 20850
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00017679543209888712,
+      "loss": 1.1344,
+      "step": 20855
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00017677939388957144,
+      "loss": 0.9309,
+      "step": 20860
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00017676335086763592,
+      "loss": 0.9748,
+      "step": 20865
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017674730303408613,
+      "loss": 1.0452,
+      "step": 20870
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00017673125038992796,
+      "loss": 1.0058,
+      "step": 20875
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017671519293616767,
+      "loss": 1.0156,
+      "step": 20880
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017669913067381167,
+      "loss": 0.8449,
+      "step": 20885
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00017668306360386685,
+      "loss": 0.9541,
+      "step": 20890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017666699172734022,
+      "loss": 0.931,
+      "step": 20895
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00017665091504523927,
+      "loss": 0.9573,
+      "step": 20900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00017663483355857164,
+      "loss": 0.9587,
+      "step": 20905
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001766187472683454,
+      "loss": 1.0087,
+      "step": 20910
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017660265617556878,
+      "loss": 1.1384,
+      "step": 20915
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00017658656028125046,
+      "loss": 1.1095,
+      "step": 20920
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017657045958639932,
+      "loss": 1.0633,
+      "step": 20925
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00017655435409202452,
+      "loss": 1.0446,
+      "step": 20930
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00017653824379913565,
+      "loss": 1.067,
+      "step": 20935
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001765221287087425,
+      "loss": 1.037,
+      "step": 20940
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00017650600882185513,
+      "loss": 0.9658,
+      "step": 20945
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.000176489884139484,
+      "loss": 0.9823,
+      "step": 20950
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017647375466263988,
+      "loss": 0.9067,
+      "step": 20955
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017645762039233368,
+      "loss": 0.9366,
+      "step": 20960
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017644148132957673,
+      "loss": 0.9275,
+      "step": 20965
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001764253374753807,
+      "loss": 0.8771,
+      "step": 20970
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017640918883075747,
+      "loss": 1.0067,
+      "step": 20975
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00017639303539671928,
+      "loss": 1.0476,
+      "step": 20980
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00017637687717427859,
+      "loss": 1.1025,
+      "step": 20985
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00017636071416444828,
+      "loss": 1.0378,
+      "step": 20990
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017634454636824144,
+      "loss": 0.9369,
+      "step": 20995
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001763283737866715,
+      "loss": 1.0305,
+      "step": 21000
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001763121964207521,
+      "loss": 1.0911,
+      "step": 21005
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00017629601427149736,
+      "loss": 0.963,
+      "step": 21010
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017627982733992153,
+      "loss": 0.9377,
+      "step": 21015
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001762636356270393,
+      "loss": 1.0163,
+      "step": 21020
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017624743913386546,
+      "loss": 1.0069,
+      "step": 21025
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017623123786141532,
+      "loss": 0.9348,
+      "step": 21030
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001762150318107044,
+      "loss": 0.9641,
+      "step": 21035
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017619882098274844,
+      "loss": 0.9824,
+      "step": 21040
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00017618260537856365,
+      "loss": 0.978,
+      "step": 21045
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00017616638499916634,
+      "loss": 1.2459,
+      "step": 21050
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001761501598455733,
+      "loss": 0.9898,
+      "step": 21055
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001761339299188015,
+      "loss": 1.0247,
+      "step": 21060
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001761176952198683,
+      "loss": 1.0085,
+      "step": 21065
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00017610145574979122,
+      "loss": 1.0736,
+      "step": 21070
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017608521150958825,
+      "loss": 1.0348,
+      "step": 21075
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017606896250027757,
+      "loss": 0.8328,
+      "step": 21080
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017605270872287767,
+      "loss": 0.8138,
+      "step": 21085
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001760364501784074,
+      "loss": 1.0117,
+      "step": 21090
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001760201868678858,
+      "loss": 0.9574,
+      "step": 21095
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017600391879233233,
+      "loss": 1.0016,
+      "step": 21100
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00017598764595276665,
+      "loss": 1.1719,
+      "step": 21105
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001759713683502088,
+      "loss": 0.8461,
+      "step": 21110
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00017595508598567902,
+      "loss": 0.9367,
+      "step": 21115
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00017593879886019799,
+      "loss": 1.0118,
+      "step": 21120
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017592250697478652,
+      "loss": 0.8151,
+      "step": 21125
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001759062103304659,
+      "loss": 0.8744,
+      "step": 21130
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001758899089282575,
+      "loss": 0.853,
+      "step": 21135
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00017587360276918323,
+      "loss": 0.9114,
+      "step": 21140
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001758572918542651,
+      "loss": 1.012,
+      "step": 21145
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017584097618452552,
+      "loss": 0.9139,
+      "step": 21150
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001758246557609872,
+      "loss": 0.8946,
+      "step": 21155
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001758083305846731,
+      "loss": 1.1681,
+      "step": 21160
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00017579200065660647,
+      "loss": 0.9387,
+      "step": 21165
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017577566597781093,
+      "loss": 1.0792,
+      "step": 21170
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017575932654931035,
+      "loss": 1.1199,
+      "step": 21175
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00017574298237212887,
+      "loss": 0.913,
+      "step": 21180
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017572663344729103,
+      "loss": 0.9306,
+      "step": 21185
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017571027977582158,
+      "loss": 0.9791,
+      "step": 21190
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017569392135874552,
+      "loss": 0.9209,
+      "step": 21195
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017567755819708828,
+      "loss": 1.0335,
+      "step": 21200
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001756611902918755,
+      "loss": 0.9791,
+      "step": 21205
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00017564481764413312,
+      "loss": 0.9551,
+      "step": 21210
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017562844025488747,
+      "loss": 0.9948,
+      "step": 21215
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.000175612058125165,
+      "loss": 0.9004,
+      "step": 21220
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00017559567125599265,
+      "loss": 0.9636,
+      "step": 21225
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001755792796483975,
+      "loss": 0.9681,
+      "step": 21230
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00017556288330340705,
+      "loss": 0.9337,
+      "step": 21235
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.000175546482222049,
+      "loss": 0.8969,
+      "step": 21240
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001755300764053514,
+      "loss": 0.8336,
+      "step": 21245
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017551366585434258,
+      "loss": 0.9793,
+      "step": 21250
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00017549725057005122,
+      "loss": 0.9903,
+      "step": 21255
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001754808305535062,
+      "loss": 0.8948,
+      "step": 21260
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017546440580573674,
+      "loss": 1.1077,
+      "step": 21265
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017544797632777236,
+      "loss": 0.9889,
+      "step": 21270
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001754315421206429,
+      "loss": 0.999,
+      "step": 21275
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001754151031853785,
+      "loss": 1.0356,
+      "step": 21280
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017539865952300953,
+      "loss": 0.9596,
+      "step": 21285
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001753822111345667,
+      "loss": 1.0203,
+      "step": 21290
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017536575802108102,
+      "loss": 1.1007,
+      "step": 21295
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017534930018358382,
+      "loss": 0.9677,
+      "step": 21300
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00017533283762310667,
+      "loss": 0.8472,
+      "step": 21305
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0,
+      "learning_rate": 0.00017531637034068142,
+      "loss": 1.0809,
+      "step": 21310
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001752998983373403,
+      "loss": 1.0434,
+      "step": 21315
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017528342161411583,
+      "loss": 1.0721,
+      "step": 21320
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017526694017204074,
+      "loss": 0.9382,
+      "step": 21325
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0001752504540121481,
+      "loss": 0.9678,
+      "step": 21330
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00017523396313547134,
+      "loss": 0.8703,
+      "step": 21335
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017521746754304407,
+      "loss": 0.9302,
+      "step": 21340
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00017520096723590024,
+      "loss": 1.0109,
+      "step": 21345
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017518446221507415,
+      "loss": 0.9523,
+      "step": 21350
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00017516795248160037,
+      "loss": 0.9953,
+      "step": 21355
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017515143803651368,
+      "loss": 0.9026,
+      "step": 21360
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017513491888084928,
+      "loss": 0.8433,
+      "step": 21365
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001751183950156426,
+      "loss": 1.0761,
+      "step": 21370
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017510186644192937,
+      "loss": 0.9921,
+      "step": 21375
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001750853331607456,
+      "loss": 0.9785,
+      "step": 21380
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017506879517312763,
+      "loss": 0.8458,
+      "step": 21385
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001750522524801121,
+      "loss": 1.0463,
+      "step": 21390
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001750357050827359,
+      "loss": 0.8554,
+      "step": 21395
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017501915298203624,
+      "loss": 0.9932,
+      "step": 21400
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00017500259617905063,
+      "loss": 0.8362,
+      "step": 21405
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017498603467481689,
+      "loss": 0.9355,
+      "step": 21410
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001749694684703731,
+      "loss": 0.9487,
+      "step": 21415
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001749528975667576,
+      "loss": 0.7915,
+      "step": 21420
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017493632196500914,
+      "loss": 1.0153,
+      "step": 21425
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001749197416661667,
+      "loss": 1.1029,
+      "step": 21430
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001749031566712695,
+      "loss": 0.8185,
+      "step": 21435
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017488656698135712,
+      "loss": 0.8799,
+      "step": 21440
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017486997259746946,
+      "loss": 0.9096,
+      "step": 21445
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00017485337352064666,
+      "loss": 1.0953,
+      "step": 21450
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00017483676975192913,
+      "loss": 0.8612,
+      "step": 21455
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017482016129235763,
+      "loss": 1.1556,
+      "step": 21460
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00017480354814297322,
+      "loss": 0.9108,
+      "step": 21465
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00017478693030481722,
+      "loss": 0.9021,
+      "step": 21470
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00017477030777893123,
+      "loss": 0.9815,
+      "step": 21475
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017475368056635724,
+      "loss": 1.0603,
+      "step": 21480
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00017473704866813737,
+      "loss": 0.9242,
+      "step": 21485
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017472041208531423,
+      "loss": 0.8487,
+      "step": 21490
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017470377081893048,
+      "loss": 1.0269,
+      "step": 21495
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017468712487002933,
+      "loss": 0.9441,
+      "step": 21500
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00017467047423965415,
+      "loss": 1.1466,
+      "step": 21505
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001746538189288486,
+      "loss": 1.0247,
+      "step": 21510
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017463715893865664,
+      "loss": 1.0308,
+      "step": 21515
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017462049427012256,
+      "loss": 0.9886,
+      "step": 21520
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001746038249242909,
+      "loss": 0.914,
+      "step": 21525
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00017458715090220656,
+      "loss": 0.9864,
+      "step": 21530
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017457047220491463,
+      "loss": 1.0514,
+      "step": 21535
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00017455378883346056,
+      "loss": 1.0187,
+      "step": 21540
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00017453710078889012,
+      "loss": 0.9507,
+      "step": 21545
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001745204080722493,
+      "loss": 1.0244,
+      "step": 21550
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017450371068458446,
+      "loss": 0.9657,
+      "step": 21555
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00017448700862694215,
+      "loss": 1.0195,
+      "step": 21560
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017447030190036932,
+      "loss": 0.9239,
+      "step": 21565
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00017445359050591313,
+      "loss": 0.9354,
+      "step": 21570
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.0001744368744446211,
+      "loss": 1.0583,
+      "step": 21575
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017442015371754103,
+      "loss": 0.9223,
+      "step": 21580
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00017440342832572095,
+      "loss": 1.046,
+      "step": 21585
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017438669827020924,
+      "loss": 1.1004,
+      "step": 21590
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017436996355205456,
+      "loss": 0.8468,
+      "step": 21595
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017435322417230586,
+      "loss": 1.0398,
+      "step": 21600
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001743364801320124,
+      "loss": 1.0182,
+      "step": 21605
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00017431973143222372,
+      "loss": 1.0255,
+      "step": 21610
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001743029780739896,
+      "loss": 0.8777,
+      "step": 21615
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00017428622005836018,
+      "loss": 0.9486,
+      "step": 21620
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001742694573863859,
+      "loss": 0.9823,
+      "step": 21625
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017425269005911744,
+      "loss": 0.9608,
+      "step": 21630
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00017423591807760582,
+      "loss": 0.9378,
+      "step": 21635
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001742191414429023,
+      "loss": 0.8654,
+      "step": 21640
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017420236015605847,
+      "loss": 0.9431,
+      "step": 21645
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0001741855742181262,
+      "loss": 0.9499,
+      "step": 21650
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017416878363015763,
+      "loss": 1.0357,
+      "step": 21655
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00017415198839320525,
+      "loss": 0.9856,
+      "step": 21660
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00017413518850832173,
+      "loss": 0.992,
+      "step": 21665
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017411838397656023,
+      "loss": 0.9234,
+      "step": 21670
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00017410157479897398,
+      "loss": 0.9881,
+      "step": 21675
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017408476097661662,
+      "loss": 1.0233,
+      "step": 21680
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017406794251054208,
+      "loss": 1.0662,
+      "step": 21685
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017405111940180454,
+      "loss": 0.9822,
+      "step": 21690
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001740342916514585,
+      "loss": 0.9748,
+      "step": 21695
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00017401745926055875,
+      "loss": 0.7273,
+      "step": 21700
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017400062223016033,
+      "loss": 0.8134,
+      "step": 21705
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00017398378056131866,
+      "loss": 1.0092,
+      "step": 21710
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017396693425508934,
+      "loss": 1.0338,
+      "step": 21715
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017395008331252834,
+      "loss": 0.9591,
+      "step": 21720
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017393322773469192,
+      "loss": 0.9183,
+      "step": 21725
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017391636752263657,
+      "loss": 1.0216,
+      "step": 21730
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001738995026774191,
+      "loss": 0.9468,
+      "step": 21735
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017388263320009667,
+      "loss": 0.9883,
+      "step": 21740
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001738657590917266,
+      "loss": 0.9116,
+      "step": 21745
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001738488803533667,
+      "loss": 1.0142,
+      "step": 21750
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017383199698607483,
+      "loss": 1.0342,
+      "step": 21755
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001738151089909093,
+      "loss": 0.9699,
+      "step": 21760
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017379821636892865,
+      "loss": 1.0844,
+      "step": 21765
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001737813191211918,
+      "loss": 1.0149,
+      "step": 21770
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001737644172487578,
+      "loss": 0.9203,
+      "step": 21775
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001737475107526861,
+      "loss": 0.928,
+      "step": 21780
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017373059963403647,
+      "loss": 1.048,
+      "step": 21785
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017371368389386887,
+      "loss": 0.955,
+      "step": 21790
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001736967635332436,
+      "loss": 0.9854,
+      "step": 21795
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001736798385532213,
+      "loss": 0.8858,
+      "step": 21800
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017366290895486276,
+      "loss": 0.9952,
+      "step": 21805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001736459747392292,
+      "loss": 0.9809,
+      "step": 21810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017362903590738204,
+      "loss": 0.9138,
+      "step": 21815
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001736120924603831,
+      "loss": 1.1694,
+      "step": 21820
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00017359514439929435,
+      "loss": 0.9206,
+      "step": 21825
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00017357819172517816,
+      "loss": 0.7781,
+      "step": 21830
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017356123443909707,
+      "loss": 0.9021,
+      "step": 21835
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017354427254211404,
+      "loss": 1.0017,
+      "step": 21840
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00017352730603529225,
+      "loss": 0.8842,
+      "step": 21845
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001735103349196952,
+      "loss": 0.9533,
+      "step": 21850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001734933591963866,
+      "loss": 1.0208,
+      "step": 21855
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00017347637886643057,
+      "loss": 0.8829,
+      "step": 21860
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001734593939308914,
+      "loss": 1.0942,
+      "step": 21865
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00017344240439083378,
+      "loss": 0.9608,
+      "step": 21870
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001734254102473226,
+      "loss": 0.916,
+      "step": 21875
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017340841150142308,
+      "loss": 1.1468,
+      "step": 21880
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00017339140815420074,
+      "loss": 0.8373,
+      "step": 21885
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00017337440020672134,
+      "loss": 1.135,
+      "step": 21890
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017335738766005097,
+      "loss": 1.0092,
+      "step": 21895
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.000173340370515256,
+      "loss": 0.9375,
+      "step": 21900
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00017332334877340303,
+      "loss": 0.9806,
+      "step": 21905
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001733063224355591,
+      "loss": 1.0827,
+      "step": 21910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017328929150279142,
+      "loss": 1.0186,
+      "step": 21915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001732722559761674,
+      "loss": 1.0553,
+      "step": 21920
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.000173255215856755,
+      "loss": 0.921,
+      "step": 21925
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00017323817114562218,
+      "loss": 0.9589,
+      "step": 21930
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017322112184383742,
+      "loss": 0.9213,
+      "step": 21935
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017320406795246933,
+      "loss": 0.9863,
+      "step": 21940
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017318700947258688,
+      "loss": 0.8876,
+      "step": 21945
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017316994640525935,
+      "loss": 0.7832,
+      "step": 21950
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00017315287875155623,
+      "loss": 0.9013,
+      "step": 21955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017313580651254738,
+      "loss": 1.0958,
+      "step": 21960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017311872968930281,
+      "loss": 1.1191,
+      "step": 21965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017310164828289305,
+      "loss": 1.0535,
+      "step": 21970
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017308456229438873,
+      "loss": 1.0017,
+      "step": 21975
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00017306747172486078,
+      "loss": 1.079,
+      "step": 21980
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001730503765753805,
+      "loss": 0.9727,
+      "step": 21985
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001730332768470194,
+      "loss": 1.0199,
+      "step": 21990
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00017301617254084938,
+      "loss": 0.936,
+      "step": 21995
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017299906365794246,
+      "loss": 1.0371,
+      "step": 22000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001729819501993711,
+      "loss": 0.9254,
+      "step": 22005
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.000172964832166208,
+      "loss": 1.041,
+      "step": 22010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00017294770955952608,
+      "loss": 0.9152,
+      "step": 22015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017293058238039867,
+      "loss": 0.9246,
+      "step": 22020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017291345062989927,
+      "loss": 0.7779,
+      "step": 22025
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017289631430910177,
+      "loss": 1.0301,
+      "step": 22030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00017287917341908025,
+      "loss": 1.1244,
+      "step": 22035
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00017286202796090917,
+      "loss": 0.7826,
+      "step": 22040
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017284487793566317,
+      "loss": 0.8691,
+      "step": 22045
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00017282772334441729,
+      "loss": 1.0271,
+      "step": 22050
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0001728105641882467,
+      "loss": 0.7768,
+      "step": 22055
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00017279340046822707,
+      "loss": 0.9809,
+      "step": 22060
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017277623218543418,
+      "loss": 0.8828,
+      "step": 22065
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001727590593409442,
+      "loss": 1.0735,
+      "step": 22070
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017274188193583346,
+      "loss": 0.7952,
+      "step": 22075
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00017272469997117878,
+      "loss": 1.0377,
+      "step": 22080
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00017270751344805702,
+      "loss": 0.9094,
+      "step": 22085
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00017269032236754556,
+      "loss": 0.9699,
+      "step": 22090
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00017267312673072187,
+      "loss": 1.0409,
+      "step": 22095
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017265592653866385,
+      "loss": 0.8788,
+      "step": 22100
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00017263872179244956,
+      "loss": 0.947,
+      "step": 22105
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00017262151249315753,
+      "loss": 0.9475,
+      "step": 22110
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017260429864186634,
+      "loss": 0.9822,
+      "step": 22115
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00017258708023965504,
+      "loss": 0.9842,
+      "step": 22120
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017256985728760288,
+      "loss": 0.8861,
+      "step": 22125
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017255262978678942,
+      "loss": 0.9062,
+      "step": 22130
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00017253539773829448,
+      "loss": 0.9959,
+      "step": 22135
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001725181611431982,
+      "loss": 1.0269,
+      "step": 22140
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.000172500920002581,
+      "loss": 1.0099,
+      "step": 22145
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017248367431752355,
+      "loss": 1.0521,
+      "step": 22150
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017246642408910685,
+      "loss": 0.9477,
+      "step": 22155
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017244916931841216,
+      "loss": 1.0187,
+      "step": 22160
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.000172431910006521,
+      "loss": 0.9319,
+      "step": 22165
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017241464615451525,
+      "loss": 0.9638,
+      "step": 22170
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.000172397377763477,
+      "loss": 1.0252,
+      "step": 22175
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00017238010483448866,
+      "loss": 1.0172,
+      "step": 22180
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00017236282736863293,
+      "loss": 0.9769,
+      "step": 22185
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00017234554536699274,
+      "loss": 1.0083,
+      "step": 22190
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001723282588306514,
+      "loss": 1.0638,
+      "step": 22195
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001723109677606924,
+      "loss": 0.9845,
+      "step": 22200
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017229367215819958,
+      "loss": 0.9822,
+      "step": 22205
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00017227637202425706,
+      "loss": 1.1154,
+      "step": 22210
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017225906735994923,
+      "loss": 0.9841,
+      "step": 22215
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00017224175816636075,
+      "loss": 0.9897,
+      "step": 22220
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017222444444457664,
+      "loss": 0.9637,
+      "step": 22225
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00017220712619568204,
+      "loss": 1.0782,
+      "step": 22230
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00017218980342076254,
+      "loss": 1.1999,
+      "step": 22235
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017217247612090396,
+      "loss": 0.9888,
+      "step": 22240
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017215514429719237,
+      "loss": 0.8246,
+      "step": 22245
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017213780795071417,
+      "loss": 0.8804,
+      "step": 22250
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.000172120467082556,
+      "loss": 1.2292,
+      "step": 22255
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001721031216938048,
+      "loss": 0.9579,
+      "step": 22260
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017208577178554787,
+      "loss": 0.9219,
+      "step": 22265
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001720684173588726,
+      "loss": 0.9643,
+      "step": 22270
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00017205105841486688,
+      "loss": 0.8721,
+      "step": 22275
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00017203369495461877,
+      "loss": 0.9526,
+      "step": 22280
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017201632697921659,
+      "loss": 0.9051,
+      "step": 22285
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017199895448974904,
+      "loss": 1.0717,
+      "step": 22290
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00017198157748730502,
+      "loss": 0.9993,
+      "step": 22295
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00017196419597297372,
+      "loss": 1.0139,
+      "step": 22300
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00017194680994784468,
+      "loss": 0.8819,
+      "step": 22305
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017192941941300763,
+      "loss": 1.0517,
+      "step": 22310
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017191202436955268,
+      "loss": 0.9249,
+      "step": 22315
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00017189462481857014,
+      "loss": 1.0895,
+      "step": 22320
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017187722076115062,
+      "loss": 0.9648,
+      "step": 22325
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00017185981219838503,
+      "loss": 0.8843,
+      "step": 22330
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017184239913136458,
+      "loss": 1.0308,
+      "step": 22335
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00017182498156118075,
+      "loss": 0.969,
+      "step": 22340
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017180755948892524,
+      "loss": 0.8646,
+      "step": 22345
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017179013291569018,
+      "loss": 0.9084,
+      "step": 22350
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00017177270184256775,
+      "loss": 1.0191,
+      "step": 22355
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00017175526627065065,
+      "loss": 1.1,
+      "step": 22360
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017173782620103176,
+      "loss": 1.0031,
+      "step": 22365
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001717203816348042,
+      "loss": 1.0448,
+      "step": 22370
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00017170293257306148,
+      "loss": 0.8559,
+      "step": 22375
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017168547901689723,
+      "loss": 0.9636,
+      "step": 22380
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017166802096740553,
+      "loss": 1.0384,
+      "step": 22385
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017165055842568067,
+      "loss": 0.9458,
+      "step": 22390
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001716330913928172,
+      "loss": 1.0262,
+      "step": 22395
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017161561986990995,
+      "loss": 1.0135,
+      "step": 22400
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001715981438580541,
+      "loss": 1.0603,
+      "step": 22405
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00017158066335834507,
+      "loss": 1.0851,
+      "step": 22410
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001715631783718785,
+      "loss": 0.8572,
+      "step": 22415
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00017154568889975042,
+      "loss": 0.9603,
+      "step": 22420
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001715281949430571,
+      "loss": 1.0527,
+      "step": 22425
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.000171510696502895,
+      "loss": 0.8258,
+      "step": 22430
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017149319358036107,
+      "loss": 1.0781,
+      "step": 22435
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001714756861765523,
+      "loss": 1.0145,
+      "step": 22440
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017145817429256612,
+      "loss": 0.8742,
+      "step": 22445
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.0001714406579295002,
+      "loss": 0.9866,
+      "step": 22450
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001714231370884525,
+      "loss": 0.8931,
+      "step": 22455
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00017140561177052117,
+      "loss": 0.9762,
+      "step": 22460
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001713880819768048,
+      "loss": 1.0102,
+      "step": 22465
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017137054770840213,
+      "loss": 1.1021,
+      "step": 22470
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00017135300896641229,
+      "loss": 1.1629,
+      "step": 22475
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00017133546575193452,
+      "loss": 1.0134,
+      "step": 22480
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017131791806606857,
+      "loss": 1.0137,
+      "step": 22485
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00017130036590991426,
+      "loss": 0.9579,
+      "step": 22490
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00017128280928457182,
+      "loss": 0.8749,
+      "step": 22495
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001712652481911417,
+      "loss": 1.0704,
+      "step": 22500
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00017124768263072467,
+      "loss": 1.0315,
+      "step": 22505
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00017123011260442174,
+      "loss": 1.0279,
+      "step": 22510
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00017121253811333423,
+      "loss": 0.9216,
+      "step": 22515
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017119495915856375,
+      "loss": 1.0082,
+      "step": 22520
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001711773757412121,
+      "loss": 0.8658,
+      "step": 22525
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017115978786238153,
+      "loss": 1.0071,
+      "step": 22530
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00017114219552317436,
+      "loss": 0.9824,
+      "step": 22535
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00017112459872469337,
+      "loss": 1.0176,
+      "step": 22540
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00017110699746804154,
+      "loss": 0.9751,
+      "step": 22545
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001710893917543221,
+      "loss": 0.8675,
+      "step": 22550
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00017107178158463863,
+      "loss": 0.9408,
+      "step": 22555
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017105416696009497,
+      "loss": 0.9809,
+      "step": 22560
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001710365478817952,
+      "loss": 0.9696,
+      "step": 22565
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001710189243508437,
+      "loss": 0.9519,
+      "step": 22570
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001710012963683451,
+      "loss": 1.0546,
+      "step": 22575
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017098366393540442,
+      "loss": 0.9163,
+      "step": 22580
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017096602705312682,
+      "loss": 1.0133,
+      "step": 22585
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017094838572261783,
+      "loss": 0.9241,
+      "step": 22590
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017093073994498318,
+      "loss": 0.9855,
+      "step": 22595
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00017091308972132905,
+      "loss": 0.9835,
+      "step": 22600
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017089543505276162,
+      "loss": 1.0074,
+      "step": 22605
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001708777759403876,
+      "loss": 0.8852,
+      "step": 22610
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00017086011238531386,
+      "loss": 0.9527,
+      "step": 22615
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001708424443886476,
+      "loss": 0.9278,
+      "step": 22620
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00017082477195149622,
+      "loss": 0.9386,
+      "step": 22625
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001708070950749675,
+      "loss": 0.8705,
+      "step": 22630
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00017078941376016938,
+      "loss": 1.0218,
+      "step": 22635
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00017077172800821018,
+      "loss": 1.1595,
+      "step": 22640
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00017075403782019848,
+      "loss": 1.0909,
+      "step": 22645
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00017073634319724309,
+      "loss": 0.8685,
+      "step": 22650
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00017071864414045318,
+      "loss": 0.9616,
+      "step": 22655
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001707009406509381,
+      "loss": 0.9912,
+      "step": 22660
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00017068323272980752,
+      "loss": 0.9322,
+      "step": 22665
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00017066552037817143,
+      "loss": 1.0414,
+      "step": 22670
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017064780359714004,
+      "loss": 0.9212,
+      "step": 22675
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00017063008238782387,
+      "loss": 1.0547,
+      "step": 22680
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017061235675133366,
+      "loss": 0.8523,
+      "step": 22685
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00017059462668878055,
+      "loss": 0.9595,
+      "step": 22690
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001705768922012758,
+      "loss": 0.9903,
+      "step": 22695
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001705591532899311,
+      "loss": 1.0138,
+      "step": 22700
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001705414099558583,
+      "loss": 0.9734,
+      "step": 22705
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00017052366220016957,
+      "loss": 1.0777,
+      "step": 22710
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001705059100239774,
+      "loss": 1.0454,
+      "step": 22715
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00017048815342839447,
+      "loss": 1.1218,
+      "step": 22720
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00017047039241453382,
+      "loss": 0.9762,
+      "step": 22725
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001704526269835087,
+      "loss": 0.9931,
+      "step": 22730
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001704348571364327,
+      "loss": 0.919,
+      "step": 22735
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017041708287441959,
+      "loss": 0.9993,
+      "step": 22740
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001703993041985836,
+      "loss": 1.067,
+      "step": 22745
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017038152111003898,
+      "loss": 0.9398,
+      "step": 22750
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00017036373360990048,
+      "loss": 1.1059,
+      "step": 22755
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00017034594169928303,
+      "loss": 1.0214,
+      "step": 22760
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00017032814537930183,
+      "loss": 0.9617,
+      "step": 22765
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00017031034465107237,
+      "loss": 0.8269,
+      "step": 22770
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00017029253951571046,
+      "loss": 0.8859,
+      "step": 22775
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00017027472997433208,
+      "loss": 0.9395,
+      "step": 22780
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001702569160280536,
+      "loss": 0.9338,
+      "step": 22785
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00017023909767799163,
+      "loss": 0.9962,
+      "step": 22790
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.000170221274925263,
+      "loss": 0.8451,
+      "step": 22795
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00017020344777098488,
+      "loss": 1.0406,
+      "step": 22800
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001701856162162747,
+      "loss": 1.0097,
+      "step": 22805
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001701677802622502,
+      "loss": 0.9807,
+      "step": 22810
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00017014993991002926,
+      "loss": 0.886,
+      "step": 22815
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00017013209516073024,
+      "loss": 1.0094,
+      "step": 22820
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017011424601547158,
+      "loss": 1.0373,
+      "step": 22825
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00017009639247537214,
+      "loss": 0.9939,
+      "step": 22830
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00017007853454155102,
+      "loss": 1.0919,
+      "step": 22835
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00017006067221512748,
+      "loss": 0.9612,
+      "step": 22840
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00017004280549722127,
+      "loss": 0.9532,
+      "step": 22845
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001700249343889522,
+      "loss": 0.8965,
+      "step": 22850
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001700070588914405,
+      "loss": 1.0316,
+      "step": 22855
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016998917900580665,
+      "loss": 1.1005,
+      "step": 22860
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00016997129473317132,
+      "loss": 0.9847,
+      "step": 22865
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00016995340607465554,
+      "loss": 0.8212,
+      "step": 22870
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016993551303138063,
+      "loss": 0.9523,
+      "step": 22875
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00016991761560446807,
+      "loss": 0.942,
+      "step": 22880
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016989971379503978,
+      "loss": 0.9276,
+      "step": 22885
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001698818076042178,
+      "loss": 0.9119,
+      "step": 22890
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001698638970331245,
+      "loss": 0.938,
+      "step": 22895
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001698459820828826,
+      "loss": 0.9448,
+      "step": 22900
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00016982806275461497,
+      "loss": 1.0021,
+      "step": 22905
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016981013904944487,
+      "loss": 1.0421,
+      "step": 22910
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016979221096849573,
+      "loss": 1.02,
+      "step": 22915
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00016977427851289133,
+      "loss": 1.1506,
+      "step": 22920
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016975634168375566,
+      "loss": 0.867,
+      "step": 22925
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00016973840048221307,
+      "loss": 0.999,
+      "step": 22930
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016972045490938812,
+      "loss": 0.9237,
+      "step": 22935
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00016970250496640564,
+      "loss": 1.0466,
+      "step": 22940
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00016968455065439076,
+      "loss": 0.832,
+      "step": 22945
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00016966659197446889,
+      "loss": 0.9661,
+      "step": 22950
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001696486289277657,
+      "loss": 0.8908,
+      "step": 22955
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001696306615154071,
+      "loss": 0.9853,
+      "step": 22960
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00016961268973851937,
+      "loss": 1.0202,
+      "step": 22965
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016959471359822895,
+      "loss": 0.8762,
+      "step": 22970
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00016957673309566258,
+      "loss": 1.0241,
+      "step": 22975
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016955874823194737,
+      "loss": 1.0761,
+      "step": 22980
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001695407590082106,
+      "loss": 1.0919,
+      "step": 22985
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00016952276542557985,
+      "loss": 1.0389,
+      "step": 22990
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.000169504767485183,
+      "loss": 1.0269,
+      "step": 22995
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016948676518814816,
+      "loss": 0.9425,
+      "step": 23000
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001694687585356037,
+      "loss": 1.1517,
+      "step": 23005
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001694507475286784,
+      "loss": 0.9304,
+      "step": 23010
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001694327321685011,
+      "loss": 0.8311,
+      "step": 23015
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001694147124562011,
+      "loss": 0.9417,
+      "step": 23020
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00016939668839290785,
+      "loss": 0.871,
+      "step": 23025
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00016937865997975116,
+      "loss": 0.9416,
+      "step": 23030
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.000169360627217861,
+      "loss": 0.9172,
+      "step": 23035
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00016934259010836775,
+      "loss": 1.0004,
+      "step": 23040
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.000169324548652402,
+      "loss": 1.0068,
+      "step": 23045
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00016930650285109454,
+      "loss": 0.9082,
+      "step": 23050
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001692884527055766,
+      "loss": 0.9053,
+      "step": 23055
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001692703982169795,
+      "loss": 1.0665,
+      "step": 23060
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016925233938643497,
+      "loss": 1.0995,
+      "step": 23065
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00016923427621507491,
+      "loss": 0.954,
+      "step": 23070
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001692162087040316,
+      "loss": 1.0213,
+      "step": 23075
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00016919813685443744,
+      "loss": 0.9033,
+      "step": 23080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001691800606674253,
+      "loss": 1.0971,
+      "step": 23085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016916198014412816,
+      "loss": 1.0208,
+      "step": 23090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016914389528567932,
+      "loss": 0.8997,
+      "step": 23095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001691258060932124,
+      "loss": 0.8185,
+      "step": 23100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001691077125678612,
+      "loss": 0.9328,
+      "step": 23105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001690896147107599,
+      "loss": 0.8601,
+      "step": 23110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016907151252304283,
+      "loss": 0.9554,
+      "step": 23115
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001690534060058447,
+      "loss": 0.931,
+      "step": 23120
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016903529516030044,
+      "loss": 0.8989,
+      "step": 23125
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016901717998754528,
+      "loss": 1.0203,
+      "step": 23130
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00016899906048871462,
+      "loss": 1.0554,
+      "step": 23135
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001689809366649443,
+      "loss": 1.0396,
+      "step": 23140
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001689628085173703,
+      "loss": 0.9693,
+      "step": 23145
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016894467604712892,
+      "loss": 0.8956,
+      "step": 23150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016892653925535672,
+      "loss": 0.9386,
+      "step": 23155
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00016890839814319057,
+      "loss": 1.0394,
+      "step": 23160
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001688902527117675,
+      "loss": 0.9978,
+      "step": 23165
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00016887210296222496,
+      "loss": 0.8688,
+      "step": 23170
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00016885394889570057,
+      "loss": 1.0429,
+      "step": 23175
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001688357905133322,
+      "loss": 1.021,
+      "step": 23180
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00016881762781625813,
+      "loss": 0.8929,
+      "step": 23185
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016879946080561675,
+      "loss": 1.0132,
+      "step": 23190
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001687812894825468,
+      "loss": 1.0581,
+      "step": 23195
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016876311384818733,
+      "loss": 0.9398,
+      "step": 23200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016874493390367756,
+      "loss": 1.0428,
+      "step": 23205
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.000168726749650157,
+      "loss": 0.972,
+      "step": 23210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00016870856108876554,
+      "loss": 0.9241,
+      "step": 23215
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00016869036822064323,
+      "loss": 0.9303,
+      "step": 23220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016867217104693036,
+      "loss": 0.9909,
+      "step": 23225
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001686539695687676,
+      "loss": 0.9815,
+      "step": 23230
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016863576378729588,
+      "loss": 1.0365,
+      "step": 23235
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016861755370365628,
+      "loss": 0.8689,
+      "step": 23240
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016859933931899027,
+      "loss": 0.9151,
+      "step": 23245
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00016858112063443956,
+      "loss": 0.9882,
+      "step": 23250
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001685628976511461,
+      "loss": 1.182,
+      "step": 23255
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016854467037025214,
+      "loss": 1.0126,
+      "step": 23260
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016852643879290016,
+      "loss": 0.9376,
+      "step": 23265
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00016850820292023294,
+      "loss": 0.9557,
+      "step": 23270
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016848996275339352,
+      "loss": 0.9623,
+      "step": 23275
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00016847171829352528,
+      "loss": 1.0208,
+      "step": 23280
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016845346954177172,
+      "loss": 0.9661,
+      "step": 23285
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016843521649927674,
+      "loss": 1.0168,
+      "step": 23290
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016841695916718443,
+      "loss": 0.9827,
+      "step": 23295
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001683986975466392,
+      "loss": 1.0155,
+      "step": 23300
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016838043163878573,
+      "loss": 0.9813,
+      "step": 23305
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00016836216144476893,
+      "loss": 0.8198,
+      "step": 23310
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.000168343886965734,
+      "loss": 1.0811,
+      "step": 23315
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00016832560820282636,
+      "loss": 0.9135,
+      "step": 23320
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001683073251571918,
+      "loss": 0.9304,
+      "step": 23325
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001682890378299763,
+      "loss": 1.0073,
+      "step": 23330
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00016827074622232616,
+      "loss": 0.99,
+      "step": 23335
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00016825245033538785,
+      "loss": 0.8615,
+      "step": 23340
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016823415017030825,
+      "loss": 0.8889,
+      "step": 23345
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016821584572823442,
+      "loss": 0.982,
+      "step": 23350
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016819753701031363,
+      "loss": 1.0136,
+      "step": 23355
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016817922401769363,
+      "loss": 1.0189,
+      "step": 23360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016816090675152214,
+      "loss": 1.0293,
+      "step": 23365
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00016814258521294744,
+      "loss": 0.9796,
+      "step": 23370
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016812425940311787,
+      "loss": 0.9483,
+      "step": 23375
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016810592932318212,
+      "loss": 0.926,
+      "step": 23380
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001680875949742892,
+      "loss": 0.907,
+      "step": 23385
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001680692563575882,
+      "loss": 0.9811,
+      "step": 23390
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016805091347422876,
+      "loss": 1.0047,
+      "step": 23395
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016803256632536053,
+      "loss": 0.8818,
+      "step": 23400
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00016801421491213358,
+      "loss": 1.011,
+      "step": 23405
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016799585923569816,
+      "loss": 0.9497,
+      "step": 23410
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016797749929720485,
+      "loss": 1.0023,
+      "step": 23415
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00016795913509780447,
+      "loss": 0.8086,
+      "step": 23420
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001679407666386481,
+      "loss": 0.9474,
+      "step": 23425
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00016792239392088708,
+      "loss": 1.0872,
+      "step": 23430
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00016790401694567305,
+      "loss": 0.9645,
+      "step": 23435
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00016788563571415793,
+      "loss": 0.9173,
+      "step": 23440
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00016786725022749382,
+      "loss": 0.8202,
+      "step": 23445
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00016784886048683322,
+      "loss": 0.9017,
+      "step": 23450
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00016783046649332872,
+      "loss": 1.0489,
+      "step": 23455
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016781206824813337,
+      "loss": 0.923,
+      "step": 23460
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00016779366575240032,
+      "loss": 0.9346,
+      "step": 23465
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001677752590072831,
+      "loss": 0.9595,
+      "step": 23470
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00016775684801393546,
+      "loss": 0.8984,
+      "step": 23475
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00016773843277351138,
+      "loss": 1.0412,
+      "step": 23480
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00016772001328716523,
+      "loss": 0.9927,
+      "step": 23485
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016770158955605152,
+      "loss": 0.8891,
+      "step": 23490
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00016768316158132505,
+      "loss": 1.0668,
+      "step": 23495
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016766472936414093,
+      "loss": 0.9172,
+      "step": 23500
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001676462929056545,
+      "loss": 0.9436,
+      "step": 23505
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016762785220702142,
+      "loss": 1.0329,
+      "step": 23510
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001676094072693975,
+      "loss": 0.862,
+      "step": 23515
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.000167590958093939,
+      "loss": 1.1708,
+      "step": 23520
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001675725046818022,
+      "loss": 1.0274,
+      "step": 23525
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016755404703414388,
+      "loss": 0.9976,
+      "step": 23530
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016753558515212095,
+      "loss": 1.0193,
+      "step": 23535
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016751711903689062,
+      "loss": 1.018,
+      "step": 23540
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016749864868961038,
+      "loss": 1.0502,
+      "step": 23545
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00016748017411143798,
+      "loss": 1.0112,
+      "step": 23550
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00016746169530353137,
+      "loss": 0.9149,
+      "step": 23555
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016744321226704888,
+      "loss": 1.0269,
+      "step": 23560
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00016742472500314904,
+      "loss": 0.8589,
+      "step": 23565
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00016740623351299067,
+      "loss": 0.9598,
+      "step": 23570
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00016738773779773278,
+      "loss": 1.0552,
+      "step": 23575
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016736923785853476,
+      "loss": 1.0401,
+      "step": 23580
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00016735073369655615,
+      "loss": 1.0078,
+      "step": 23585
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001673322253129569,
+      "loss": 0.9583,
+      "step": 23590
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016731371270889707,
+      "loss": 0.9447,
+      "step": 23595
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016729519588553704,
+      "loss": 0.8466,
+      "step": 23600
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016727667484403748,
+      "loss": 0.9608,
+      "step": 23605
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016725814958555932,
+      "loss": 0.9509,
+      "step": 23610
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016723962011126376,
+      "loss": 0.9129,
+      "step": 23615
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016722108642231224,
+      "loss": 0.9439,
+      "step": 23620
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00016720254851986647,
+      "loss": 1.2049,
+      "step": 23625
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001671840064050884,
+      "loss": 1.0429,
+      "step": 23630
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001671654600791403,
+      "loss": 0.9667,
+      "step": 23635
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016714690954318465,
+      "loss": 0.9415,
+      "step": 23640
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016712835479838428,
+      "loss": 0.9917,
+      "step": 23645
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00016710979584590215,
+      "loss": 0.9524,
+      "step": 23650
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016709123268690158,
+      "loss": 1.0102,
+      "step": 23655
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016707266532254615,
+      "loss": 1.0184,
+      "step": 23660
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016705409375399963,
+      "loss": 0.9577,
+      "step": 23665
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00016703551798242621,
+      "loss": 0.9503,
+      "step": 23670
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00016701693800899014,
+      "loss": 1.2045,
+      "step": 23675
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00016699835383485604,
+      "loss": 1.0171,
+      "step": 23680
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016697976546118886,
+      "loss": 0.859,
+      "step": 23685
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00016696117288915368,
+      "loss": 0.9017,
+      "step": 23690
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016694257611991594,
+      "loss": 1.0393,
+      "step": 23695
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016692397515464125,
+      "loss": 0.9315,
+      "step": 23700
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00016690536999449561,
+      "loss": 0.933,
+      "step": 23705
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00016688676064064516,
+      "loss": 0.8206,
+      "step": 23710
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001668681470942564,
+      "loss": 0.9649,
+      "step": 23715
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.000166849529356496,
+      "loss": 0.939,
+      "step": 23720
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016683090742853097,
+      "loss": 0.9569,
+      "step": 23725
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016681228131152856,
+      "loss": 0.9348,
+      "step": 23730
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016679365100665626,
+      "loss": 1.0092,
+      "step": 23735
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016677501651508184,
+      "loss": 0.9932,
+      "step": 23740
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001667563778379733,
+      "loss": 0.9631,
+      "step": 23745
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.000166737734976499,
+      "loss": 1.0865,
+      "step": 23750
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001667190879318275,
+      "loss": 1.0609,
+      "step": 23755
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016670043670512753,
+      "loss": 1.0621,
+      "step": 23760
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00016668178129756824,
+      "loss": 1.0944,
+      "step": 23765
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00016666312171031896,
+      "loss": 0.8663,
+      "step": 23770
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016664445794454928,
+      "loss": 1.04,
+      "step": 23775
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00016662579000142907,
+      "loss": 0.93,
+      "step": 23780
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016660711788212847,
+      "loss": 0.9787,
+      "step": 23785
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001665884415878179,
+      "loss": 0.9356,
+      "step": 23790
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001665697611196679,
+      "loss": 0.9909,
+      "step": 23795
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016655107647884946,
+      "loss": 0.9665,
+      "step": 23800
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001665323876665338,
+      "loss": 0.774,
+      "step": 23805
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00016651369468389228,
+      "loss": 1.0398,
+      "step": 23810
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016649499753209666,
+      "loss": 1.0225,
+      "step": 23815
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00016647629621231882,
+      "loss": 0.8501,
+      "step": 23820
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016645759072573104,
+      "loss": 1.0198,
+      "step": 23825
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00016643888107350577,
+      "loss": 0.9066,
+      "step": 23830
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001664201672568158,
+      "loss": 0.9364,
+      "step": 23835
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016640144927683407,
+      "loss": 0.8964,
+      "step": 23840
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00016638272713473387,
+      "loss": 0.9231,
+      "step": 23845
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016636400083168878,
+      "loss": 1.0691,
+      "step": 23850
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00016634527036887245,
+      "loss": 0.8508,
+      "step": 23855
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001663265357474591,
+      "loss": 0.9175,
+      "step": 23860
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001663077969686229,
+      "loss": 0.927,
+      "step": 23865
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001662890540335385,
+      "loss": 0.9039,
+      "step": 23870
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00016627030694338067,
+      "loss": 1.0943,
+      "step": 23875
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00016625155569932455,
+      "loss": 0.9051,
+      "step": 23880
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00016623280030254542,
+      "loss": 0.9562,
+      "step": 23885
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.000166214040754219,
+      "loss": 0.9317,
+      "step": 23890
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016619527705552103,
+      "loss": 0.8951,
+      "step": 23895
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00016617650920762773,
+      "loss": 0.8851,
+      "step": 23900
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00016615773721171545,
+      "loss": 1.0109,
+      "step": 23905
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00016613896106896085,
+      "loss": 1.1413,
+      "step": 23910
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001661201807805409,
+      "loss": 1.0307,
+      "step": 23915
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016610139634763265,
+      "loss": 0.9929,
+      "step": 23920
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016608260777141361,
+      "loss": 0.8565,
+      "step": 23925
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016606381505306149,
+      "loss": 1.0399,
+      "step": 23930
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00016604501819375415,
+      "loss": 1.004,
+      "step": 23935
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016602621719466988,
+      "loss": 0.9961,
+      "step": 23940
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00016600741205698714,
+      "loss": 0.9562,
+      "step": 23945
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00016598860278188457,
+      "loss": 0.9537,
+      "step": 23950
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016596978937054129,
+      "loss": 1.1004,
+      "step": 23955
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00016595097182413643,
+      "loss": 0.9118,
+      "step": 23960
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016593215014384957,
+      "loss": 0.9165,
+      "step": 23965
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016591332433086044,
+      "loss": 1.1373,
+      "step": 23970
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001658944943863491,
+      "loss": 0.9159,
+      "step": 23975
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016587566031149576,
+      "loss": 0.8789,
+      "step": 23980
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016585682210748103,
+      "loss": 1.0135,
+      "step": 23985
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001658379797754857,
+      "loss": 0.864,
+      "step": 23990
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001658191333166908,
+      "loss": 1.0405,
+      "step": 23995
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016580028273227763,
+      "loss": 0.909,
+      "step": 24000
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001657814280234278,
+      "loss": 0.8624,
+      "step": 24005
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016576256919132321,
+      "loss": 0.9113,
+      "step": 24010
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016574370623714582,
+      "loss": 0.9907,
+      "step": 24015
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016572483916207808,
+      "loss": 0.9848,
+      "step": 24020
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016570596796730257,
+      "loss": 0.9872,
+      "step": 24025
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016568709265400212,
+      "loss": 0.9887,
+      "step": 24030
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00016566821322335992,
+      "loss": 0.9673,
+      "step": 24035
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00016564932967655933,
+      "loss": 0.9441,
+      "step": 24040
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00016563044201478396,
+      "loss": 1.0899,
+      "step": 24045
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001656115502392178,
+      "loss": 0.8546,
+      "step": 24050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016559265435104486,
+      "loss": 0.8485,
+      "step": 24055
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016557375435144968,
+      "loss": 0.9084,
+      "step": 24060
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00016555485024161693,
+      "loss": 0.8803,
+      "step": 24065
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016553594202273146,
+      "loss": 0.949,
+      "step": 24070
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00016551702969597854,
+      "loss": 1.1374,
+      "step": 24075
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016549811326254353,
+      "loss": 1.0132,
+      "step": 24080
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016547919272361222,
+      "loss": 0.9281,
+      "step": 24085
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00016546026808037054,
+      "loss": 0.8351,
+      "step": 24090
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001654413393340047,
+      "loss": 0.9989,
+      "step": 24095
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016542240648570114,
+      "loss": 0.9545,
+      "step": 24100
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016540346953664668,
+      "loss": 0.9323,
+      "step": 24105
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016538452848802824,
+      "loss": 1.0501,
+      "step": 24110
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001653655833410331,
+      "loss": 0.9696,
+      "step": 24115
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00016534663409684876,
+      "loss": 0.9939,
+      "step": 24120
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016532768075666295,
+      "loss": 1.0044,
+      "step": 24125
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001653087233216637,
+      "loss": 0.9425,
+      "step": 24130
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00016528976179303932,
+      "loss": 0.9097,
+      "step": 24135
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001652707961719783,
+      "loss": 0.927,
+      "step": 24140
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001652518264596694,
+      "loss": 1.0928,
+      "step": 24145
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016523285265730177,
+      "loss": 1.136,
+      "step": 24150
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016521387476606462,
+      "loss": 1.1036,
+      "step": 24155
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001651948927871475,
+      "loss": 0.9845,
+      "step": 24160
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016517590672174027,
+      "loss": 0.8934,
+      "step": 24165
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016515691657103298,
+      "loss": 0.9262,
+      "step": 24170
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00016513792233621594,
+      "loss": 0.9136,
+      "step": 24175
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016511892401847975,
+      "loss": 0.9593,
+      "step": 24180
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016509992161901525,
+      "loss": 0.9288,
+      "step": 24185
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.0001650809151390135,
+      "loss": 0.9901,
+      "step": 24190
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016506190457966585,
+      "loss": 0.9776,
+      "step": 24195
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016504288994216397,
+      "loss": 0.9804,
+      "step": 24200
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016502387122769964,
+      "loss": 0.8687,
+      "step": 24205
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.000165004848437465,
+      "loss": 0.9509,
+      "step": 24210
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016498582157265245,
+      "loss": 0.9531,
+      "step": 24215
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016496679063445456,
+      "loss": 0.9361,
+      "step": 24220
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016494775562406425,
+      "loss": 0.841,
+      "step": 24225
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.0001649287165426747,
+      "loss": 1.1791,
+      "step": 24230
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001649096733914792,
+      "loss": 0.9415,
+      "step": 24235
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00016489062617167145,
+      "loss": 0.9468,
+      "step": 24240
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016487157488444536,
+      "loss": 0.9885,
+      "step": 24245
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00016485251953099505,
+      "loss": 0.95,
+      "step": 24250
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00016483346011251498,
+      "loss": 1.0973,
+      "step": 24255
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001648143966301998,
+      "loss": 0.9447,
+      "step": 24260
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00016479532908524438,
+      "loss": 0.9326,
+      "step": 24265
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.000164776257478844,
+      "loss": 0.909,
+      "step": 24270
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00016475718181219398,
+      "loss": 0.985,
+      "step": 24275
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016473810208649003,
+      "loss": 0.8886,
+      "step": 24280
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016471901830292815,
+      "loss": 1.0953,
+      "step": 24285
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001646999304627045,
+      "loss": 0.9343,
+      "step": 24290
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001646808385670155,
+      "loss": 0.9034,
+      "step": 24295
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00016466174261705785,
+      "loss": 0.9041,
+      "step": 24300
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00016464264261402859,
+      "loss": 0.8977,
+      "step": 24305
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001646235385591248,
+      "loss": 1.0345,
+      "step": 24310
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001646044304535441,
+      "loss": 1.0311,
+      "step": 24315
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016458531829848406,
+      "loss": 1.018,
+      "step": 24320
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001645662020951427,
+      "loss": 0.8999,
+      "step": 24325
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001645470818447183,
+      "loss": 1.0338,
+      "step": 24330
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016452795754840928,
+      "loss": 0.9788,
+      "step": 24335
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001645088292074144,
+      "loss": 0.9003,
+      "step": 24340
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001644896968229326,
+      "loss": 0.9124,
+      "step": 24345
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001644705603961632,
+      "loss": 0.9349,
+      "step": 24350
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00016445141992830562,
+      "loss": 0.9792,
+      "step": 24355
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016443227542055965,
+      "loss": 1.0906,
+      "step": 24360
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016441312687412527,
+      "loss": 0.8854,
+      "step": 24365
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016439397429020272,
+      "loss": 1.1025,
+      "step": 24370
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016437481766999254,
+      "loss": 0.9103,
+      "step": 24375
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00016435565701469548,
+      "loss": 1.1122,
+      "step": 24380
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00016433649232551253,
+      "loss": 0.9879,
+      "step": 24385
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016431732360364497,
+      "loss": 0.9769,
+      "step": 24390
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001642981508502943,
+      "loss": 0.8533,
+      "step": 24395
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00016427897406666233,
+      "loss": 1.0708,
+      "step": 24400
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00016425979325395104,
+      "loss": 0.8953,
+      "step": 24405
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00016424060841336275,
+      "loss": 0.9551,
+      "step": 24410
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00016422141954609994,
+      "loss": 1.033,
+      "step": 24415
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001642022266533654,
+      "loss": 1.1218,
+      "step": 24420
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016418302973636223,
+      "loss": 0.8115,
+      "step": 24425
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001641638287962936,
+      "loss": 1.0355,
+      "step": 24430
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00016414462383436312,
+      "loss": 0.9497,
+      "step": 24435
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00016412541485177456,
+      "loss": 0.9697,
+      "step": 24440
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00016410620184973196,
+      "loss": 1.0461,
+      "step": 24445
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016408698482943962,
+      "loss": 0.9266,
+      "step": 24450
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001640677637921021,
+      "loss": 0.9798,
+      "step": 24455
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00016404853873892416,
+      "loss": 0.9772,
+      "step": 24460
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00016402930967111088,
+      "loss": 0.8672,
+      "step": 24465
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00016401007658986753,
+      "loss": 1.0654,
+      "step": 24470
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016399083949639968,
+      "loss": 0.9599,
+      "step": 24475
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016397159839191315,
+      "loss": 1.083,
+      "step": 24480
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016395235327761395,
+      "loss": 0.9926,
+      "step": 24485
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00016393310415470844,
+      "loss": 1.0446,
+      "step": 24490
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016391385102440314,
+      "loss": 1.0361,
+      "step": 24495
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016389459388790485,
+      "loss": 1.0226,
+      "step": 24500
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00016387533274642065,
+      "loss": 0.9835,
+      "step": 24505
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00016385606760115784,
+      "loss": 1.0376,
+      "step": 24510
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00016383679845332401,
+      "loss": 0.9766,
+      "step": 24515
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016381752530412693,
+      "loss": 0.9625,
+      "step": 24520
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016379824815477466,
+      "loss": 1.0135,
+      "step": 24525
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0001637789670064756,
+      "loss": 0.905,
+      "step": 24530
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001637596818604382,
+      "loss": 0.963,
+      "step": 24535
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00016374039271787133,
+      "loss": 0.9131,
+      "step": 24540
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00016372109957998404,
+      "loss": 0.8397,
+      "step": 24545
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016370180244798567,
+      "loss": 0.7781,
+      "step": 24550
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016368250132308578,
+      "loss": 0.9317,
+      "step": 24555
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001636631962064942,
+      "loss": 0.9688,
+      "step": 24560
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.91015625,
+      "learning_rate": 0.00016364388709942093,
+      "loss": 1.1305,
+      "step": 24565
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016362457400307637,
+      "loss": 0.9365,
+      "step": 24570
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.000163605256918671,
+      "loss": 1.0317,
+      "step": 24575
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0703125,
+      "learning_rate": 0.00016358593584741576,
+      "loss": 1.0129,
+      "step": 24580
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016356661079052157,
+      "loss": 0.882,
+      "step": 24585
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016354728174919984,
+      "loss": 0.9529,
+      "step": 24590
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016352794872466215,
+      "loss": 0.9644,
+      "step": 24595
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016350861171812023,
+      "loss": 0.8416,
+      "step": 24600
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016348927073078624,
+      "loss": 0.9917,
+      "step": 24605
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00016346992576387242,
+      "loss": 1.1237,
+      "step": 24610
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001634505768185914,
+      "loss": 0.8601,
+      "step": 24615
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00016343122389615594,
+      "loss": 1.0272,
+      "step": 24620
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016341186699777912,
+      "loss": 0.9921,
+      "step": 24625
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016339250612467426,
+      "loss": 0.7782,
+      "step": 24630
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00016337314127805495,
+      "loss": 0.9831,
+      "step": 24635
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001633537724591349,
+      "loss": 0.8844,
+      "step": 24640
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00016333439966912828,
+      "loss": 0.9384,
+      "step": 24645
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00016331502290924937,
+      "loss": 0.8616,
+      "step": 24650
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016329564218071273,
+      "loss": 0.7969,
+      "step": 24655
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001632762574847331,
+      "loss": 1.045,
+      "step": 24660
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001632568688225256,
+      "loss": 1.0361,
+      "step": 24665
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00016323747619530554,
+      "loss": 0.9363,
+      "step": 24670
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00016321807960428843,
+      "loss": 0.9388,
+      "step": 24675
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016319867905069009,
+      "loss": 1.0529,
+      "step": 24680
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016317927453572657,
+      "loss": 0.8853,
+      "step": 24685
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00016315986606061416,
+      "loss": 0.9914,
+      "step": 24690
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016314045362656945,
+      "loss": 1.0792,
+      "step": 24695
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001631210372348092,
+      "loss": 0.9683,
+      "step": 24700
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00016310161688655036,
+      "loss": 1.0693,
+      "step": 24705
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00016308219258301038,
+      "loss": 1.0178,
+      "step": 24710
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001630627643254067,
+      "loss": 1.0458,
+      "step": 24715
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00016304333211495715,
+      "loss": 0.9108,
+      "step": 24720
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016302389595287975,
+      "loss": 0.8966,
+      "step": 24725
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00016300445584039274,
+      "loss": 0.8942,
+      "step": 24730
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001629850117787147,
+      "loss": 0.8664,
+      "step": 24735
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001629655637690644,
+      "loss": 0.9213,
+      "step": 24740
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00016294611181266082,
+      "loss": 0.9186,
+      "step": 24745
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00016292665591072328,
+      "loss": 0.8974,
+      "step": 24750
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001629071960644713,
+      "loss": 0.9719,
+      "step": 24755
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00016288773227512459,
+      "loss": 0.9197,
+      "step": 24760
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001628682645439032,
+      "loss": 0.9864,
+      "step": 24765
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001628487928720274,
+      "loss": 0.9193,
+      "step": 24770
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001628293172607177,
+      "loss": 0.8703,
+      "step": 24775
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001628098377111948,
+      "loss": 1.0337,
+      "step": 24780
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00016279035422467976,
+      "loss": 0.8703,
+      "step": 24785
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00016277086680239382,
+      "loss": 0.8611,
+      "step": 24790
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00016275137544555842,
+      "loss": 1.1237,
+      "step": 24795
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00016273188015539537,
+      "loss": 1.1163,
+      "step": 24800
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016271238093312662,
+      "loss": 0.9018,
+      "step": 24805
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016269287777997442,
+      "loss": 0.9558,
+      "step": 24810
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016267337069716123,
+      "loss": 0.926,
+      "step": 24815
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00016265385968590977,
+      "loss": 1.006,
+      "step": 24820
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00016263434474744304,
+      "loss": 0.9435,
+      "step": 24825
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00016261482588298426,
+      "loss": 0.9433,
+      "step": 24830
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00016259530309375685,
+      "loss": 0.952,
+      "step": 24835
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00016257577638098457,
+      "loss": 0.9448,
+      "step": 24840
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016255624574589136,
+      "loss": 0.9726,
+      "step": 24845
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001625367111897014,
+      "loss": 0.9088,
+      "step": 24850
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001625171727136392,
+      "loss": 0.8722,
+      "step": 24855
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001624976303189294,
+      "loss": 1.1047,
+      "step": 24860
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00016247808400679693,
+      "loss": 1.0619,
+      "step": 24865
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016245853377846702,
+      "loss": 1.0366,
+      "step": 24870
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016243897963516508,
+      "loss": 1.0427,
+      "step": 24875
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00016241942157811678,
+      "loss": 1.0127,
+      "step": 24880
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016239985960854805,
+      "loss": 0.9216,
+      "step": 24885
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00016238029372768505,
+      "loss": 0.8924,
+      "step": 24890
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016236072393675417,
+      "loss": 0.9695,
+      "step": 24895
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001623411502369821,
+      "loss": 0.9501,
+      "step": 24900
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016232157262959573,
+      "loss": 1.0881,
+      "step": 24905
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7265625,
+      "learning_rate": 0.00016230199111582226,
+      "loss": 0.9827,
+      "step": 24910
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00016228240569688898,
+      "loss": 0.9261,
+      "step": 24915
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001622628163740236,
+      "loss": 0.9561,
+      "step": 24920
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016224322314845394,
+      "loss": 0.979,
+      "step": 24925
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00016222362602140818,
+      "loss": 1.0244,
+      "step": 24930
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001622040249941147,
+      "loss": 0.9016,
+      "step": 24935
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016218442006780208,
+      "loss": 0.928,
+      "step": 24940
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016216481124369918,
+      "loss": 1.0769,
+      "step": 24945
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001621451985230351,
+      "loss": 1.0409,
+      "step": 24950
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00016212558190703923,
+      "loss": 0.987,
+      "step": 24955
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016210596139694112,
+      "loss": 1.0778,
+      "step": 24960
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001620863369939706,
+      "loss": 0.9534,
+      "step": 24965
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.0001620667086993578,
+      "loss": 1.0017,
+      "step": 24970
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016204707651433297,
+      "loss": 0.8229,
+      "step": 24975
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016202744044012675,
+      "loss": 0.8823,
+      "step": 24980
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001620078004779699,
+      "loss": 0.9721,
+      "step": 24985
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001619881566290935,
+      "loss": 1.0124,
+      "step": 24990
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00016196850889472887,
+      "loss": 0.9635,
+      "step": 24995
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00016194885727610747,
+      "loss": 1.0222,
+      "step": 25000
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016192920177446118,
+      "loss": 0.9272,
+      "step": 25005
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00016190954239102197,
+      "loss": 0.8514,
+      "step": 25010
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016188987912702215,
+      "loss": 0.9072,
+      "step": 25015
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016187021198369426,
+      "loss": 1.005,
+      "step": 25020
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016185054096227094,
+      "loss": 0.9324,
+      "step": 25025
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00016183086606398533,
+      "loss": 0.8911,
+      "step": 25030
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001618111872900706,
+      "loss": 1.1515,
+      "step": 25035
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00016179150464176023,
+      "loss": 1.0191,
+      "step": 25040
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.000161771818120288,
+      "loss": 0.8963,
+      "step": 25045
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016175212772688786,
+      "loss": 0.9288,
+      "step": 25050
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016173243346279402,
+      "loss": 0.9832,
+      "step": 25055
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001617127353292409,
+      "loss": 0.8714,
+      "step": 25060
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0001616930333274633,
+      "loss": 1.0185,
+      "step": 25065
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001616733274586961,
+      "loss": 0.9872,
+      "step": 25070
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016165361772417448,
+      "loss": 0.9913,
+      "step": 25075
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001616339041251339,
+      "loss": 0.9422,
+      "step": 25080
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016161418666281,
+      "loss": 0.9239,
+      "step": 25085
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001615944653384387,
+      "loss": 0.8856,
+      "step": 25090
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016157474015325617,
+      "loss": 0.9683,
+      "step": 25095
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001615550111084988,
+      "loss": 0.9395,
+      "step": 25100
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00016153527820540324,
+      "loss": 0.8642,
+      "step": 25105
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016151554144520637,
+      "loss": 1.0026,
+      "step": 25110
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00016149580082914526,
+      "loss": 0.9313,
+      "step": 25115
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016147605635845734,
+      "loss": 1.0754,
+      "step": 25120
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00016145630803438018,
+      "loss": 0.9872,
+      "step": 25125
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016143655585815165,
+      "loss": 0.9619,
+      "step": 25130
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016141679983100983,
+      "loss": 0.965,
+      "step": 25135
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00016139703995419303,
+      "loss": 0.9198,
+      "step": 25140
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016137727622893988,
+      "loss": 1.0207,
+      "step": 25145
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016135750865648913,
+      "loss": 0.8741,
+      "step": 25150
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00016133773723807986,
+      "loss": 0.9791,
+      "step": 25155
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00016131796197495134,
+      "loss": 0.8629,
+      "step": 25160
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016129818286834315,
+      "loss": 0.875,
+      "step": 25165
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00016127839991949503,
+      "loss": 0.9423,
+      "step": 25170
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00016125861312964705,
+      "loss": 0.9443,
+      "step": 25175
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016123882250003942,
+      "loss": 0.8957,
+      "step": 25180
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001612190280319126,
+      "loss": 1.0252,
+      "step": 25185
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016119922972650743,
+      "loss": 0.8536,
+      "step": 25190
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016117942758506483,
+      "loss": 0.9057,
+      "step": 25195
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016115962160882604,
+      "loss": 0.917,
+      "step": 25200
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001611398117990325,
+      "loss": 0.9083,
+      "step": 25205
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00016111999815692594,
+      "loss": 1.0793,
+      "step": 25210
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00016110018068374825,
+      "loss": 0.9917,
+      "step": 25215
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001610803593807417,
+      "loss": 1.0187,
+      "step": 25220
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001610605342491486,
+      "loss": 0.8417,
+      "step": 25225
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00016104070529021172,
+      "loss": 0.9645,
+      "step": 25230
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016102087250517388,
+      "loss": 0.882,
+      "step": 25235
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016100103589527826,
+      "loss": 1.0314,
+      "step": 25240
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016098119546176825,
+      "loss": 1.0143,
+      "step": 25245
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00016096135120588744,
+      "loss": 0.9423,
+      "step": 25250
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00016094150312887973,
+      "loss": 1.1706,
+      "step": 25255
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001609216512319892,
+      "loss": 0.9264,
+      "step": 25260
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016090179551646013,
+      "loss": 0.8544,
+      "step": 25265
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.78515625,
+      "learning_rate": 0.00016088193598353724,
+      "loss": 0.968,
+      "step": 25270
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016086207263446518,
+      "loss": 0.9744,
+      "step": 25275
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00016084220547048916,
+      "loss": 1.0384,
+      "step": 25280
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016082233449285437,
+      "loss": 1.0272,
+      "step": 25285
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00016080245970280638,
+      "loss": 0.9706,
+      "step": 25290
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.000160782581101591,
+      "loss": 1.0,
+      "step": 25295
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00016076269869045418,
+      "loss": 0.912,
+      "step": 25300
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001607428124706422,
+      "loss": 0.8741,
+      "step": 25305
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00016072292244340158,
+      "loss": 0.8362,
+      "step": 25310
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00016070302860997902,
+      "loss": 0.8294,
+      "step": 25315
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00016068313097162147,
+      "loss": 0.9885,
+      "step": 25320
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001606632295295762,
+      "loss": 0.8356,
+      "step": 25325
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00016064332428509056,
+      "loss": 1.0694,
+      "step": 25330
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00016062341523941234,
+      "loss": 1.0154,
+      "step": 25335
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016060350239378935,
+      "loss": 0.9468,
+      "step": 25340
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00016058358574946985,
+      "loss": 0.9269,
+      "step": 25345
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001605636653077022,
+      "loss": 1.0031,
+      "step": 25350
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.000160543741069735,
+      "loss": 0.9001,
+      "step": 25355
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001605238130368172,
+      "loss": 0.9295,
+      "step": 25360
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00016050388121019782,
+      "loss": 0.8978,
+      "step": 25365
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00016048394559112626,
+      "loss": 0.847,
+      "step": 25370
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00016046400618085214,
+      "loss": 0.9054,
+      "step": 25375
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00016044406298062526,
+      "loss": 0.9974,
+      "step": 25380
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016042411599169563,
+      "loss": 1.1149,
+      "step": 25385
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001604041652153136,
+      "loss": 0.9138,
+      "step": 25390
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001603842106527297,
+      "loss": 0.8508,
+      "step": 25395
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00016036425230519475,
+      "loss": 0.9259,
+      "step": 25400
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016034429017395966,
+      "loss": 0.8668,
+      "step": 25405
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00016032432426027577,
+      "loss": 1.0463,
+      "step": 25410
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00016030435456539452,
+      "loss": 0.9397,
+      "step": 25415
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016028438109056762,
+      "loss": 0.9873,
+      "step": 25420
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00016026440383704708,
+      "loss": 1.0089,
+      "step": 25425
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00016024442280608507,
+      "loss": 1.0219,
+      "step": 25430
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00016022443799893404,
+      "loss": 0.9601,
+      "step": 25435
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00016020444941684662,
+      "loss": 1.0713,
+      "step": 25440
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00016018445706107576,
+      "loss": 0.8197,
+      "step": 25445
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00016016446093287457,
+      "loss": 0.9444,
+      "step": 25450
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00016014446103349648,
+      "loss": 0.9886,
+      "step": 25455
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00016012445736419503,
+      "loss": 0.8165,
+      "step": 25460
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00016010444992622415,
+      "loss": 1.0564,
+      "step": 25465
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00016008443872083788,
+      "loss": 0.9417,
+      "step": 25470
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00016006442374929058,
+      "loss": 0.9726,
+      "step": 25475
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016004440501283677,
+      "loss": 0.9835,
+      "step": 25480
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00016002438251273127,
+      "loss": 0.9092,
+      "step": 25485
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00016000435625022913,
+      "loss": 0.926,
+      "step": 25490
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015998432622658557,
+      "loss": 0.939,
+      "step": 25495
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015996429244305617,
+      "loss": 0.9014,
+      "step": 25500
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015994425490089659,
+      "loss": 0.9478,
+      "step": 25505
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015992421360136282,
+      "loss": 1.0787,
+      "step": 25510
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00015990416854571115,
+      "loss": 0.8994,
+      "step": 25515
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00015988411973519794,
+      "loss": 0.9534,
+      "step": 25520
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001598640671710799,
+      "loss": 1.0001,
+      "step": 25525
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00015984401085461397,
+      "loss": 0.9634,
+      "step": 25530
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00015982395078705729,
+      "loss": 0.9582,
+      "step": 25535
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015980388696966723,
+      "loss": 0.8748,
+      "step": 25540
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0001597838194037014,
+      "loss": 0.9613,
+      "step": 25545
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001597637480904177,
+      "loss": 1.0516,
+      "step": 25550
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.0001597436730310742,
+      "loss": 0.9361,
+      "step": 25555
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015972359422692923,
+      "loss": 0.9223,
+      "step": 25560
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015970351167924138,
+      "loss": 0.98,
+      "step": 25565
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001596834253892694,
+      "loss": 1.0118,
+      "step": 25570
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015966333535827234,
+      "loss": 1.0899,
+      "step": 25575
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00015964324158750947,
+      "loss": 0.8592,
+      "step": 25580
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001596231440782403,
+      "loss": 0.9544,
+      "step": 25585
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015960304283172452,
+      "loss": 1.0262,
+      "step": 25590
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00015958293784922218,
+      "loss": 0.9641,
+      "step": 25595
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001595628291319934,
+      "loss": 1.0267,
+      "step": 25600
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001595427166812987,
+      "loss": 0.9957,
+      "step": 25605
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015952260049839864,
+      "loss": 1.0398,
+      "step": 25610
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00015950248058455423,
+      "loss": 1.001,
+      "step": 25615
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001594823569410266,
+      "loss": 1.0848,
+      "step": 25620
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00015946222956907704,
+      "loss": 0.9839,
+      "step": 25625
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015944209846996722,
+      "loss": 0.8242,
+      "step": 25630
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00015942196364495897,
+      "loss": 0.8995,
+      "step": 25635
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00015940182509531435,
+      "loss": 0.9686,
+      "step": 25640
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015938168282229572,
+      "loss": 1.0651,
+      "step": 25645
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00015936153682716557,
+      "loss": 1.0298,
+      "step": 25650
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00015934138711118666,
+      "loss": 0.9333,
+      "step": 25655
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.000159321233675622,
+      "loss": 0.9606,
+      "step": 25660
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015930107652173492,
+      "loss": 0.8906,
+      "step": 25665
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015928091565078879,
+      "loss": 0.9417,
+      "step": 25670
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00015926075106404737,
+      "loss": 0.9536,
+      "step": 25675
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015924058276277453,
+      "loss": 0.925,
+      "step": 25680
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015922041074823456,
+      "loss": 0.987,
+      "step": 25685
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015920023502169174,
+      "loss": 1.0728,
+      "step": 25690
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00015918005558441078,
+      "loss": 0.9732,
+      "step": 25695
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00015915987243765657,
+      "loss": 1.0096,
+      "step": 25700
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00015913968558269414,
+      "loss": 0.9989,
+      "step": 25705
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001591194950207889,
+      "loss": 1.0489,
+      "step": 25710
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00015909930075320633,
+      "loss": 0.9951,
+      "step": 25715
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015907910278121232,
+      "loss": 0.8418,
+      "step": 25720
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015905890110607285,
+      "loss": 1.0019,
+      "step": 25725
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00015903869572905422,
+      "loss": 1.1367,
+      "step": 25730
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00015901848665142288,
+      "loss": 0.9247,
+      "step": 25735
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015899827387444554,
+      "loss": 0.9569,
+      "step": 25740
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015897805739938927,
+      "loss": 0.8595,
+      "step": 25745
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00015895783722752116,
+      "loss": 0.9867,
+      "step": 25750
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.84375,
+      "learning_rate": 0.00015893761336010866,
+      "loss": 1.0285,
+      "step": 25755
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00015891738579841943,
+      "loss": 1.0182,
+      "step": 25760
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015889715454372137,
+      "loss": 0.9272,
+      "step": 25765
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00015887691959728256,
+      "loss": 1.0613,
+      "step": 25770
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001588566809603714,
+      "loss": 0.9419,
+      "step": 25775
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00015883643863425645,
+      "loss": 0.8621,
+      "step": 25780
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001588161926202065,
+      "loss": 0.9175,
+      "step": 25785
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001587959429194906,
+      "loss": 0.9916,
+      "step": 25790
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00015877568953337806,
+      "loss": 1.0484,
+      "step": 25795
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015875543246313836,
+      "loss": 1.063,
+      "step": 25800
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015873517171004125,
+      "loss": 1.0483,
+      "step": 25805
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00015871490727535666,
+      "loss": 1.0646,
+      "step": 25810
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001586946391603548,
+      "loss": 0.9879,
+      "step": 25815
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001586743673663061,
+      "loss": 0.9607,
+      "step": 25820
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015865409189448127,
+      "loss": 0.9211,
+      "step": 25825
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015863381274615115,
+      "loss": 1.0857,
+      "step": 25830
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00015861352992258684,
+      "loss": 0.8483,
+      "step": 25835
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00015859324342505974,
+      "loss": 1.0105,
+      "step": 25840
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001585729532548414,
+      "loss": 0.9075,
+      "step": 25845
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00015855265941320366,
+      "loss": 1.1377,
+      "step": 25850
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00015853236190141855,
+      "loss": 0.8974,
+      "step": 25855
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015851206072075829,
+      "loss": 0.9576,
+      "step": 25860
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00015849175587249545,
+      "loss": 1.062,
+      "step": 25865
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001584714473579027,
+      "loss": 0.9231,
+      "step": 25870
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00015845113517825313,
+      "loss": 0.9307,
+      "step": 25875
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015843081933481976,
+      "loss": 0.942,
+      "step": 25880
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001584104998288761,
+      "loss": 1.075,
+      "step": 25885
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001583901766616958,
+      "loss": 0.9113,
+      "step": 25890
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001583698498345527,
+      "loss": 0.8825,
+      "step": 25895
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015834951934872094,
+      "loss": 0.9867,
+      "step": 25900
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015832918520547487,
+      "loss": 0.9442,
+      "step": 25905
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00015830884740608906,
+      "loss": 1.0392,
+      "step": 25910
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00015828850595183823,
+      "loss": 1.052,
+      "step": 25915
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001582681608439975,
+      "loss": 0.8458,
+      "step": 25920
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015824781208384207,
+      "loss": 0.9396,
+      "step": 25925
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015822745967264745,
+      "loss": 1.1154,
+      "step": 25930
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00015820710361168935,
+      "loss": 0.9944,
+      "step": 25935
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001581867439022437,
+      "loss": 0.8879,
+      "step": 25940
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015816638054558666,
+      "loss": 0.9119,
+      "step": 25945
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015814601354299462,
+      "loss": 0.8787,
+      "step": 25950
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015812564289574427,
+      "loss": 1.0819,
+      "step": 25955
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00015810526860511243,
+      "loss": 0.9735,
+      "step": 25960
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00015808489067237614,
+      "loss": 1.1039,
+      "step": 25965
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00015806450909881277,
+      "loss": 0.9072,
+      "step": 25970
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00015804412388569986,
+      "loss": 0.8738,
+      "step": 25975
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00015802373503431513,
+      "loss": 0.9162,
+      "step": 25980
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00015800334254593661,
+      "loss": 0.9719,
+      "step": 25985
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00015798294642184251,
+      "loss": 0.9771,
+      "step": 25990
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00015796254666331131,
+      "loss": 1.0905,
+      "step": 25995
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00015794214327162167,
+      "loss": 0.8707,
+      "step": 26000
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00015792173624805245,
+      "loss": 0.8273,
+      "step": 26005
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00015790132559388291,
+      "loss": 0.9846,
+      "step": 26010
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.734375,
+      "learning_rate": 0.0001578809113103923,
+      "loss": 0.9594,
+      "step": 26015
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00015786049339886024,
+      "loss": 0.9845,
+      "step": 26020
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00015784007186056656,
+      "loss": 0.9394,
+      "step": 26025
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00015781964669679132,
+      "loss": 1.0991,
+      "step": 26030
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00015779921790881474,
+      "loss": 1.0305,
+      "step": 26035
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001577787854979174,
+      "loss": 1.0352,
+      "step": 26040
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00015775834946537995,
+      "loss": 0.9475,
+      "step": 26045
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001577379098124834,
+      "loss": 1.0782,
+      "step": 26050
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015771746654050887,
+      "loss": 0.9941,
+      "step": 26055
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00015769701965073782,
+      "loss": 0.9338,
+      "step": 26060
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00015767656914445188,
+      "loss": 1.0494,
+      "step": 26065
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001576561150229329,
+      "loss": 0.9931,
+      "step": 26070
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00015763565728746292,
+      "loss": 0.8672,
+      "step": 26075
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00015761519593932434,
+      "loss": 0.6515,
+      "step": 26080
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00015759473097979964,
+      "loss": 0.9082,
+      "step": 26085
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00015757426241017161,
+      "loss": 0.9968,
+      "step": 26090
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00015755379023172327,
+      "loss": 0.8977,
+      "step": 26095
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00015753331444573777,
+      "loss": 1.0706,
+      "step": 26100
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00015751283505349863,
+      "loss": 0.8743,
+      "step": 26105
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00015749235205628946,
+      "loss": 1.0556,
+      "step": 26110
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00015747186545539418,
+      "loss": 0.9144,
+      "step": 26115
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00015745137525209694,
+      "loss": 0.929,
+      "step": 26120
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015743088144768209,
+      "loss": 0.8177,
+      "step": 26125
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015741038404343412,
+      "loss": 0.9324,
+      "step": 26130
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015738988304063792,
+      "loss": 0.805,
+      "step": 26135
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015736937844057852,
+      "loss": 0.9338,
+      "step": 26140
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00015734887024454111,
+      "loss": 1.0461,
+      "step": 26145
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001573283584538112,
+      "loss": 1.0292,
+      "step": 26150
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001573078430696745,
+      "loss": 0.8662,
+      "step": 26155
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001572873240934169,
+      "loss": 0.9804,
+      "step": 26160
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00015726680152632462,
+      "loss": 1.1208,
+      "step": 26165
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.000157246275369684,
+      "loss": 1.0409,
+      "step": 26170
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001572257456247816,
+      "loss": 0.8872,
+      "step": 26175
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00015720521229290434,
+      "loss": 0.8249,
+      "step": 26180
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001571846753753392,
+      "loss": 0.7681,
+      "step": 26185
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00015716413487337346,
+      "loss": 0.9822,
+      "step": 26190
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00015714359078829467,
+      "loss": 0.9559,
+      "step": 26195
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001571230431213905,
+      "loss": 0.9754,
+      "step": 26200
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00015710249187394896,
+      "loss": 0.8946,
+      "step": 26205
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00015708193704725817,
+      "loss": 0.9902,
+      "step": 26210
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001570613786426066,
+      "loss": 0.9934,
+      "step": 26215
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00015704081666128276,
+      "loss": 1.0759,
+      "step": 26220
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00015702025110457562,
+      "loss": 0.9217,
+      "step": 26225
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001569996819737742,
+      "loss": 0.9059,
+      "step": 26230
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00015697910927016775,
+      "loss": 0.9285,
+      "step": 26235
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00015695853299504587,
+      "loss": 1.0308,
+      "step": 26240
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015693795314969825,
+      "loss": 0.9831,
+      "step": 26245
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015691736973541493,
+      "loss": 0.9238,
+      "step": 26250
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.000156896782753486,
+      "loss": 0.9768,
+      "step": 26255
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00015687619220520194,
+      "loss": 0.9166,
+      "step": 26260
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001568555980918534,
+      "loss": 0.8379,
+      "step": 26265
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.0001568350004147312,
+      "loss": 0.9155,
+      "step": 26270
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015681439917512646,
+      "loss": 0.8975,
+      "step": 26275
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015679379437433046,
+      "loss": 0.9756,
+      "step": 26280
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00015677318601363472,
+      "loss": 1.0602,
+      "step": 26285
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00015675257409433107,
+      "loss": 0.8572,
+      "step": 26290
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00015673195861771143,
+      "loss": 0.849,
+      "step": 26295
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.000156711339585068,
+      "loss": 1.04,
+      "step": 26300
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001566907169976932,
+      "loss": 0.9551,
+      "step": 26305
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00015667009085687972,
+      "loss": 1.0188,
+      "step": 26310
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001566494611639204,
+      "loss": 0.8871,
+      "step": 26315
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00015662882792010828,
+      "loss": 0.8827,
+      "step": 26320
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015660819112673678,
+      "loss": 1.01,
+      "step": 26325
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001565875507850994,
+      "loss": 0.8856,
+      "step": 26330
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001565669068964899,
+      "loss": 1.0583,
+      "step": 26335
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.21875,
+      "learning_rate": 0.0001565462594622022,
+      "loss": 1.0725,
+      "step": 26340
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.00015652560848353057,
+      "loss": 1.0632,
+      "step": 26345
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001565049539617695,
+      "loss": 1.0049,
+      "step": 26350
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00015648429589821348,
+      "loss": 1.0311,
+      "step": 26355
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00015646363429415748,
+      "loss": 0.9113,
+      "step": 26360
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00015644296915089657,
+      "loss": 0.9195,
+      "step": 26365
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015642230046972606,
+      "loss": 0.891,
+      "step": 26370
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015640162825194151,
+      "loss": 1.0039,
+      "step": 26375
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.0001563809524988387,
+      "loss": 1.0073,
+      "step": 26380
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00015636027321171353,
+      "loss": 1.0067,
+      "step": 26385
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00015633959039186227,
+      "loss": 1.0397,
+      "step": 26390
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001563189040405813,
+      "loss": 0.9546,
+      "step": 26395
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.921875,
+      "learning_rate": 0.0001562982141591673,
+      "loss": 0.8314,
+      "step": 26400
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001562775207489171,
+      "loss": 0.9185,
+      "step": 26405
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001562568238111278,
+      "loss": 1.0778,
+      "step": 26410
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00015623612334709675,
+      "loss": 1.086,
+      "step": 26415
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001562154193581214,
+      "loss": 1.0063,
+      "step": 26420
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00015619471184549955,
+      "loss": 0.8892,
+      "step": 26425
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001561740008105292,
+      "loss": 0.8712,
+      "step": 26430
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00015615328625450848,
+      "loss": 1.0723,
+      "step": 26435
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001561325681787358,
+      "loss": 0.9601,
+      "step": 26440
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00015611184658450983,
+      "loss": 0.9405,
+      "step": 26445
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001560911214731294,
+      "loss": 0.8906,
+      "step": 26450
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001560703928458936,
+      "loss": 1.0349,
+      "step": 26455
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00015604966070410176,
+      "loss": 0.9879,
+      "step": 26460
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001560289250490533,
+      "loss": 0.9159,
+      "step": 26465
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00015600818588204805,
+      "loss": 0.9955,
+      "step": 26470
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00015598744320438588,
+      "loss": 0.9993,
+      "step": 26475
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.000155966697017367,
+      "loss": 0.868,
+      "step": 26480
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00015594594732229187,
+      "loss": 0.9074,
+      "step": 26485
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00015592519412046098,
+      "loss": 0.9526,
+      "step": 26490
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015590443741317524,
+      "loss": 0.8759,
+      "step": 26495
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001558836772017357,
+      "loss": 0.9871,
+      "step": 26500
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00015586291348744364,
+      "loss": 1.1237,
+      "step": 26505
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015584214627160055,
+      "loss": 0.9317,
+      "step": 26510
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001558213755555081,
+      "loss": 0.8633,
+      "step": 26515
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001558006013404683,
+      "loss": 1.0021,
+      "step": 26520
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001557798236277832,
+      "loss": 0.9625,
+      "step": 26525
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00015575904241875525,
+      "loss": 0.8798,
+      "step": 26530
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00015573825771468704,
+      "loss": 1.0856,
+      "step": 26535
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015571746951688136,
+      "loss": 0.9614,
+      "step": 26540
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00015569667782664118,
+      "loss": 0.8792,
+      "step": 26545
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015567588264526984,
+      "loss": 1.0623,
+      "step": 26550
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001556550839740708,
+      "loss": 1.0637,
+      "step": 26555
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00015563428181434764,
+      "loss": 1.0017,
+      "step": 26560
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00015561347616740436,
+      "loss": 1.1512,
+      "step": 26565
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00015559266703454508,
+      "loss": 0.9056,
+      "step": 26570
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015557185441707406,
+      "loss": 0.9349,
+      "step": 26575
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00015555103831629597,
+      "loss": 0.8511,
+      "step": 26580
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001555302187335155,
+      "loss": 0.9237,
+      "step": 26585
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00015550939567003771,
+      "loss": 0.9405,
+      "step": 26590
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00015548856912716774,
+      "loss": 0.9342,
+      "step": 26595
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00015546773910621106,
+      "loss": 0.9176,
+      "step": 26600
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015544690560847336,
+      "loss": 1.0353,
+      "step": 26605
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.80078125,
+      "learning_rate": 0.0001554260686352604,
+      "loss": 0.9468,
+      "step": 26610
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001554052281878784,
+      "loss": 0.8511,
+      "step": 26615
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001553843842676336,
+      "loss": 0.9141,
+      "step": 26620
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015536353687583247,
+      "loss": 1.1151,
+      "step": 26625
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001553426860137818,
+      "loss": 1.0005,
+      "step": 26630
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00015532183168278854,
+      "loss": 1.0243,
+      "step": 26635
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001553009738841599,
+      "loss": 0.9182,
+      "step": 26640
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001552801126192032,
+      "loss": 0.9837,
+      "step": 26645
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0001552592478892261,
+      "loss": 0.8765,
+      "step": 26650
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00015523837969553644,
+      "loss": 1.1424,
+      "step": 26655
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00015521750803944214,
+      "loss": 0.9813,
+      "step": 26660
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001551966329222516,
+      "loss": 0.903,
+      "step": 26665
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001551757543452733,
+      "loss": 0.9145,
+      "step": 26670
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001551548723098158,
+      "loss": 1.0128,
+      "step": 26675
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001551339868171881,
+      "loss": 0.8879,
+      "step": 26680
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015511309786869935,
+      "loss": 1.0045,
+      "step": 26685
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00015509220546565882,
+      "loss": 0.9818,
+      "step": 26690
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00015507130960937612,
+      "loss": 0.9332,
+      "step": 26695
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00015505041030116102,
+      "loss": 0.9466,
+      "step": 26700
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00015502950754232349,
+      "loss": 1.0231,
+      "step": 26705
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00015500860133417374,
+      "loss": 0.9721,
+      "step": 26710
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00015498769167802222,
+      "loss": 0.9158,
+      "step": 26715
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001549667785751796,
+      "loss": 0.9728,
+      "step": 26720
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00015494586202695665,
+      "loss": 0.9821,
+      "step": 26725
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015492494203466452,
+      "loss": 1.0645,
+      "step": 26730
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00015490401859961445,
+      "loss": 0.8899,
+      "step": 26735
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00015488309172311798,
+      "loss": 0.9968,
+      "step": 26740
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00015486216140648678,
+      "loss": 0.873,
+      "step": 26745
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00015484122765103286,
+      "loss": 1.048,
+      "step": 26750
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0001548202904580683,
+      "loss": 0.8578,
+      "step": 26755
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00015479934982890551,
+      "loss": 0.9117,
+      "step": 26760
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001547784057648571,
+      "loss": 0.9955,
+      "step": 26765
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015475745826723576,
+      "loss": 0.9677,
+      "step": 26770
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015473650733735463,
+      "loss": 0.9753,
+      "step": 26775
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015471555297652686,
+      "loss": 0.9088,
+      "step": 26780
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001546945951860659,
+      "loss": 0.9215,
+      "step": 26785
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00015467363396728543,
+      "loss": 0.8744,
+      "step": 26790
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015465266932149932,
+      "loss": 0.9558,
+      "step": 26795
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015463170125002166,
+      "loss": 0.9333,
+      "step": 26800
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00015461072975416675,
+      "loss": 0.951,
+      "step": 26805
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015458975483524905,
+      "loss": 0.8719,
+      "step": 26810
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001545687764945834,
+      "loss": 0.8961,
+      "step": 26815
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00015454779473348467,
+      "loss": 0.9622,
+      "step": 26820
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015452680955326803,
+      "loss": 0.9103,
+      "step": 26825
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00015450582095524892,
+      "loss": 0.9518,
+      "step": 26830
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015448482894074282,
+      "loss": 0.9383,
+      "step": 26835
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00015446383351106562,
+      "loss": 1.0018,
+      "step": 26840
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001544428346675333,
+      "loss": 0.9988,
+      "step": 26845
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00015442183241146207,
+      "loss": 0.96,
+      "step": 26850
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015440082674416842,
+      "loss": 0.9955,
+      "step": 26855
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.000154379817666969,
+      "loss": 0.9501,
+      "step": 26860
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015435880518118066,
+      "loss": 1.0952,
+      "step": 26865
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001543377892881205,
+      "loss": 0.866,
+      "step": 26870
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00015431676998910586,
+      "loss": 0.9214,
+      "step": 26875
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015429574728545418,
+      "loss": 1.0176,
+      "step": 26880
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00015427472117848323,
+      "loss": 0.9703,
+      "step": 26885
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00015425369166951095,
+      "loss": 0.9798,
+      "step": 26890
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001542326587598555,
+      "loss": 0.9244,
+      "step": 26895
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001542116224508352,
+      "loss": 1.0506,
+      "step": 26900
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015419058274376867,
+      "loss": 0.8334,
+      "step": 26905
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00015416953963997472,
+      "loss": 1.0585,
+      "step": 26910
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.0001541484931407723,
+      "loss": 1.1043,
+      "step": 26915
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015412744324748068,
+      "loss": 0.9439,
+      "step": 26920
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00015410638996141927,
+      "loss": 0.8591,
+      "step": 26925
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001540853332839077,
+      "loss": 0.9574,
+      "step": 26930
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015406427321626586,
+      "loss": 1.0304,
+      "step": 26935
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015404320975981382,
+      "loss": 0.8397,
+      "step": 26940
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001540221429158718,
+      "loss": 1.142,
+      "step": 26945
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00015400107268576037,
+      "loss": 1.2038,
+      "step": 26950
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015397999907080015,
+      "loss": 1.0749,
+      "step": 26955
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015395892207231216,
+      "loss": 0.9764,
+      "step": 26960
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00015393784169161746,
+      "loss": 0.8951,
+      "step": 26965
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00015391675793003742,
+      "loss": 1.0228,
+      "step": 26970
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015389567078889356,
+      "loss": 0.9191,
+      "step": 26975
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001538745802695077,
+      "loss": 1.1133,
+      "step": 26980
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00015385348637320182,
+      "loss": 0.9428,
+      "step": 26985
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00015383238910129804,
+      "loss": 0.8891,
+      "step": 26990
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001538112884551188,
+      "loss": 0.9961,
+      "step": 26995
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015379018443598672,
+      "loss": 0.9033,
+      "step": 27000
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00015376907704522464,
+      "loss": 0.9199,
+      "step": 27005
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00015374796628415556,
+      "loss": 1.0566,
+      "step": 27010
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00015372685215410273,
+      "loss": 1.057,
+      "step": 27015
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00015370573465638962,
+      "loss": 1.0201,
+      "step": 27020
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015368461379233986,
+      "loss": 0.9872,
+      "step": 27025
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00015366348956327743,
+      "loss": 0.9782,
+      "step": 27030
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015364236197052634,
+      "loss": 0.9006,
+      "step": 27035
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001536212310154109,
+      "loss": 0.9762,
+      "step": 27040
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015360009669925564,
+      "loss": 0.8898,
+      "step": 27045
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015357895902338523,
+      "loss": 0.9673,
+      "step": 27050
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00015355781798912467,
+      "loss": 1.0248,
+      "step": 27055
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00015353667359779908,
+      "loss": 1.0253,
+      "step": 27060
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00015351552585073384,
+      "loss": 0.9211,
+      "step": 27065
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001534943747492545,
+      "loss": 0.7949,
+      "step": 27070
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001534732202946868,
+      "loss": 0.9069,
+      "step": 27075
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00015345206248835674,
+      "loss": 1.0244,
+      "step": 27080
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00015343090133159053,
+      "loss": 0.9724,
+      "step": 27085
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00015340973682571459,
+      "loss": 0.9931,
+      "step": 27090
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015338856897205552,
+      "loss": 0.9437,
+      "step": 27095
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00015336739777194013,
+      "loss": 0.9188,
+      "step": 27100
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00015334622322669543,
+      "loss": 1.1199,
+      "step": 27105
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00015332504533764876,
+      "loss": 0.9538,
+      "step": 27110
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001533038641061275,
+      "loss": 0.9649,
+      "step": 27115
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00015328267953345934,
+      "loss": 1.0045,
+      "step": 27120
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00015326149162097212,
+      "loss": 1.0143,
+      "step": 27125
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00015324030036999395,
+      "loss": 0.969,
+      "step": 27130
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0001532191057818531,
+      "loss": 0.9579,
+      "step": 27135
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001531979078578781,
+      "loss": 0.8333,
+      "step": 27140
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00015317670659939768,
+      "loss": 0.941,
+      "step": 27145
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00015315550200774072,
+      "loss": 0.9879,
+      "step": 27150
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015313429408423632,
+      "loss": 0.8847,
+      "step": 27155
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00015311308283021385,
+      "loss": 0.9423,
+      "step": 27160
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001530918682470029,
+      "loss": 0.9994,
+      "step": 27165
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00015307065033593316,
+      "loss": 0.9144,
+      "step": 27170
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00015304942909833463,
+      "loss": 0.9032,
+      "step": 27175
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015302820453553745,
+      "loss": 0.9647,
+      "step": 27180
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.000153006976648872,
+      "loss": 0.9738,
+      "step": 27185
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001529857454396689,
+      "loss": 0.9856,
+      "step": 27190
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00015296451090925897,
+      "loss": 0.8408,
+      "step": 27195
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001529432730589731,
+      "loss": 0.9164,
+      "step": 27200
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00015292203189014262,
+      "loss": 1.1075,
+      "step": 27205
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001529007874040989,
+      "loss": 0.8321,
+      "step": 27210
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00015287953960217357,
+      "loss": 0.9266,
+      "step": 27215
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001528582884856985,
+      "loss": 1.0955,
+      "step": 27220
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001528370340560057,
+      "loss": 0.8284,
+      "step": 27225
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001528157763144274,
+      "loss": 0.9035,
+      "step": 27230
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001527945152622961,
+      "loss": 0.7278,
+      "step": 27235
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00015277325090094443,
+      "loss": 0.9253,
+      "step": 27240
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00015275198323170535,
+      "loss": 0.8693,
+      "step": 27245
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00015273071225591187,
+      "loss": 0.9612,
+      "step": 27250
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015270943797489724,
+      "loss": 0.9629,
+      "step": 27255
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015268816038999504,
+      "loss": 0.9402,
+      "step": 27260
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015266687950253894,
+      "loss": 0.9483,
+      "step": 27265
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00015264559531386285,
+      "loss": 0.9645,
+      "step": 27270
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015262430782530092,
+      "loss": 1.0064,
+      "step": 27275
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001526030170381874,
+      "loss": 0.9229,
+      "step": 27280
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00015258172295385688,
+      "loss": 1.114,
+      "step": 27285
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00015256042557364405,
+      "loss": 1.021,
+      "step": 27290
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001525391248988839,
+      "loss": 1.0266,
+      "step": 27295
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001525178209309116,
+      "loss": 1.1505,
+      "step": 27300
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001524965136710624,
+      "loss": 0.9701,
+      "step": 27305
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00015247520312067198,
+      "loss": 1.1152,
+      "step": 27310
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00015245388928107606,
+      "loss": 0.8663,
+      "step": 27315
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015243257215361063,
+      "loss": 0.8906,
+      "step": 27320
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00015241125173961185,
+      "loss": 0.9771,
+      "step": 27325
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015238992804041612,
+      "loss": 0.9428,
+      "step": 27330
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00015236860105736003,
+      "loss": 0.9696,
+      "step": 27335
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00015234727079178038,
+      "loss": 0.9693,
+      "step": 27340
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00015232593724501419,
+      "loss": 0.8225,
+      "step": 27345
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00015230460041839862,
+      "loss": 1.0572,
+      "step": 27350
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001522832603132712,
+      "loss": 0.923,
+      "step": 27355
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015226191693096944,
+      "loss": 0.9713,
+      "step": 27360
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00015224057027283117,
+      "loss": 1.0176,
+      "step": 27365
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001522192203401945,
+      "loss": 0.9373,
+      "step": 27370
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015219786713439762,
+      "loss": 1.1225,
+      "step": 27375
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015217651065677898,
+      "loss": 0.8477,
+      "step": 27380
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015215515090867722,
+      "loss": 0.9262,
+      "step": 27385
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001521337878914312,
+      "loss": 0.8049,
+      "step": 27390
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00015211242160637997,
+      "loss": 0.942,
+      "step": 27395
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00015209105205486284,
+      "loss": 0.9257,
+      "step": 27400
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015206967923821923,
+      "loss": 0.9311,
+      "step": 27405
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001520483031577888,
+      "loss": 0.9391,
+      "step": 27410
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00015202692381491146,
+      "loss": 0.9197,
+      "step": 27415
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00015200554121092726,
+      "loss": 1.159,
+      "step": 27420
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015198415534717653,
+      "loss": 0.9105,
+      "step": 27425
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00015196276622499977,
+      "loss": 0.933,
+      "step": 27430
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0001519413738457376,
+      "loss": 0.8771,
+      "step": 27435
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00015191997821073098,
+      "loss": 0.9563,
+      "step": 27440
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00015189857932132098,
+      "loss": 0.9751,
+      "step": 27445
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015187717717884893,
+      "loss": 0.8955,
+      "step": 27450
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00015185577178465631,
+      "loss": 0.8752,
+      "step": 27455
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00015183436314008487,
+      "loss": 0.8825,
+      "step": 27460
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00015181295124647653,
+      "loss": 0.9836,
+      "step": 27465
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00015179153610517338,
+      "loss": 1.0022,
+      "step": 27470
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00015177011771751777,
+      "loss": 0.8779,
+      "step": 27475
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001517486960848522,
+      "loss": 0.9912,
+      "step": 27480
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015172727120851947,
+      "loss": 0.9619,
+      "step": 27485
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001517058430898624,
+      "loss": 1.092,
+      "step": 27490
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00015168441173022426,
+      "loss": 0.928,
+      "step": 27495
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015166297713094828,
+      "loss": 0.8879,
+      "step": 27500
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001516415392933781,
+      "loss": 1.0288,
+      "step": 27505
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00015162009821885738,
+      "loss": 0.9421,
+      "step": 27510
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00015159865390873014,
+      "loss": 0.9683,
+      "step": 27515
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001515772063643405,
+      "loss": 0.8493,
+      "step": 27520
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00015155575558703282,
+      "loss": 1.0029,
+      "step": 27525
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015153430157815168,
+      "loss": 0.9784,
+      "step": 27530
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001515128443390418,
+      "loss": 0.9032,
+      "step": 27535
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00015149138387104817,
+      "loss": 0.9295,
+      "step": 27540
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00015146992017551597,
+      "loss": 0.9388,
+      "step": 27545
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00015144845325379053,
+      "loss": 0.9346,
+      "step": 27550
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00015142698310721748,
+      "loss": 1.0505,
+      "step": 27555
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015140550973714251,
+      "loss": 1.0346,
+      "step": 27560
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001513840331449117,
+      "loss": 0.9626,
+      "step": 27565
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00015136255333187115,
+      "loss": 1.002,
+      "step": 27570
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00015134107029936725,
+      "loss": 0.8865,
+      "step": 27575
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001513195840487466,
+      "loss": 0.9616,
+      "step": 27580
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00015129809458135597,
+      "loss": 0.9437,
+      "step": 27585
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015127660189854237,
+      "loss": 1.066,
+      "step": 27590
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015125510600165295,
+      "loss": 0.8881,
+      "step": 27595
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00015123360689203507,
+      "loss": 0.8513,
+      "step": 27600
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015121210457103642,
+      "loss": 0.9547,
+      "step": 27605
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00015119059904000466,
+      "loss": 0.9742,
+      "step": 27610
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015116909030028793,
+      "loss": 0.8072,
+      "step": 27615
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001511475783532343,
+      "loss": 0.8026,
+      "step": 27620
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001511260632001922,
+      "loss": 0.9882,
+      "step": 27625
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015110454484251027,
+      "loss": 0.8986,
+      "step": 27630
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001510830232815372,
+      "loss": 0.9226,
+      "step": 27635
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015106149851862213,
+      "loss": 0.954,
+      "step": 27640
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015103997055511414,
+      "loss": 0.9185,
+      "step": 27645
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00015101843939236263,
+      "loss": 0.8623,
+      "step": 27650
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001509969050317173,
+      "loss": 1.0706,
+      "step": 27655
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015097536747452785,
+      "loss": 1.0242,
+      "step": 27660
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00015095382672214428,
+      "loss": 0.8957,
+      "step": 27665
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00015093228277591688,
+      "loss": 0.8766,
+      "step": 27670
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00015091073563719596,
+      "loss": 0.874,
+      "step": 27675
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015088918530733217,
+      "loss": 0.8828,
+      "step": 27680
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015086763178767627,
+      "loss": 0.8816,
+      "step": 27685
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00015084607507957924,
+      "loss": 0.8533,
+      "step": 27690
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00015082451518439238,
+      "loss": 0.9149,
+      "step": 27695
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.000150802952103467,
+      "loss": 1.0048,
+      "step": 27700
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001507813858381547,
+      "loss": 0.9428,
+      "step": 27705
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00015075981638980733,
+      "loss": 0.8444,
+      "step": 27710
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001507382437597768,
+      "loss": 0.9639,
+      "step": 27715
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015071666794941544,
+      "loss": 0.9995,
+      "step": 27720
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00015069508896007553,
+      "loss": 0.9617,
+      "step": 27725
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00015067350679310971,
+      "loss": 0.901,
+      "step": 27730
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00015065192144987077,
+      "loss": 1.0665,
+      "step": 27735
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00015063033293171173,
+      "loss": 1.0087,
+      "step": 27740
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00015060874123998575,
+      "loss": 1.0189,
+      "step": 27745
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015058714637604623,
+      "loss": 0.858,
+      "step": 27750
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00015056554834124675,
+      "loss": 0.9466,
+      "step": 27755
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001505439471369411,
+      "loss": 0.9885,
+      "step": 27760
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001505223427644833,
+      "loss": 0.9319,
+      "step": 27765
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00015050073522522751,
+      "loss": 1.2127,
+      "step": 27770
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00015047912452052813,
+      "loss": 1.0039,
+      "step": 27775
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00015045751065173972,
+      "loss": 0.9694,
+      "step": 27780
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00015043589362021708,
+      "loss": 1.0176,
+      "step": 27785
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001504142734273152,
+      "loss": 1.0114,
+      "step": 27790
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001503926500743892,
+      "loss": 0.8299,
+      "step": 27795
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00015037102356279457,
+      "loss": 0.961,
+      "step": 27800
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00015034939389388678,
+      "loss": 1.0191,
+      "step": 27805
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001503277610690216,
+      "loss": 0.8792,
+      "step": 27810
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001503061250895551,
+      "loss": 0.9465,
+      "step": 27815
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00015028448595684336,
+      "loss": 1.059,
+      "step": 27820
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00015026284367224276,
+      "loss": 0.9291,
+      "step": 27825
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00015024119823710987,
+      "loss": 0.9444,
+      "step": 27830
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00015021954965280148,
+      "loss": 0.8182,
+      "step": 27835
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001501978979206745,
+      "loss": 0.9046,
+      "step": 27840
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001501762430420861,
+      "loss": 1.1551,
+      "step": 27845
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00015015458501839367,
+      "loss": 0.9656,
+      "step": 27850
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00015013292385095475,
+      "loss": 0.9725,
+      "step": 27855
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.000150111259541127,
+      "loss": 0.9957,
+      "step": 27860
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00015008959209026848,
+      "loss": 1.0478,
+      "step": 27865
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001500679214997373,
+      "loss": 0.9713,
+      "step": 27870
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001500462477708917,
+      "loss": 0.948,
+      "step": 27875
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00015002457090509033,
+      "loss": 1.055,
+      "step": 27880
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001500028909036919,
+      "loss": 0.9306,
+      "step": 27885
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001499812077680553,
+      "loss": 0.8554,
+      "step": 27890
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001499595214995397,
+      "loss": 0.9094,
+      "step": 27895
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00014993783209950437,
+      "loss": 0.9488,
+      "step": 27900
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014991613956930885,
+      "loss": 0.8167,
+      "step": 27905
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00014989444391031283,
+      "loss": 0.8708,
+      "step": 27910
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00014987274512387631,
+      "loss": 0.9122,
+      "step": 27915
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00014985104321135927,
+      "loss": 1.0521,
+      "step": 27920
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00014982933817412207,
+      "loss": 1.0315,
+      "step": 27925
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00014980763001352522,
+      "loss": 1.0003,
+      "step": 27930
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00014978591873092938,
+      "loss": 0.9905,
+      "step": 27935
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00014976420432769545,
+      "loss": 0.9145,
+      "step": 27940
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014974248680518453,
+      "loss": 0.8594,
+      "step": 27945
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00014972076616475785,
+      "loss": 0.9862,
+      "step": 27950
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00014969904240777696,
+      "loss": 1.0253,
+      "step": 27955
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001496773155356035,
+      "loss": 0.9986,
+      "step": 27960
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001496555855495993,
+      "loss": 1.0846,
+      "step": 27965
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00014963385245112643,
+      "loss": 0.9269,
+      "step": 27970
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00014961211624154715,
+      "loss": 1.0928,
+      "step": 27975
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00014959037692222396,
+      "loss": 0.9256,
+      "step": 27980
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00014956863449451947,
+      "loss": 0.8836,
+      "step": 27985
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001495468889597965,
+      "loss": 0.9389,
+      "step": 27990
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001495251403194181,
+      "loss": 0.932,
+      "step": 27995
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00014950338857474751,
+      "loss": 1.0042,
+      "step": 28000
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00014948163372714812,
+      "loss": 0.9303,
+      "step": 28005
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001494598757779836,
+      "loss": 1.0224,
+      "step": 28010
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001494381147286177,
+      "loss": 0.9624,
+      "step": 28015
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001494163505804145,
+      "loss": 1.1779,
+      "step": 28020
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014939458333473814,
+      "loss": 0.8101,
+      "step": 28025
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014937281299295306,
+      "loss": 0.8563,
+      "step": 28030
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00014935103955642385,
+      "loss": 0.9414,
+      "step": 28035
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00014932926302651525,
+      "loss": 1.0134,
+      "step": 28040
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00014930748340459223,
+      "loss": 1.0407,
+      "step": 28045
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00014928570069202004,
+      "loss": 0.9893,
+      "step": 28050
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.000149263914890164,
+      "loss": 0.8799,
+      "step": 28055
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00014924212600038962,
+      "loss": 1.0822,
+      "step": 28060
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00014922033402406273,
+      "loss": 0.9435,
+      "step": 28065
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00014919853896254925,
+      "loss": 0.9677,
+      "step": 28070
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00014917674081721532,
+      "loss": 0.9603,
+      "step": 28075
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00014915493958942726,
+      "loss": 0.8813,
+      "step": 28080
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001491331352805516,
+      "loss": 0.9499,
+      "step": 28085
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00014911132789195507,
+      "loss": 1.0815,
+      "step": 28090
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001490895174250046,
+      "loss": 0.9011,
+      "step": 28095
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00014906770388106722,
+      "loss": 0.9064,
+      "step": 28100
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001490458872615103,
+      "loss": 0.936,
+      "step": 28105
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00014902406756770131,
+      "loss": 1.0428,
+      "step": 28110
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00014900224480100794,
+      "loss": 1.0521,
+      "step": 28115
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00014898041896279805,
+      "loss": 0.8189,
+      "step": 28120
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001489585900544397,
+      "loss": 0.8784,
+      "step": 28125
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00014893675807730117,
+      "loss": 0.9528,
+      "step": 28130
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0001489149230327509,
+      "loss": 1.0401,
+      "step": 28135
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00014889308492215756,
+      "loss": 0.9957,
+      "step": 28140
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00014887124374688999,
+      "loss": 1.0581,
+      "step": 28145
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00014884939950831716,
+      "loss": 0.9572,
+      "step": 28150
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00014882755220780837,
+      "loss": 0.8592,
+      "step": 28155
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.000148805701846733,
+      "loss": 0.8136,
+      "step": 28160
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001487838484264606,
+      "loss": 0.9009,
+      "step": 28165
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001487619919483611,
+      "loss": 0.9096,
+      "step": 28170
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00014874013241380436,
+      "loss": 0.9505,
+      "step": 28175
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00014871826982416062,
+      "loss": 0.9748,
+      "step": 28180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00014869640418080024,
+      "loss": 1.0135,
+      "step": 28185
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00014867453548509377,
+      "loss": 0.9244,
+      "step": 28190
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00014865266373841204,
+      "loss": 1.0503,
+      "step": 28195
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00014863078894212587,
+      "loss": 0.9883,
+      "step": 28200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00014860891109760646,
+      "loss": 1.0211,
+      "step": 28205
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0001485870302062252,
+      "loss": 0.9979,
+      "step": 28210
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001485651462693535,
+      "loss": 0.9707,
+      "step": 28215
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00014854325928836314,
+      "loss": 0.8423,
+      "step": 28220
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014852136926462602,
+      "loss": 0.878,
+      "step": 28225
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001484994761995142,
+      "loss": 0.9854,
+      "step": 28230
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00014847758009439995,
+      "loss": 1.0014,
+      "step": 28235
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00014845568095065578,
+      "loss": 0.9161,
+      "step": 28240
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00014843377876965437,
+      "loss": 0.9832,
+      "step": 28245
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014841187355276852,
+      "loss": 1.0405,
+      "step": 28250
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001483899653013713,
+      "loss": 0.9107,
+      "step": 28255
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014836805401683597,
+      "loss": 0.843,
+      "step": 28260
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00014834613970053588,
+      "loss": 0.9375,
+      "step": 28265
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001483242223538447,
+      "loss": 1.0245,
+      "step": 28270
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00014830230197813627,
+      "loss": 0.9011,
+      "step": 28275
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00014828037857478451,
+      "loss": 0.8082,
+      "step": 28280
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014825845214516364,
+      "loss": 0.9006,
+      "step": 28285
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00014823652269064803,
+      "loss": 0.9635,
+      "step": 28290
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00014821459021261224,
+      "loss": 0.9608,
+      "step": 28295
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.0625,
+      "learning_rate": 0.00014819265471243103,
+      "loss": 1.0338,
+      "step": 28300
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00014817071619147932,
+      "loss": 1.0203,
+      "step": 28305
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00014814877465113227,
+      "loss": 1.1001,
+      "step": 28310
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00014812683009276517,
+      "loss": 0.944,
+      "step": 28315
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00014810488251775357,
+      "loss": 1.0509,
+      "step": 28320
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014808293192747313,
+      "loss": 0.9797,
+      "step": 28325
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00014806097832329975,
+      "loss": 0.88,
+      "step": 28330
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00014803902170660953,
+      "loss": 1.0054,
+      "step": 28335
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001480170620787787,
+      "loss": 1.0066,
+      "step": 28340
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014799509944118374,
+      "loss": 1.0437,
+      "step": 28345
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00014797313379520132,
+      "loss": 0.9357,
+      "step": 28350
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00014795116514220818,
+      "loss": 0.9221,
+      "step": 28355
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00014792919348358144,
+      "loss": 0.913,
+      "step": 28360
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00014790721882069823,
+      "loss": 0.8986,
+      "step": 28365
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.000147885241154936,
+      "loss": 0.9824,
+      "step": 28370
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00014786326048767232,
+      "loss": 0.9492,
+      "step": 28375
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.000147841276820285,
+      "loss": 0.865,
+      "step": 28380
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001478192901541519,
+      "loss": 1.1203,
+      "step": 28385
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00014779730049065124,
+      "loss": 0.8508,
+      "step": 28390
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00014777530783116136,
+      "loss": 0.9426,
+      "step": 28395
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00014775331217706077,
+      "loss": 1.1074,
+      "step": 28400
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001477313135297282,
+      "loss": 0.9565,
+      "step": 28405
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00014770931189054252,
+      "loss": 0.9497,
+      "step": 28410
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00014768730726088286,
+      "loss": 0.8835,
+      "step": 28415
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00014766529964212844,
+      "loss": 1.0526,
+      "step": 28420
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00014764328903565875,
+      "loss": 0.9905,
+      "step": 28425
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001476212754428535,
+      "loss": 0.9627,
+      "step": 28430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00014759925886509241,
+      "loss": 0.8762,
+      "step": 28435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00014757723930375555,
+      "loss": 0.9423,
+      "step": 28440
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001475552167602232,
+      "loss": 1.1222,
+      "step": 28445
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00014753319123587567,
+      "loss": 1.0916,
+      "step": 28450
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00014751116273209358,
+      "loss": 0.9933,
+      "step": 28455
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00014748913125025773,
+      "loss": 0.9168,
+      "step": 28460
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.000147467096791749,
+      "loss": 0.9554,
+      "step": 28465
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00014744505935794858,
+      "loss": 0.8522,
+      "step": 28470
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00014742301895023785,
+      "loss": 1.1012,
+      "step": 28475
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014740097556999824,
+      "loss": 0.9742,
+      "step": 28480
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00014737892921861155,
+      "loss": 1.1805,
+      "step": 28485
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00014735687989745957,
+      "loss": 0.9158,
+      "step": 28490
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001473348276079244,
+      "loss": 1.0106,
+      "step": 28495
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00014731277235138839,
+      "loss": 1.0168,
+      "step": 28500
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001472907141292339,
+      "loss": 0.8944,
+      "step": 28505
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00014726865294284358,
+      "loss": 0.8638,
+      "step": 28510
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00014724658879360027,
+      "loss": 0.7973,
+      "step": 28515
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00014722452168288694,
+      "loss": 0.7767,
+      "step": 28520
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00014720245161208682,
+      "loss": 0.9628,
+      "step": 28525
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001471803785825833,
+      "loss": 0.9037,
+      "step": 28530
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00014715830259575988,
+      "loss": 1.0658,
+      "step": 28535
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8046875,
+      "learning_rate": 0.00014713622365300038,
+      "loss": 0.971,
+      "step": 28540
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014711414175568865,
+      "loss": 1.1279,
+      "step": 28545
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00014709205690520888,
+      "loss": 0.9409,
+      "step": 28550
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014706996910294535,
+      "loss": 1.0696,
+      "step": 28555
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00014704787835028257,
+      "loss": 0.8597,
+      "step": 28560
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00014702578464860516,
+      "loss": 0.9773,
+      "step": 28565
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00014700368799929804,
+      "loss": 0.9193,
+      "step": 28570
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014698158840374619,
+      "loss": 0.9545,
+      "step": 28575
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001469594858633349,
+      "loss": 0.9189,
+      "step": 28580
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00014693738037944954,
+      "loss": 0.9593,
+      "step": 28585
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00014691527195347573,
+      "loss": 0.9682,
+      "step": 28590
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00014689316058679922,
+      "loss": 0.8859,
+      "step": 28595
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.000146871046280806,
+      "loss": 1.1101,
+      "step": 28600
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00014684892903688224,
+      "loss": 0.9146,
+      "step": 28605
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00014682680885641424,
+      "loss": 0.989,
+      "step": 28610
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00014680468574078853,
+      "loss": 0.9693,
+      "step": 28615
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00014678255969139184,
+      "loss": 0.9098,
+      "step": 28620
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00014676043070961097,
+      "loss": 0.9663,
+      "step": 28625
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0001467382987968331,
+      "loss": 0.8683,
+      "step": 28630
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001467161639544454,
+      "loss": 0.9827,
+      "step": 28635
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00014669402618383535,
+      "loss": 1.0172,
+      "step": 28640
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00014667188548639056,
+      "loss": 1.003,
+      "step": 28645
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00014664974186349883,
+      "loss": 1.0974,
+      "step": 28650
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00014662759531654812,
+      "loss": 0.8167,
+      "step": 28655
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001466054458469267,
+      "loss": 0.959,
+      "step": 28660
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00014658329345602282,
+      "loss": 1.1323,
+      "step": 28665
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00014656113814522502,
+      "loss": 0.9222,
+      "step": 28670
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001465389799159221,
+      "loss": 0.926,
+      "step": 28675
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00014651681876950287,
+      "loss": 1.0413,
+      "step": 28680
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001464946547073565,
+      "loss": 0.9231,
+      "step": 28685
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00014647248773087219,
+      "loss": 0.9352,
+      "step": 28690
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00014645031784143946,
+      "loss": 0.8001,
+      "step": 28695
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014642814504044787,
+      "loss": 0.9051,
+      "step": 28700
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001464059693292873,
+      "loss": 0.9376,
+      "step": 28705
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00014638379070934767,
+      "loss": 1.1241,
+      "step": 28710
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00014636160918201927,
+      "loss": 0.9326,
+      "step": 28715
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001463394247486924,
+      "loss": 0.9006,
+      "step": 28720
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00014631723741075759,
+      "loss": 1.0671,
+      "step": 28725
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00014629504716960558,
+      "loss": 1.1036,
+      "step": 28730
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001462728540266273,
+      "loss": 1.0322,
+      "step": 28735
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00014625065798321382,
+      "loss": 0.9795,
+      "step": 28740
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00014622845904075643,
+      "loss": 0.9386,
+      "step": 28745
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00014620625720064657,
+      "loss": 1.0119,
+      "step": 28750
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00014618405246427592,
+      "loss": 1.0182,
+      "step": 28755
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00014616184483303622,
+      "loss": 0.9796,
+      "step": 28760
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00014613963430831948,
+      "loss": 0.9254,
+      "step": 28765
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.000146117420891518,
+      "loss": 0.8252,
+      "step": 28770
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00014609520458402404,
+      "loss": 0.8693,
+      "step": 28775
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0001460729853872301,
+      "loss": 0.9317,
+      "step": 28780
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.000146050763302529,
+      "loss": 0.8887,
+      "step": 28785
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00014602853833131361,
+      "loss": 0.9777,
+      "step": 28790
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00014600631047497698,
+      "loss": 0.9747,
+      "step": 28795
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00014598407973491248,
+      "loss": 1.0291,
+      "step": 28800
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00014596184611251345,
+      "loss": 0.9252,
+      "step": 28805
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014593960960917354,
+      "loss": 0.8716,
+      "step": 28810
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00014591737022628663,
+      "loss": 1.051,
+      "step": 28815
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001458951279652466,
+      "loss": 0.9231,
+      "step": 28820
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00014587288282744774,
+      "loss": 0.9758,
+      "step": 28825
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001458506348142843,
+      "loss": 0.9687,
+      "step": 28830
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00014582838392715087,
+      "loss": 0.8625,
+      "step": 28835
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00014580613016744213,
+      "loss": 0.9746,
+      "step": 28840
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00014578387353655296,
+      "loss": 0.9406,
+      "step": 28845
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001457616140358785,
+      "loss": 0.8996,
+      "step": 28850
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00014573935166681392,
+      "loss": 0.9932,
+      "step": 28855
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00014571708643075468,
+      "loss": 0.8784,
+      "step": 28860
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001456948183290964,
+      "loss": 1.002,
+      "step": 28865
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001456725473632349,
+      "loss": 0.9869,
+      "step": 28870
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00014565027353456608,
+      "loss": 0.9852,
+      "step": 28875
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00014562799684448617,
+      "loss": 0.8989,
+      "step": 28880
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00014560571729439138,
+      "loss": 0.9583,
+      "step": 28885
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00014558343488567835,
+      "loss": 0.9679,
+      "step": 28890
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00014556114961974368,
+      "loss": 1.1131,
+      "step": 28895
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00014553886149798424,
+      "loss": 1.0294,
+      "step": 28900
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00014551657052179712,
+      "loss": 0.971,
+      "step": 28905
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00014549427669257955,
+      "loss": 0.9676,
+      "step": 28910
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00014547198001172885,
+      "loss": 0.9163,
+      "step": 28915
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00014544968048064267,
+      "loss": 1.026,
+      "step": 28920
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00014542737810071879,
+      "loss": 0.8377,
+      "step": 28925
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014540507287335506,
+      "loss": 0.9355,
+      "step": 28930
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001453827647999497,
+      "loss": 0.8847,
+      "step": 28935
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00014536045388190093,
+      "loss": 1.0317,
+      "step": 28940
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001453381401206072,
+      "loss": 0.8708,
+      "step": 28945
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001453158235174673,
+      "loss": 0.9982,
+      "step": 28950
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00014529350407387995,
+      "loss": 0.9613,
+      "step": 28955
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00014527118179124415,
+      "loss": 0.8559,
+      "step": 28960
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00014524885667095914,
+      "loss": 1.0255,
+      "step": 28965
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00014522652871442425,
+      "loss": 0.9109,
+      "step": 28970
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.000145204197923039,
+      "loss": 0.8796,
+      "step": 28975
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001451818642982032,
+      "loss": 1.0481,
+      "step": 28980
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00014515952784131665,
+      "loss": 0.9254,
+      "step": 28985
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00014513718855377949,
+      "loss": 1.0509,
+      "step": 28990
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001451148464369919,
+      "loss": 0.9491,
+      "step": 28995
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00014509250149235438,
+      "loss": 0.9705,
+      "step": 29000
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00014507015372126753,
+      "loss": 0.9703,
+      "step": 29005
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00014504780312513208,
+      "loss": 1.0045,
+      "step": 29010
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00014502544970534906,
+      "loss": 0.9571,
+      "step": 29015
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00014500309346331954,
+      "loss": 1.0786,
+      "step": 29020
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00014498073440044487,
+      "loss": 0.9729,
+      "step": 29025
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00014495837251812655,
+      "loss": 0.9124,
+      "step": 29030
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00014493600781776626,
+      "loss": 0.8797,
+      "step": 29035
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001449136403007658,
+      "loss": 0.8734,
+      "step": 29040
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001448912699685272,
+      "loss": 1.0155,
+      "step": 29045
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00014486889682245268,
+      "loss": 0.9959,
+      "step": 29050
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0001448465208639446,
+      "loss": 0.9737,
+      "step": 29055
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014482414209440555,
+      "loss": 0.7726,
+      "step": 29060
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001448017605152382,
+      "loss": 0.8841,
+      "step": 29065
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001447793761278455,
+      "loss": 0.8884,
+      "step": 29070
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00014475698893363044,
+      "loss": 0.896,
+      "step": 29075
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00014473459893399638,
+      "loss": 1.1103,
+      "step": 29080
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00014471220613034675,
+      "loss": 1.0756,
+      "step": 29085
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00014468981052408506,
+      "loss": 1.0535,
+      "step": 29090
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00014466741211661516,
+      "loss": 0.8998,
+      "step": 29095
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.000144645010909341,
+      "loss": 0.9208,
+      "step": 29100
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0078125,
+      "learning_rate": 0.00014462260690366668,
+      "loss": 1.0528,
+      "step": 29105
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001446002001009966,
+      "loss": 1.0644,
+      "step": 29110
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001445777905027352,
+      "loss": 0.9569,
+      "step": 29115
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00014455537811028702,
+      "loss": 0.9952,
+      "step": 29120
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00014453296292505708,
+      "loss": 1.0159,
+      "step": 29125
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00014451054494845025,
+      "loss": 0.8709,
+      "step": 29130
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001444881241818718,
+      "loss": 0.92,
+      "step": 29135
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00014446570062672705,
+      "loss": 1.0585,
+      "step": 29140
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00014444327428442155,
+      "loss": 1.0007,
+      "step": 29145
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.000144420845156361,
+      "loss": 0.9254,
+      "step": 29150
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001443984132439513,
+      "loss": 0.9686,
+      "step": 29155
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00014437597854859846,
+      "loss": 0.9258,
+      "step": 29160
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014435354107170876,
+      "loss": 0.8689,
+      "step": 29165
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00014433110081468859,
+      "loss": 0.9698,
+      "step": 29170
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00014430865777894454,
+      "loss": 1.0518,
+      "step": 29175
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00014428621196588336,
+      "loss": 0.9741,
+      "step": 29180
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00014426376337691195,
+      "loss": 0.8741,
+      "step": 29185
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00014424131201343747,
+      "loss": 0.9883,
+      "step": 29190
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00014421885787686714,
+      "loss": 0.9213,
+      "step": 29195
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00014419640096860844,
+      "loss": 0.9058,
+      "step": 29200
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.000144173941290069,
+      "loss": 0.892,
+      "step": 29205
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001441514788426566,
+      "loss": 0.9142,
+      "step": 29210
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00014412901362777922,
+      "loss": 1.0063,
+      "step": 29215
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00014410654564684505,
+      "loss": 1.0145,
+      "step": 29220
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00014408407490126233,
+      "loss": 0.9077,
+      "step": 29225
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00014406160139243958,
+      "loss": 1.0044,
+      "step": 29230
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.0001440391251217855,
+      "loss": 1.1243,
+      "step": 29235
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00014401664609070889,
+      "loss": 1.1431,
+      "step": 29240
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00014399416430061875,
+      "loss": 0.789,
+      "step": 29245
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001439716797529243,
+      "loss": 0.8925,
+      "step": 29250
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001439491924490349,
+      "loss": 0.9749,
+      "step": 29255
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00014392670239036004,
+      "loss": 1.0446,
+      "step": 29260
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00014390420957830947,
+      "loss": 0.8864,
+      "step": 29265
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00014388171401429306,
+      "loss": 0.9329,
+      "step": 29270
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00014385921569972079,
+      "loss": 0.843,
+      "step": 29275
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00014383671463600293,
+      "loss": 0.8605,
+      "step": 29280
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00014381421082454988,
+      "loss": 0.9682,
+      "step": 29285
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001437917042667722,
+      "loss": 0.8813,
+      "step": 29290
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00014376919496408063,
+      "loss": 1.0273,
+      "step": 29295
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00014374668291788608,
+      "loss": 1.1245,
+      "step": 29300
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001437241681295996,
+      "loss": 0.9684,
+      "step": 29305
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00014370165060063246,
+      "loss": 0.858,
+      "step": 29310
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001436791303323961,
+      "loss": 0.9644,
+      "step": 29315
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001436566073263021,
+      "loss": 1.017,
+      "step": 29320
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00014363408158376225,
+      "loss": 0.9339,
+      "step": 29325
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00014361155310618846,
+      "loss": 1.0671,
+      "step": 29330
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00014358902189499283,
+      "loss": 0.9671,
+      "step": 29335
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00014356648795158766,
+      "loss": 0.9803,
+      "step": 29340
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00014354395127738544,
+      "loss": 1.1102,
+      "step": 29345
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00014352141187379876,
+      "loss": 1.1862,
+      "step": 29350
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001434988697422404,
+      "loss": 0.9587,
+      "step": 29355
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00014347632488412337,
+      "loss": 1.0016,
+      "step": 29360
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.00014345377730086076,
+      "loss": 0.9902,
+      "step": 29365
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00014343122699386589,
+      "loss": 0.9437,
+      "step": 29370
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00014340867396455228,
+      "loss": 0.9887,
+      "step": 29375
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001433861182143335,
+      "loss": 0.8093,
+      "step": 29380
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00014336355974462346,
+      "loss": 1.0003,
+      "step": 29385
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00014334099855683606,
+      "loss": 0.871,
+      "step": 29390
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00014331843465238556,
+      "loss": 1.0591,
+      "step": 29395
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00014329586803268625,
+      "loss": 0.9311,
+      "step": 29400
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001432732986991526,
+      "loss": 0.9105,
+      "step": 29405
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00014325072665319927,
+      "loss": 0.9017,
+      "step": 29410
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00014322815189624118,
+      "loss": 1.0109,
+      "step": 29415
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00014320557442969325,
+      "loss": 0.9553,
+      "step": 29420
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00014318299425497075,
+      "loss": 0.9934,
+      "step": 29425
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00014316041137348896,
+      "loss": 1.0326,
+      "step": 29430
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00014313782578666346,
+      "loss": 0.8656,
+      "step": 29435
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001431152374959099,
+      "loss": 1.0418,
+      "step": 29440
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00014309264650264414,
+      "loss": 1.0044,
+      "step": 29445
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001430700528082822,
+      "loss": 0.8477,
+      "step": 29450
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00014304745641424035,
+      "loss": 1.0616,
+      "step": 29455
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00014302485732193486,
+      "loss": 0.9678,
+      "step": 29460
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00014300225553278234,
+      "loss": 0.8745,
+      "step": 29465
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00014297965104819946,
+      "loss": 0.8087,
+      "step": 29470
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00014295704386960308,
+      "loss": 1.0838,
+      "step": 29475
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001429344339984103,
+      "loss": 0.9963,
+      "step": 29480
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00014291182143603826,
+      "loss": 0.8762,
+      "step": 29485
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001428892061839044,
+      "loss": 0.839,
+      "step": 29490
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00014286658824342626,
+      "loss": 0.8925,
+      "step": 29495
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00014284396761602152,
+      "loss": 0.8431,
+      "step": 29500
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00014282134430310814,
+      "loss": 0.9033,
+      "step": 29505
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00014279871830610413,
+      "loss": 0.9588,
+      "step": 29510
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00014277608962642765,
+      "loss": 0.9106,
+      "step": 29515
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00014275345826549718,
+      "loss": 0.951,
+      "step": 29520
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001427308242247313,
+      "loss": 0.8358,
+      "step": 29525
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00014270818750554862,
+      "loss": 1.019,
+      "step": 29530
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00014268554810936818,
+      "loss": 1.0055,
+      "step": 29535
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00014266290603760892,
+      "loss": 0.8804,
+      "step": 29540
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014264026129169014,
+      "loss": 0.9749,
+      "step": 29545
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001426176138730312,
+      "loss": 0.8841,
+      "step": 29550
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001425949637830517,
+      "loss": 0.943,
+      "step": 29555
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00014257231102317137,
+      "loss": 1.0354,
+      "step": 29560
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00014254965559481005,
+      "loss": 0.8461,
+      "step": 29565
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001425269974993879,
+      "loss": 0.959,
+      "step": 29570
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00014250433673832513,
+      "loss": 1.0894,
+      "step": 29575
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001424816733130421,
+      "loss": 0.9041,
+      "step": 29580
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001424590072249594,
+      "loss": 0.9017,
+      "step": 29585
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00014243633847549778,
+      "loss": 0.9902,
+      "step": 29590
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001424136670660781,
+      "loss": 1.0025,
+      "step": 29595
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001423909929981215,
+      "loss": 0.8774,
+      "step": 29600
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00014236831627304918,
+      "loss": 0.9068,
+      "step": 29605
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00014234563689228253,
+      "loss": 0.8929,
+      "step": 29610
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00014232295485724314,
+      "loss": 0.9433,
+      "step": 29615
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00014230027016935273,
+      "loss": 0.8688,
+      "step": 29620
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001422775828300332,
+      "loss": 0.9854,
+      "step": 29625
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00014225489284070667,
+      "loss": 0.9152,
+      "step": 29630
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001422322002027953,
+      "loss": 0.9432,
+      "step": 29635
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001422095049177215,
+      "loss": 1.1606,
+      "step": 29640
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00014218680698690792,
+      "loss": 0.9514,
+      "step": 29645
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00014216410641177717,
+      "loss": 0.87,
+      "step": 29650
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00014214140319375224,
+      "loss": 0.9848,
+      "step": 29655
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00014211869733425615,
+      "loss": 0.904,
+      "step": 29660
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00014209598883471215,
+      "loss": 0.9153,
+      "step": 29665
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00014207327769654365,
+      "loss": 0.9631,
+      "step": 29670
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00014205056392117414,
+      "loss": 1.0254,
+      "step": 29675
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00014202784751002744,
+      "loss": 0.898,
+      "step": 29680
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00014200512846452735,
+      "loss": 1.1391,
+      "step": 29685
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.000141982406786098,
+      "loss": 0.9663,
+      "step": 29690
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00014195968247616357,
+      "loss": 0.9682,
+      "step": 29695
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00014193695553614845,
+      "loss": 1.0978,
+      "step": 29700
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00014191422596747716,
+      "loss": 0.8681,
+      "step": 29705
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00014189149377157452,
+      "loss": 0.8885,
+      "step": 29710
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001418687589498653,
+      "loss": 1.0428,
+      "step": 29715
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00014184602150377456,
+      "loss": 0.814,
+      "step": 29720
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001418232814347276,
+      "loss": 0.9806,
+      "step": 29725
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00014180053874414967,
+      "loss": 0.907,
+      "step": 29730
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014177779343346638,
+      "loss": 0.9666,
+      "step": 29735
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00014175504550410342,
+      "loss": 1.0781,
+      "step": 29740
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00014173229495748665,
+      "loss": 0.9353,
+      "step": 29745
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00014170954179504213,
+      "loss": 0.982,
+      "step": 29750
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00014168678601819598,
+      "loss": 0.9841,
+      "step": 29755
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00014166402762837464,
+      "loss": 0.9569,
+      "step": 29760
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00014164126662700457,
+      "loss": 1.0546,
+      "step": 29765
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001416185030155125,
+      "loss": 0.869,
+      "step": 29770
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00014159573679532525,
+      "loss": 0.9614,
+      "step": 29775
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00014157296796786989,
+      "loss": 0.9047,
+      "step": 29780
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001415501965345735,
+      "loss": 0.9159,
+      "step": 29785
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00014152742249686347,
+      "loss": 0.896,
+      "step": 29790
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00014150464585616733,
+      "loss": 0.9605,
+      "step": 29795
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001414818666139127,
+      "loss": 0.9356,
+      "step": 29800
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00014145908477152742,
+      "loss": 0.9055,
+      "step": 29805
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00014143630033043952,
+      "loss": 1.1045,
+      "step": 29810
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00014141351329207706,
+      "loss": 1.0658,
+      "step": 29815
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001413907236578685,
+      "loss": 0.9089,
+      "step": 29820
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00014136793142924218,
+      "loss": 1.1593,
+      "step": 29825
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00014134513660762682,
+      "loss": 0.8836,
+      "step": 29830
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00014132233919445124,
+      "loss": 0.8684,
+      "step": 29835
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00014129953919114435,
+      "loss": 0.9364,
+      "step": 29840
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001412767365991353,
+      "loss": 1.017,
+      "step": 29845
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00014125393141985342,
+      "loss": 0.9966,
+      "step": 29850
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00014123112365472808,
+      "loss": 1.0908,
+      "step": 29855
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.000141208313305189,
+      "loss": 0.9723,
+      "step": 29860
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001411855003726659,
+      "loss": 0.927,
+      "step": 29865
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001411626848585887,
+      "loss": 0.9745,
+      "step": 29870
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0001411398667643876,
+      "loss": 1.08,
+      "step": 29875
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014111704609149275,
+      "loss": 0.7581,
+      "step": 29880
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001410942228413346,
+      "loss": 0.9267,
+      "step": 29885
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001410713970153438,
+      "loss": 0.961,
+      "step": 29890
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00014104856861495103,
+      "loss": 0.9943,
+      "step": 29895
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001410257376415873,
+      "loss": 0.9378,
+      "step": 29900
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00014100290409668352,
+      "loss": 0.9556,
+      "step": 29905
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00014098006798167104,
+      "loss": 1.1233,
+      "step": 29910
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00014095722929798122,
+      "loss": 0.7322,
+      "step": 29915
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00014093438804704567,
+      "loss": 0.925,
+      "step": 29920
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00014091154423029602,
+      "loss": 0.9481,
+      "step": 29925
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00014088869784916418,
+      "loss": 0.9175,
+      "step": 29930
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001408658489050822,
+      "loss": 0.9337,
+      "step": 29935
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00014084299739948225,
+      "loss": 0.8176,
+      "step": 29940
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00014082014333379673,
+      "loss": 0.9413,
+      "step": 29945
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001407972867094581,
+      "loss": 0.8769,
+      "step": 29950
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00014077442752789913,
+      "loss": 0.8575,
+      "step": 29955
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001407515657905526,
+      "loss": 0.9906,
+      "step": 29960
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00014072870149885148,
+      "loss": 0.9162,
+      "step": 29965
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00014070583465422898,
+      "loss": 0.9681,
+      "step": 29970
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00014068296525811838,
+      "loss": 0.9826,
+      "step": 29975
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.0001406600933119532,
+      "loss": 0.9846,
+      "step": 29980
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00014063721881716707,
+      "loss": 0.9939,
+      "step": 29985
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00014061434177519376,
+      "loss": 0.9893,
+      "step": 29990
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00014059146218746728,
+      "loss": 0.979,
+      "step": 29995
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00014056858005542173,
+      "loss": 0.9693,
+      "step": 30000
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00014054569538049134,
+      "loss": 1.0135,
+      "step": 30005
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001405228081641106,
+      "loss": 0.9971,
+      "step": 30010
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.0001404999184077141,
+      "loss": 1.0018,
+      "step": 30015
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00014047702611273658,
+      "loss": 0.9322,
+      "step": 30020
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.000140454131280613,
+      "loss": 0.9966,
+      "step": 30025
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00014043123391277836,
+      "loss": 1.0644,
+      "step": 30030
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00014040833401066793,
+      "loss": 1.1402,
+      "step": 30035
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00014038543157571712,
+      "loss": 0.9642,
+      "step": 30040
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00014036252660936142,
+      "loss": 0.9608,
+      "step": 30045
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00014033961911303665,
+      "loss": 0.9696,
+      "step": 30050
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00014031670908817856,
+      "loss": 0.9089,
+      "step": 30055
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00014029379653622326,
+      "loss": 0.993,
+      "step": 30060
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001402708814586069,
+      "loss": 0.8867,
+      "step": 30065
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001402479638567658,
+      "loss": 0.9627,
+      "step": 30070
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.0001402250437321365,
+      "loss": 0.9215,
+      "step": 30075
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00014020212108615564,
+      "loss": 0.9507,
+      "step": 30080
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00014017919592026006,
+      "loss": 0.9175,
+      "step": 30085
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001401562682358867,
+      "loss": 0.9142,
+      "step": 30090
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00014013333803447275,
+      "loss": 1.0266,
+      "step": 30095
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00014011040531745542,
+      "loss": 0.8784,
+      "step": 30100
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001400874700862722,
+      "loss": 0.8888,
+      "step": 30105
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00014006453234236075,
+      "loss": 1.0052,
+      "step": 30110
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00014004159208715874,
+      "loss": 0.9737,
+      "step": 30115
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.859375,
+      "learning_rate": 0.00014001864932210417,
+      "loss": 1.165,
+      "step": 30120
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00013999570404863503,
+      "loss": 1.0133,
+      "step": 30125
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00013997275626818965,
+      "loss": 0.9089,
+      "step": 30130
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001399498059822064,
+      "loss": 1.0132,
+      "step": 30135
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00013992685319212376,
+      "loss": 0.8784,
+      "step": 30140
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00013990389789938053,
+      "loss": 0.9172,
+      "step": 30145
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001398809401054155,
+      "loss": 1.0431,
+      "step": 30150
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00013985797981166774,
+      "loss": 1.0053,
+      "step": 30155
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001398350170195764,
+      "loss": 1.0129,
+      "step": 30160
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00013981205173058082,
+      "loss": 1.0763,
+      "step": 30165
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00013978908394612053,
+      "loss": 1.1376,
+      "step": 30170
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00013976611366763514,
+      "loss": 0.9536,
+      "step": 30175
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001397431408965644,
+      "loss": 1.0176,
+      "step": 30180
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00013972016563434838,
+      "loss": 1.008,
+      "step": 30185
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00013969718788242713,
+      "loss": 1.0744,
+      "step": 30190
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00013967420764224092,
+      "loss": 0.8735,
+      "step": 30195
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001396512249152302,
+      "loss": 1.1065,
+      "step": 30200
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00013962823970283553,
+      "loss": 0.9289,
+      "step": 30205
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00013960525200649765,
+      "loss": 1.0908,
+      "step": 30210
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00013958226182765753,
+      "loss": 1.0867,
+      "step": 30215
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001395592691677561,
+      "loss": 0.9663,
+      "step": 30220
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00013953627402823465,
+      "loss": 0.9168,
+      "step": 30225
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001395132764105345,
+      "loss": 0.8632,
+      "step": 30230
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00013949027631609718,
+      "loss": 0.9526,
+      "step": 30235
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001394672737463644,
+      "loss": 0.9847,
+      "step": 30240
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00013944426870277793,
+      "loss": 0.9977,
+      "step": 30245
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00013942126118677973,
+      "loss": 0.8099,
+      "step": 30250
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00013939825119981206,
+      "loss": 0.9267,
+      "step": 30255
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001393752387433171,
+      "loss": 0.9748,
+      "step": 30260
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00013935222381873728,
+      "loss": 1.0083,
+      "step": 30265
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00013932920642751535,
+      "loss": 0.9495,
+      "step": 30270
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.0001393061865710939,
+      "loss": 0.913,
+      "step": 30275
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00013928316425091593,
+      "loss": 0.9324,
+      "step": 30280
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00013926013946842449,
+      "loss": 0.8735,
+      "step": 30285
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00013923711222506277,
+      "loss": 0.903,
+      "step": 30290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00013921408252227422,
+      "loss": 0.935,
+      "step": 30295
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.78125,
+      "learning_rate": 0.00013919105036150226,
+      "loss": 1.0949,
+      "step": 30300
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00013916801574419068,
+      "loss": 0.8978,
+      "step": 30305
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00013914497867178322,
+      "loss": 1.004,
+      "step": 30310
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.953125,
+      "learning_rate": 0.00013912193914572391,
+      "loss": 1.0612,
+      "step": 30315
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00013909889716745693,
+      "loss": 0.9002,
+      "step": 30320
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00013907585273842656,
+      "loss": 0.8754,
+      "step": 30325
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00013905280586007718,
+      "loss": 0.9143,
+      "step": 30330
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0001390297565338535,
+      "loss": 0.6927,
+      "step": 30335
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00013900670476120022,
+      "loss": 0.9526,
+      "step": 30340
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00013898365054356226,
+      "loss": 0.9771,
+      "step": 30345
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00013896059388238466,
+      "loss": 0.9911,
+      "step": 30350
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00013893753477911268,
+      "loss": 0.9632,
+      "step": 30355
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001389144732351917,
+      "loss": 1.0988,
+      "step": 30360
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001388914092520672,
+      "loss": 0.9119,
+      "step": 30365
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013886834283118483,
+      "loss": 0.8508,
+      "step": 30370
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001388452739739905,
+      "loss": 0.9863,
+      "step": 30375
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001388222026819302,
+      "loss": 0.9309,
+      "step": 30380
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00013879912895644995,
+      "loss": 0.9081,
+      "step": 30385
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00013877605279899612,
+      "loss": 0.9678,
+      "step": 30390
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00013875297421101518,
+      "loss": 0.8776,
+      "step": 30395
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00013872989319395367,
+      "loss": 0.8552,
+      "step": 30400
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001387068097492583,
+      "loss": 0.9758,
+      "step": 30405
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00013868372387837604,
+      "loss": 0.9454,
+      "step": 30410
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00013866063558275394,
+      "loss": 1.0248,
+      "step": 30415
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00013863754486383913,
+      "loss": 1.0927,
+      "step": 30420
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00013861445172307904,
+      "loss": 0.9265,
+      "step": 30425
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001385913561619211,
+      "loss": 1.1097,
+      "step": 30430
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00013856825818181302,
+      "loss": 1.0002,
+      "step": 30435
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001385451577842026,
+      "loss": 1.1563,
+      "step": 30440
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00013852205497053775,
+      "loss": 0.8349,
+      "step": 30445
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00013849894974226666,
+      "loss": 0.9589,
+      "step": 30450
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00013847584210083754,
+      "loss": 1.0755,
+      "step": 30455
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00013845273204769884,
+      "loss": 0.9441,
+      "step": 30460
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00013842961958429907,
+      "loss": 1.0065,
+      "step": 30465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.000138406504712087,
+      "loss": 0.9354,
+      "step": 30470
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00013838338743251148,
+      "loss": 0.9546,
+      "step": 30475
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001383602677470215,
+      "loss": 0.9654,
+      "step": 30480
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00013833714565706626,
+      "loss": 0.9255,
+      "step": 30485
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00013831402116409506,
+      "loss": 0.9128,
+      "step": 30490
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00013829089426955736,
+      "loss": 0.8833,
+      "step": 30495
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001382677649749028,
+      "loss": 0.9549,
+      "step": 30500
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001382446332815812,
+      "loss": 0.9081,
+      "step": 30505
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001382214991910424,
+      "loss": 0.9331,
+      "step": 30510
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001381983627047365,
+      "loss": 0.9937,
+      "step": 30515
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001381752238241137,
+      "loss": 0.8444,
+      "step": 30520
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001381520825506244,
+      "loss": 1.0127,
+      "step": 30525
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00013812893888571917,
+      "loss": 0.9867,
+      "step": 30530
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00013810579283084855,
+      "loss": 0.9169,
+      "step": 30535
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00013808264438746346,
+      "loss": 0.915,
+      "step": 30540
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00013805949355701487,
+      "loss": 0.9577,
+      "step": 30545
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00013803634034095384,
+      "loss": 0.9792,
+      "step": 30550
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00013801318474073167,
+      "loss": 0.9618,
+      "step": 30555
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00013799002675779983,
+      "loss": 0.8944,
+      "step": 30560
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00013796686639360982,
+      "loss": 0.919,
+      "step": 30565
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00013794370364961342,
+      "loss": 1.0096,
+      "step": 30570
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.69921875,
+      "learning_rate": 0.00013792053852726242,
+      "loss": 1.1821,
+      "step": 30575
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00013789737102800888,
+      "loss": 1.1147,
+      "step": 30580
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.000137874201153305,
+      "loss": 1.0756,
+      "step": 30585
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00013785102890460303,
+      "loss": 0.9199,
+      "step": 30590
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00013782785428335546,
+      "loss": 0.9286,
+      "step": 30595
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001378046772910149,
+      "loss": 0.9718,
+      "step": 30600
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.83984375,
+      "learning_rate": 0.0001377814979290341,
+      "loss": 1.0316,
+      "step": 30605
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00013775831619886603,
+      "loss": 0.9671,
+      "step": 30610
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001377351321019636,
+      "loss": 0.9524,
+      "step": 30615
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00013771194563978024,
+      "loss": 0.9654,
+      "step": 30620
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013768875681376908,
+      "loss": 0.9857,
+      "step": 30625
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00013766556562538375,
+      "loss": 1.0038,
+      "step": 30630
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001376423720760779,
+      "loss": 0.9687,
+      "step": 30635
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00013761917616730523,
+      "loss": 0.9222,
+      "step": 30640
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00013759597790051982,
+      "loss": 0.9985,
+      "step": 30645
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00013757277727717565,
+      "loss": 1.0839,
+      "step": 30650
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.000137549574298727,
+      "loss": 0.9799,
+      "step": 30655
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001375263689666283,
+      "loss": 0.9521,
+      "step": 30660
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00013750316128233406,
+      "loss": 0.9973,
+      "step": 30665
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00013747995124729892,
+      "loss": 0.8904,
+      "step": 30670
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00013745673886297782,
+      "loss": 0.9553,
+      "step": 30675
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001374335241308256,
+      "loss": 1.0507,
+      "step": 30680
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00013741030705229746,
+      "loss": 0.9111,
+      "step": 30685
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013738708762884872,
+      "loss": 0.9156,
+      "step": 30690
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00013736386586193472,
+      "loss": 1.0744,
+      "step": 30695
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.000137340641753011,
+      "loss": 1.0734,
+      "step": 30700
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001373174153035334,
+      "loss": 0.8312,
+      "step": 30705
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00013729418651495766,
+      "loss": 0.829,
+      "step": 30710
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00013727095538873987,
+      "loss": 0.8859,
+      "step": 30715
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00013724772192633616,
+      "loss": 0.8469,
+      "step": 30720
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00013722448612920278,
+      "loss": 0.9399,
+      "step": 30725
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001372012479987962,
+      "loss": 0.8732,
+      "step": 30730
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00013717800753657307,
+      "loss": 0.8908,
+      "step": 30735
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001371547647439901,
+      "loss": 0.9621,
+      "step": 30740
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00013713151962250412,
+      "loss": 1.0011,
+      "step": 30745
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001371082721735722,
+      "loss": 1.0021,
+      "step": 30750
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00013708502239865152,
+      "loss": 0.9714,
+      "step": 30755
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00013706177029919943,
+      "loss": 1.1514,
+      "step": 30760
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001370385158766733,
+      "loss": 0.8423,
+      "step": 30765
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00013701525913253088,
+      "loss": 0.9267,
+      "step": 30770
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00013699200006822985,
+      "loss": 0.9194,
+      "step": 30775
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001369687386852281,
+      "loss": 0.9124,
+      "step": 30780
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00013694547498498372,
+      "loss": 0.9982,
+      "step": 30785
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001369222089689549,
+      "loss": 0.8867,
+      "step": 30790
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00013689894063859996,
+      "loss": 0.9391,
+      "step": 30795
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001368756699953774,
+      "loss": 0.7938,
+      "step": 30800
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00013685239704074582,
+      "loss": 1.0281,
+      "step": 30805
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00013682912177616404,
+      "loss": 1.002,
+      "step": 30810
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00013680584420309097,
+      "loss": 0.8366,
+      "step": 30815
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00013678256432298567,
+      "loss": 0.981,
+      "step": 30820
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00013675928213730736,
+      "loss": 1.0917,
+      "step": 30825
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00013673599764751535,
+      "loss": 0.9331,
+      "step": 30830
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00013671271085506916,
+      "loss": 0.9488,
+      "step": 30835
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00013668942176142847,
+      "loss": 0.9893,
+      "step": 30840
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00013666613036805302,
+      "loss": 0.8505,
+      "step": 30845
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0001366428366764028,
+      "loss": 1.0239,
+      "step": 30850
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001366195406879378,
+      "loss": 0.9631,
+      "step": 30855
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00013659624240411827,
+      "loss": 0.8366,
+      "step": 30860
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00013657294182640463,
+      "loss": 0.9888,
+      "step": 30865
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001365496389562573,
+      "loss": 0.9006,
+      "step": 30870
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00013652633379513703,
+      "loss": 0.9934,
+      "step": 30875
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00013650302634450454,
+      "loss": 0.9593,
+      "step": 30880
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00013647971660582075,
+      "loss": 1.047,
+      "step": 30885
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001364564045805468,
+      "loss": 0.9426,
+      "step": 30890
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00013643309027014389,
+      "loss": 1.0609,
+      "step": 30895
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001364097736760734,
+      "loss": 0.8709,
+      "step": 30900
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001363864547997968,
+      "loss": 0.9918,
+      "step": 30905
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001363631336427758,
+      "loss": 0.8467,
+      "step": 30910
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00013633981020647214,
+      "loss": 0.9783,
+      "step": 30915
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0001363164844923478,
+      "loss": 1.0742,
+      "step": 30920
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00013629315650186488,
+      "loss": 0.929,
+      "step": 30925
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00013626982623648554,
+      "loss": 0.8576,
+      "step": 30930
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001362464936976722,
+      "loss": 0.873,
+      "step": 30935
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013622315888688736,
+      "loss": 0.9349,
+      "step": 30940
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00013619982180559364,
+      "loss": 0.9348,
+      "step": 30945
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00013617648245525392,
+      "loss": 0.8357,
+      "step": 30950
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00013615314083733102,
+      "loss": 0.9747,
+      "step": 30955
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00013612979695328808,
+      "loss": 0.9304,
+      "step": 30960
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00013610645080458833,
+      "loss": 0.9786,
+      "step": 30965
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00013608310239269513,
+      "loss": 0.8328,
+      "step": 30970
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00013605975171907195,
+      "loss": 0.8963,
+      "step": 30975
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.0001360363987851825,
+      "loss": 1.0268,
+      "step": 30980
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00013601304359249053,
+      "loss": 0.9399,
+      "step": 30985
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00013598968614245995,
+      "loss": 0.9845,
+      "step": 30990
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00013596632643655485,
+      "loss": 0.7482,
+      "step": 30995
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00013594296447623947,
+      "loss": 0.95,
+      "step": 31000
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00013591960026297813,
+      "loss": 0.8751,
+      "step": 31005
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00013589623379823532,
+      "loss": 0.85,
+      "step": 31010
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00013587286508347574,
+      "loss": 0.8601,
+      "step": 31015
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001358494941201641,
+      "loss": 0.7568,
+      "step": 31020
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9921875,
+      "learning_rate": 0.0001358261209097653,
+      "loss": 0.9786,
+      "step": 31025
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001358027454537445,
+      "loss": 0.9896,
+      "step": 31030
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00013577936775356676,
+      "loss": 1.0181,
+      "step": 31035
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0001357559878106976,
+      "loss": 0.8901,
+      "step": 31040
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00013573260562660232,
+      "loss": 1.1547,
+      "step": 31045
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013570922120274666,
+      "loss": 1.0983,
+      "step": 31050
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00013568583454059632,
+      "loss": 0.9642,
+      "step": 31055
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00013566244564161727,
+      "loss": 0.9549,
+      "step": 31060
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00013563905450727546,
+      "loss": 1.0332,
+      "step": 31065
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013561566113903716,
+      "loss": 1.0243,
+      "step": 31070
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00013559226553836866,
+      "loss": 0.9611,
+      "step": 31075
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00013556886770673638,
+      "loss": 0.9881,
+      "step": 31080
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00013554546764560702,
+      "loss": 0.8778,
+      "step": 31085
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00013552206535644723,
+      "loss": 0.9477,
+      "step": 31090
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00013549866084072392,
+      "loss": 0.8397,
+      "step": 31095
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00013547525409990415,
+      "loss": 0.9632,
+      "step": 31100
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.000135451845135455,
+      "loss": 0.7316,
+      "step": 31105
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001354284339488439,
+      "loss": 1.0173,
+      "step": 31110
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00013540502054153815,
+      "loss": 0.8948,
+      "step": 31115
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001353816049150054,
+      "loss": 0.8243,
+      "step": 31120
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00013535818707071336,
+      "loss": 0.8822,
+      "step": 31125
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001353347670101299,
+      "loss": 0.958,
+      "step": 31130
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00013531134473472297,
+      "loss": 0.906,
+      "step": 31135
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00013528792024596075,
+      "loss": 0.9097,
+      "step": 31140
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001352644935453115,
+      "loss": 1.0661,
+      "step": 31145
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00013524106463424365,
+      "loss": 0.9701,
+      "step": 31150
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00013521763351422573,
+      "loss": 1.0976,
+      "step": 31155
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001351942001867264,
+      "loss": 1.0888,
+      "step": 31160
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00013517076465321455,
+      "loss": 0.9993,
+      "step": 31165
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00013514732691515907,
+      "loss": 1.0969,
+      "step": 31170
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00013512388697402912,
+      "loss": 0.9507,
+      "step": 31175
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00013510044483129398,
+      "loss": 1.1196,
+      "step": 31180
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00013507700048842296,
+      "loss": 0.8669,
+      "step": 31185
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00013505355394688562,
+      "loss": 1.0395,
+      "step": 31190
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00013503010520815156,
+      "loss": 1.0378,
+      "step": 31195
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001350066542736906,
+      "loss": 1.0775,
+      "step": 31200
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00013498320114497273,
+      "loss": 0.8533,
+      "step": 31205
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00013495974582346793,
+      "loss": 0.9821,
+      "step": 31210
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00013493628831064646,
+      "loss": 0.9691,
+      "step": 31215
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00013491282860797864,
+      "loss": 0.9099,
+      "step": 31220
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00013488936671693496,
+      "loss": 0.9978,
+      "step": 31225
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00013486590263898603,
+      "loss": 0.9646,
+      "step": 31230
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00013484243637560265,
+      "loss": 0.8978,
+      "step": 31235
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00013481896792825565,
+      "loss": 0.8367,
+      "step": 31240
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001347954972984161,
+      "loss": 0.9076,
+      "step": 31245
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00013477202448755515,
+      "loss": 1.0487,
+      "step": 31250
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00013474854949714407,
+      "loss": 0.7774,
+      "step": 31255
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00013472507232865436,
+      "loss": 1.0968,
+      "step": 31260
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00013470159298355758,
+      "loss": 1.0582,
+      "step": 31265
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001346781114633254,
+      "loss": 1.0763,
+      "step": 31270
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00013465462776942973,
+      "loss": 1.0131,
+      "step": 31275
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001346311419033425,
+      "loss": 0.9431,
+      "step": 31280
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00013460765386653583,
+      "loss": 0.9907,
+      "step": 31285
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00013458416366048204,
+      "loss": 0.9854,
+      "step": 31290
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00013456067128665347,
+      "loss": 1.0112,
+      "step": 31295
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001345371767465227,
+      "loss": 1.0002,
+      "step": 31300
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00013451368004156232,
+      "loss": 0.9572,
+      "step": 31305
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00013449018117324516,
+      "loss": 0.9806,
+      "step": 31310
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00013446668014304418,
+      "loss": 0.9183,
+      "step": 31315
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00013444317695243243,
+      "loss": 0.8783,
+      "step": 31320
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001344196716028831,
+      "loss": 1.051,
+      "step": 31325
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001343961640958696,
+      "loss": 0.8848,
+      "step": 31330
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001343726544328653,
+      "loss": 0.9742,
+      "step": 31335
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001343491426153439,
+      "loss": 1.0643,
+      "step": 31340
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00013432562864477916,
+      "loss": 0.8818,
+      "step": 31345
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001343021125226449,
+      "loss": 0.9056,
+      "step": 31350
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00013427859425041514,
+      "loss": 0.929,
+      "step": 31355
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00013425507382956405,
+      "loss": 0.8192,
+      "step": 31360
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00013423155126156592,
+      "loss": 1.0017,
+      "step": 31365
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001342080265478952,
+      "loss": 0.981,
+      "step": 31370
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00013418449969002636,
+      "loss": 0.9689,
+      "step": 31375
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013416097068943422,
+      "loss": 0.8402,
+      "step": 31380
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0001341374395475935,
+      "loss": 0.9719,
+      "step": 31385
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00013411390626597917,
+      "loss": 1.0305,
+      "step": 31390
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00013409037084606635,
+      "loss": 0.9859,
+      "step": 31395
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001340668332893303,
+      "loss": 1.1799,
+      "step": 31400
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001340432935972463,
+      "loss": 1.0158,
+      "step": 31405
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001340197517712899,
+      "loss": 0.8952,
+      "step": 31410
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001339962078129367,
+      "loss": 0.9101,
+      "step": 31415
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00013397266172366246,
+      "loss": 1.0594,
+      "step": 31420
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00013394911350494314,
+      "loss": 0.8013,
+      "step": 31425
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001339255631582547,
+      "loss": 0.9715,
+      "step": 31430
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001339020106850733,
+      "loss": 0.9856,
+      "step": 31435
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00013387845608687528,
+      "loss": 0.9213,
+      "step": 31440
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00013385489936513703,
+      "loss": 1.1167,
+      "step": 31445
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00013383134052133517,
+      "loss": 0.9305,
+      "step": 31450
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013380777955694634,
+      "loss": 0.9467,
+      "step": 31455
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00013378421647344737,
+      "loss": 0.8589,
+      "step": 31460
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00013376065127231523,
+      "loss": 0.9975,
+      "step": 31465
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00013373708395502705,
+      "loss": 0.9272,
+      "step": 31470
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00013371351452306,
+      "loss": 1.0828,
+      "step": 31475
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00013368994297789145,
+      "loss": 0.9316,
+      "step": 31480
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013366636932099892,
+      "loss": 1.1062,
+      "step": 31485
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00013364279355386003,
+      "loss": 0.975,
+      "step": 31490
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001336192156779525,
+      "loss": 0.8687,
+      "step": 31495
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00013359563569475424,
+      "loss": 1.0445,
+      "step": 31500
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001335720536057433,
+      "loss": 0.8226,
+      "step": 31505
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001335484694123978,
+      "loss": 0.8799,
+      "step": 31510
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00013352488311619602,
+      "loss": 0.9934,
+      "step": 31515
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00013350129471861638,
+      "loss": 0.7989,
+      "step": 31520
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00013347770422113744,
+      "loss": 1.0236,
+      "step": 31525
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00013345411162523786,
+      "loss": 0.9755,
+      "step": 31530
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00013343051693239647,
+      "loss": 1.1008,
+      "step": 31535
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00013340692014409222,
+      "loss": 0.9241,
+      "step": 31540
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00013338332126180415,
+      "loss": 1.1343,
+      "step": 31545
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00013335972028701148,
+      "loss": 0.9986,
+      "step": 31550
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00013333611722119357,
+      "loss": 0.9672,
+      "step": 31555
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00013331251206582983,
+      "loss": 1.0488,
+      "step": 31560
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00013328890482239994,
+      "loss": 1.0026,
+      "step": 31565
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00013326529549238352,
+      "loss": 1.0177,
+      "step": 31570
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00013324168407726056,
+      "loss": 0.9506,
+      "step": 31575
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00013321807057851094,
+      "loss": 0.9342,
+      "step": 31580
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00013319445499761486,
+      "loss": 1.0342,
+      "step": 31585
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00013317083733605252,
+      "loss": 1.0099,
+      "step": 31590
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001331472175953043,
+      "loss": 0.9359,
+      "step": 31595
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013312359577685072,
+      "loss": 0.971,
+      "step": 31600
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00013309997188217247,
+      "loss": 0.93,
+      "step": 31605
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00013307634591275028,
+      "loss": 0.9013,
+      "step": 31610
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00013305271787006503,
+      "loss": 0.9941,
+      "step": 31615
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001330290877555978,
+      "loss": 1.1095,
+      "step": 31620
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001330054555708297,
+      "loss": 0.9006,
+      "step": 31625
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00013298182131724211,
+      "loss": 0.9277,
+      "step": 31630
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00013295818499631636,
+      "loss": 1.009,
+      "step": 31635
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00013293454660953403,
+      "loss": 0.9823,
+      "step": 31640
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00013291090615837685,
+      "loss": 1.0153,
+      "step": 31645
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00013288726364432652,
+      "loss": 0.8506,
+      "step": 31650
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00013286361906886512,
+      "loss": 1.0333,
+      "step": 31655
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00013283997243347464,
+      "loss": 1.1103,
+      "step": 31660
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00013281632373963727,
+      "loss": 0.9147,
+      "step": 31665
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00013279267298883535,
+      "loss": 0.9931,
+      "step": 31670
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00013276902018255132,
+      "loss": 1.0037,
+      "step": 31675
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001327453653222678,
+      "loss": 0.9983,
+      "step": 31680
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00013272170840946754,
+      "loss": 1.1003,
+      "step": 31685
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00013269804944563327,
+      "loss": 1.095,
+      "step": 31690
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00013267438843224803,
+      "loss": 1.0487,
+      "step": 31695
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00013265072537079492,
+      "loss": 1.1067,
+      "step": 31700
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00013262706026275712,
+      "loss": 0.8894,
+      "step": 31705
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00013260339310961806,
+      "loss": 0.8737,
+      "step": 31710
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0001325797239128612,
+      "loss": 0.7964,
+      "step": 31715
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00013255605267397012,
+      "loss": 0.8535,
+      "step": 31720
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.0001325323793944286,
+      "loss": 1.0003,
+      "step": 31725
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00013250870407572045,
+      "loss": 0.8312,
+      "step": 31730
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00013248502671932971,
+      "loss": 1.0002,
+      "step": 31735
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00013246134732674056,
+      "loss": 0.912,
+      "step": 31740
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001324376658994371,
+      "loss": 0.928,
+      "step": 31745
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00013241398243890386,
+      "loss": 1.0814,
+      "step": 31750
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00013239029694662527,
+      "loss": 0.7331,
+      "step": 31755
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00013236660942408596,
+      "loss": 0.9826,
+      "step": 31760
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00013234291987277076,
+      "loss": 0.8716,
+      "step": 31765
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00013231922829416443,
+      "loss": 0.8813,
+      "step": 31770
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001322955346897521,
+      "loss": 1.0243,
+      "step": 31775
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001322718390610189,
+      "loss": 0.9718,
+      "step": 31780
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00013224814140945003,
+      "loss": 0.838,
+      "step": 31785
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00013222444173653097,
+      "loss": 0.8757,
+      "step": 31790
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001322007400437472,
+      "loss": 0.9052,
+      "step": 31795
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00013217703633258433,
+      "loss": 0.9327,
+      "step": 31800
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013215333060452816,
+      "loss": 0.9738,
+      "step": 31805
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00013212962286106468,
+      "loss": 0.8292,
+      "step": 31810
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00013210591310367978,
+      "loss": 0.9225,
+      "step": 31815
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00013208220133385974,
+      "loss": 1.0917,
+      "step": 31820
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00013205848755309073,
+      "loss": 0.9889,
+      "step": 31825
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00013203477176285924,
+      "loss": 0.9825,
+      "step": 31830
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00013201105396465178,
+      "loss": 1.0294,
+      "step": 31835
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.75,
+      "learning_rate": 0.00013198733415995494,
+      "loss": 1.1127,
+      "step": 31840
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00013196361235025562,
+      "loss": 0.8949,
+      "step": 31845
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00013193988853704068,
+      "loss": 0.845,
+      "step": 31850
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00013191616272179713,
+      "loss": 0.8679,
+      "step": 31855
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00013189243490601215,
+      "loss": 0.9756,
+      "step": 31860
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00013186870509117302,
+      "loss": 1.0327,
+      "step": 31865
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00013184497327876717,
+      "loss": 0.9826,
+      "step": 31870
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00013182123947028216,
+      "loss": 0.7791,
+      "step": 31875
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00013179750366720556,
+      "loss": 0.9861,
+      "step": 31880
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00013177376587102522,
+      "loss": 0.9623,
+      "step": 31885
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001317500260832291,
+      "loss": 0.9592,
+      "step": 31890
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00013172628430530513,
+      "loss": 0.9209,
+      "step": 31895
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00013170254053874157,
+      "loss": 0.7683,
+      "step": 31900
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00013167879478502665,
+      "loss": 0.9936,
+      "step": 31905
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00013165504704564876,
+      "loss": 1.0267,
+      "step": 31910
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00013163129732209652,
+      "loss": 0.9512,
+      "step": 31915
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001316075456158585,
+      "loss": 0.9878,
+      "step": 31920
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00013158379192842353,
+      "loss": 1.0096,
+      "step": 31925
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00013156003626128054,
+      "loss": 0.9902,
+      "step": 31930
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00013153627861591847,
+      "loss": 0.7962,
+      "step": 31935
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00013151251899382662,
+      "loss": 0.873,
+      "step": 31940
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00013148875739649413,
+      "loss": 0.9694,
+      "step": 31945
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.00013146499382541048,
+      "loss": 1.0138,
+      "step": 31950
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00013144122828206523,
+      "loss": 0.9132,
+      "step": 31955
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00013141746076794793,
+      "loss": 0.9452,
+      "step": 31960
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001313936912845484,
+      "loss": 1.0403,
+      "step": 31965
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00013136991983335656,
+      "loss": 0.9956,
+      "step": 31970
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00013134614641586244,
+      "loss": 0.8793,
+      "step": 31975
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00013132237103355613,
+      "loss": 0.9511,
+      "step": 31980
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00013129859368792794,
+      "loss": 0.9065,
+      "step": 31985
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00013127481438046824,
+      "loss": 0.9021,
+      "step": 31990
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00013125103311266756,
+      "loss": 0.982,
+      "step": 31995
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00013122724988601656,
+      "loss": 0.8726,
+      "step": 32000
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00013120346470200594,
+      "loss": 0.845,
+      "step": 32005
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00013117967756212667,
+      "loss": 1.1373,
+      "step": 32010
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00013115588846786963,
+      "loss": 0.9724,
+      "step": 32015
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00013113209742072606,
+      "loss": 1.0094,
+      "step": 32020
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00013110830442218714,
+      "loss": 0.9913,
+      "step": 32025
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001310845094737443,
+      "loss": 0.919,
+      "step": 32030
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00013106071257688897,
+      "loss": 1.0586,
+      "step": 32035
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001310369137331128,
+      "loss": 1.0326,
+      "step": 32040
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00013101311294390756,
+      "loss": 0.996,
+      "step": 32045
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00013098931021076506,
+      "loss": 1.1005,
+      "step": 32050
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00013096550553517734,
+      "loss": 0.9351,
+      "step": 32055
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001309416989186364,
+      "loss": 0.8718,
+      "step": 32060
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001309178903626346,
+      "loss": 0.9508,
+      "step": 32065
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00013089407986866414,
+      "loss": 0.9767,
+      "step": 32070
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00013087026743821765,
+      "loss": 0.9856,
+      "step": 32075
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00013084645307278762,
+      "loss": 0.9379,
+      "step": 32080
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00013082263677386674,
+      "loss": 0.9073,
+      "step": 32085
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00013079881854294792,
+      "loss": 0.7946,
+      "step": 32090
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013077499838152405,
+      "loss": 0.9327,
+      "step": 32095
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00013075117629108825,
+      "loss": 0.963,
+      "step": 32100
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001307273522731337,
+      "loss": 1.0912,
+      "step": 32105
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.75,
+      "learning_rate": 0.0001307035263291537,
+      "loss": 1.0582,
+      "step": 32110
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001306796984606417,
+      "loss": 1.0168,
+      "step": 32115
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00013065586866909128,
+      "loss": 0.9312,
+      "step": 32120
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00013063203695599606,
+      "loss": 1.0982,
+      "step": 32125
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.09375,
+      "learning_rate": 0.0001306082033228499,
+      "loss": 1.0406,
+      "step": 32130
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00013058436777114673,
+      "loss": 1.037,
+      "step": 32135
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001305605303023805,
+      "loss": 0.8963,
+      "step": 32140
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00013053669091804546,
+      "loss": 0.9866,
+      "step": 32145
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00013051284961963585,
+      "loss": 1.0073,
+      "step": 32150
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00013048900640864606,
+      "loss": 1.0038,
+      "step": 32155
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00013046516128657065,
+      "loss": 1.0067,
+      "step": 32160
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00013044131425490418,
+      "loss": 0.9225,
+      "step": 32165
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001304174653151415,
+      "loss": 0.8713,
+      "step": 32170
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00013039361446877745,
+      "loss": 0.8281,
+      "step": 32175
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.000130369761717307,
+      "loss": 0.9131,
+      "step": 32180
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00013034590706222538,
+      "loss": 0.891,
+      "step": 32185
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001303220505050277,
+      "loss": 1.1046,
+      "step": 32190
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00013029819204720932,
+      "loss": 1.0849,
+      "step": 32195
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001302743316902658,
+      "loss": 0.9852,
+      "step": 32200
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00013025046943569268,
+      "loss": 0.9682,
+      "step": 32205
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00013022660528498568,
+      "loss": 0.937,
+      "step": 32210
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00013020273923964064,
+      "loss": 0.8642,
+      "step": 32215
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00013017887130115349,
+      "loss": 0.8892,
+      "step": 32220
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00013015500147102032,
+      "loss": 0.9228,
+      "step": 32225
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00013013112975073733,
+      "loss": 1.0077,
+      "step": 32230
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001301072561418008,
+      "loss": 1.0031,
+      "step": 32235
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00013008338064570717,
+      "loss": 0.9054,
+      "step": 32240
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.000130059503263953,
+      "loss": 1.0037,
+      "step": 32245
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00013003562399803488,
+      "loss": 0.8962,
+      "step": 32250
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00013001174284944968,
+      "loss": 0.8093,
+      "step": 32255
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00012998785981969423,
+      "loss": 0.9555,
+      "step": 32260
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00012996397491026558,
+      "loss": 0.9951,
+      "step": 32265
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001299400881226609,
+      "loss": 0.764,
+      "step": 32270
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00012991619945837735,
+      "loss": 0.9576,
+      "step": 32275
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00012989230891891236,
+      "loss": 0.8473,
+      "step": 32280
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001298684165057634,
+      "loss": 0.803,
+      "step": 32285
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001298445222204281,
+      "loss": 0.8959,
+      "step": 32290
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00012982062606440412,
+      "loss": 1.0361,
+      "step": 32295
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012979672803918938,
+      "loss": 0.9159,
+      "step": 32300
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00012977282814628172,
+      "loss": 1.0206,
+      "step": 32305
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00012974892638717932,
+      "loss": 0.9815,
+      "step": 32310
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00012972502276338034,
+      "loss": 1.0164,
+      "step": 32315
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001297011172763831,
+      "loss": 1.0121,
+      "step": 32320
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00012967720992768596,
+      "loss": 1.0233,
+      "step": 32325
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00012965330071878752,
+      "loss": 1.0093,
+      "step": 32330
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012962938965118643,
+      "loss": 0.935,
+      "step": 32335
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00012960547672638144,
+      "loss": 0.9044,
+      "step": 32340
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00012958156194587146,
+      "loss": 1.0922,
+      "step": 32345
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012955764531115548,
+      "loss": 0.9507,
+      "step": 32350
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012953372682373264,
+      "loss": 0.9014,
+      "step": 32355
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012950980648510213,
+      "loss": 1.078,
+      "step": 32360
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012948588429676335,
+      "loss": 1.0288,
+      "step": 32365
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00012946196026021578,
+      "loss": 1.0478,
+      "step": 32370
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.000129438034376959,
+      "loss": 0.9876,
+      "step": 32375
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001294141066484927,
+      "loss": 0.7554,
+      "step": 32380
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00012939017707631664,
+      "loss": 0.9796,
+      "step": 32385
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00012936624566193086,
+      "loss": 1.0194,
+      "step": 32390
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00012934231240683536,
+      "loss": 0.8431,
+      "step": 32395
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012931837731253027,
+      "loss": 0.9893,
+      "step": 32400
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001292944403805159,
+      "loss": 1.1741,
+      "step": 32405
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012927050161229265,
+      "loss": 0.6753,
+      "step": 32410
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012924656100936103,
+      "loss": 0.9545,
+      "step": 32415
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001292226185732217,
+      "loss": 1.087,
+      "step": 32420
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012919867430537525,
+      "loss": 1.0627,
+      "step": 32425
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00012917472820732272,
+      "loss": 1.1009,
+      "step": 32430
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012915078028056498,
+      "loss": 0.9163,
+      "step": 32435
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00012912683052660313,
+      "loss": 0.922,
+      "step": 32440
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00012910287894693834,
+      "loss": 0.8632,
+      "step": 32445
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.000129078925543072,
+      "loss": 1.0744,
+      "step": 32450
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012905497031650548,
+      "loss": 0.9194,
+      "step": 32455
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00012903101326874032,
+      "loss": 1.0821,
+      "step": 32460
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00012900705440127818,
+      "loss": 0.9133,
+      "step": 32465
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00012898309371562084,
+      "loss": 0.9024,
+      "step": 32470
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001289591312132702,
+      "loss": 1.0314,
+      "step": 32475
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001289351668957282,
+      "loss": 1.0295,
+      "step": 32480
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012891120076449699,
+      "loss": 0.9198,
+      "step": 32485
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001288872328210788,
+      "loss": 1.0714,
+      "step": 32490
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00012886326306697595,
+      "loss": 0.9769,
+      "step": 32495
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00012883929150369093,
+      "loss": 1.016,
+      "step": 32500
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001288153181327262,
+      "loss": 1.0573,
+      "step": 32505
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00012879134295558457,
+      "loss": 1.1267,
+      "step": 32510
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00012876736597376874,
+      "loss": 1.0102,
+      "step": 32515
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00012874338718878167,
+      "loss": 0.9099,
+      "step": 32520
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012871940660212636,
+      "loss": 1.0459,
+      "step": 32525
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00012869542421530594,
+      "loss": 1.0324,
+      "step": 32530
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001286714400298236,
+      "loss": 0.8854,
+      "step": 32535
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001286474540471828,
+      "loss": 1.0234,
+      "step": 32540
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012862346626888694,
+      "loss": 1.0255,
+      "step": 32545
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012859947669643958,
+      "loss": 0.9126,
+      "step": 32550
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00012857548533134452,
+      "loss": 1.0146,
+      "step": 32555
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012855149217510544,
+      "loss": 0.852,
+      "step": 32560
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001285274972292263,
+      "loss": 0.8511,
+      "step": 32565
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001285035004952112,
+      "loss": 1.1151,
+      "step": 32570
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012847950197456416,
+      "loss": 1.0862,
+      "step": 32575
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00012845550166878957,
+      "loss": 0.9294,
+      "step": 32580
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001284314995793917,
+      "loss": 1.1179,
+      "step": 32585
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.000128407495707875,
+      "loss": 0.9428,
+      "step": 32590
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00012838349005574417,
+      "loss": 1.0087,
+      "step": 32595
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012835948262450385,
+      "loss": 0.8835,
+      "step": 32600
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00012833547341565887,
+      "loss": 1.0477,
+      "step": 32605
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00012831146243071415,
+      "loss": 1.0284,
+      "step": 32610
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00012828744967117469,
+      "loss": 1.0356,
+      "step": 32615
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00012826343513854568,
+      "loss": 1.0029,
+      "step": 32620
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00012823941883433236,
+      "loss": 0.9734,
+      "step": 32625
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00012821540076004016,
+      "loss": 1.0066,
+      "step": 32630
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012819138091717445,
+      "loss": 0.9152,
+      "step": 32635
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00012816735930724088,
+      "loss": 0.8258,
+      "step": 32640
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00012814333593174515,
+      "loss": 1.0175,
+      "step": 32645
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00012811931079219309,
+      "loss": 1.1105,
+      "step": 32650
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00012809528389009058,
+      "loss": 0.878,
+      "step": 32655
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001280712552269437,
+      "loss": 0.9104,
+      "step": 32660
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012804722480425857,
+      "loss": 1.0456,
+      "step": 32665
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00012802319262354142,
+      "loss": 0.8883,
+      "step": 32670
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00012799915868629867,
+      "loss": 1.0246,
+      "step": 32675
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00012797512299403673,
+      "loss": 0.9482,
+      "step": 32680
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00012795108554826228,
+      "loss": 0.9019,
+      "step": 32685
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001279270463504819,
+      "loss": 0.963,
+      "step": 32690
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001279030054022025,
+      "loss": 0.8933,
+      "step": 32695
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00012787896270493088,
+      "loss": 1.0861,
+      "step": 32700
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00012785491826017414,
+      "loss": 0.9753,
+      "step": 32705
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00012783087206943942,
+      "loss": 0.9748,
+      "step": 32710
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00012780682413423395,
+      "loss": 0.9589,
+      "step": 32715
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00012778277445606506,
+      "loss": 0.805,
+      "step": 32720
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00012775872303644021,
+      "loss": 1.01,
+      "step": 32725
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.000127734669876867,
+      "loss": 1.0963,
+      "step": 32730
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00012771061497885312,
+      "loss": 0.9843,
+      "step": 32735
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001276865583439063,
+      "loss": 0.9517,
+      "step": 32740
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00012766249997353448,
+      "loss": 0.8473,
+      "step": 32745
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00012763843986924564,
+      "loss": 0.919,
+      "step": 32750
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00012761437803254793,
+      "loss": 0.949,
+      "step": 32755
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00012759031446494957,
+      "loss": 0.9639,
+      "step": 32760
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012756624916795885,
+      "loss": 0.9237,
+      "step": 32765
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00012754218214308427,
+      "loss": 0.947,
+      "step": 32770
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0001275181133918343,
+      "loss": 0.9556,
+      "step": 32775
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00012749404291571766,
+      "loss": 1.0168,
+      "step": 32780
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001274699707162431,
+      "loss": 0.9174,
+      "step": 32785
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001274458967949195,
+      "loss": 0.8608,
+      "step": 32790
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00012742182115325584,
+      "loss": 1.1732,
+      "step": 32795
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00012739774379276117,
+      "loss": 1.1252,
+      "step": 32800
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.00012737366471494472,
+      "loss": 1.0629,
+      "step": 32805
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001273495839213158,
+      "loss": 1.0418,
+      "step": 32810
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001273255014133838,
+      "loss": 0.868,
+      "step": 32815
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012730141719265826,
+      "loss": 0.9079,
+      "step": 32820
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001272773312606488,
+      "loss": 0.9518,
+      "step": 32825
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00012725324361886515,
+      "loss": 0.9046,
+      "step": 32830
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00012722915426881715,
+      "loss": 0.9271,
+      "step": 32835
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00012720506321201472,
+      "loss": 0.9904,
+      "step": 32840
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00012718097044996798,
+      "loss": 0.9101,
+      "step": 32845
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00012715687598418706,
+      "loss": 0.937,
+      "step": 32850
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012713277981618218,
+      "loss": 1.0701,
+      "step": 32855
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001271086819474638,
+      "loss": 0.8446,
+      "step": 32860
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00012708458237954234,
+      "loss": 0.926,
+      "step": 32865
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00012706048111392845,
+      "loss": 1.0323,
+      "step": 32870
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012703637815213273,
+      "loss": 0.9371,
+      "step": 32875
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00012701227349566608,
+      "loss": 0.8327,
+      "step": 32880
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00012698816714603933,
+      "loss": 1.0061,
+      "step": 32885
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001269640591047635,
+      "loss": 0.91,
+      "step": 32890
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00012693994937334983,
+      "loss": 0.823,
+      "step": 32895
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001269158379533094,
+      "loss": 0.9868,
+      "step": 32900
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00012689172484615357,
+      "loss": 0.9551,
+      "step": 32905
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00012686761005339383,
+      "loss": 0.8869,
+      "step": 32910
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00012684349357654166,
+      "loss": 0.941,
+      "step": 32915
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00012681937541710877,
+      "loss": 0.9691,
+      "step": 32920
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001267952555766069,
+      "loss": 0.8838,
+      "step": 32925
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00012677113405654784,
+      "loss": 0.8744,
+      "step": 32930
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00012674701085844365,
+      "loss": 1.0615,
+      "step": 32935
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00012672288598380632,
+      "loss": 0.9642,
+      "step": 32940
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001266987594341481,
+      "loss": 0.9284,
+      "step": 32945
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00012667463121098123,
+      "loss": 1.0717,
+      "step": 32950
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00012665050131581806,
+      "loss": 0.9979,
+      "step": 32955
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00012662636975017114,
+      "loss": 1.0399,
+      "step": 32960
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00012660223651555304,
+      "loss": 1.1574,
+      "step": 32965
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00012657810161347644,
+      "loss": 0.8294,
+      "step": 32970
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012655396504545417,
+      "loss": 1.0261,
+      "step": 32975
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00012652982681299915,
+      "loss": 0.9312,
+      "step": 32980
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00012650568691762435,
+      "loss": 0.8175,
+      "step": 32985
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00012648154536084292,
+      "loss": 1.0048,
+      "step": 32990
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00012645740214416805,
+      "loss": 0.8705,
+      "step": 32995
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001264332572691131,
+      "loss": 0.8976,
+      "step": 33000
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00012640911073719146,
+      "loss": 1.0398,
+      "step": 33005
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00012638496254991667,
+      "loss": 1.1284,
+      "step": 33010
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001263608127088024,
+      "loss": 0.9641,
+      "step": 33015
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00012633666121536236,
+      "loss": 0.8835,
+      "step": 33020
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00012631250807111042,
+      "loss": 0.8836,
+      "step": 33025
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001262883532775605,
+      "loss": 0.9024,
+      "step": 33030
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00012626419683622664,
+      "loss": 0.8939,
+      "step": 33035
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.000126240038748623,
+      "loss": 0.8697,
+      "step": 33040
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00012621587901626385,
+      "loss": 0.9675,
+      "step": 33045
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00012619171764066358,
+      "loss": 1.0066,
+      "step": 33050
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0001261675546233366,
+      "loss": 0.9397,
+      "step": 33055
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00012614338996579748,
+      "loss": 0.9274,
+      "step": 33060
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0001261192236695609,
+      "loss": 0.8634,
+      "step": 33065
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00012609505573614167,
+      "loss": 0.9893,
+      "step": 33070
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001260708861670546,
+      "loss": 0.927,
+      "step": 33075
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001260467149638147,
+      "loss": 0.8995,
+      "step": 33080
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00012602254212793702,
+      "loss": 0.9559,
+      "step": 33085
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012599836766093677,
+      "loss": 0.8693,
+      "step": 33090
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012597419156432923,
+      "loss": 0.8482,
+      "step": 33095
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001259500138396298,
+      "loss": 1.0206,
+      "step": 33100
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012592583448835394,
+      "loss": 1.0884,
+      "step": 33105
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.703125,
+      "learning_rate": 0.00012590165351201725,
+      "loss": 1.1486,
+      "step": 33110
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.0001258774709121354,
+      "loss": 0.8503,
+      "step": 33115
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001258532866902242,
+      "loss": 0.9517,
+      "step": 33120
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00012582910084779956,
+      "loss": 1.1299,
+      "step": 33125
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00012580491338637744,
+      "loss": 0.9218,
+      "step": 33130
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.000125780724307474,
+      "loss": 0.9923,
+      "step": 33135
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001257565336126054,
+      "loss": 0.9179,
+      "step": 33140
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00012573234130328789,
+      "loss": 0.8966,
+      "step": 33145
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00012570814738103794,
+      "loss": 0.7933,
+      "step": 33150
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00012568395184737205,
+      "loss": 0.9099,
+      "step": 33155
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001256597547038068,
+      "loss": 1.0154,
+      "step": 33160
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001256355559518589,
+      "loss": 0.9243,
+      "step": 33165
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00012561135559304516,
+      "loss": 1.0208,
+      "step": 33170
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00012558715362888246,
+      "loss": 1.0687,
+      "step": 33175
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00012556295006088783,
+      "loss": 0.9965,
+      "step": 33180
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001255387448905784,
+      "loss": 1.0748,
+      "step": 33185
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00012551453811947136,
+      "loss": 0.8903,
+      "step": 33190
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.000125490329749084,
+      "loss": 1.049,
+      "step": 33195
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001254661197809337,
+      "loss": 0.7615,
+      "step": 33200
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00012544190821653806,
+      "loss": 0.8462,
+      "step": 33205
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00012541769505741465,
+      "loss": 0.8998,
+      "step": 33210
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00012539348030508115,
+      "loss": 0.8996,
+      "step": 33215
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00012536926396105534,
+      "loss": 0.9038,
+      "step": 33220
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012534504602685522,
+      "loss": 0.8514,
+      "step": 33225
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00012532082650399873,
+      "loss": 0.8233,
+      "step": 33230
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.000125296605394004,
+      "loss": 0.9716,
+      "step": 33235
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001252723826983892,
+      "loss": 1.0804,
+      "step": 33240
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00012524815841867272,
+      "loss": 1.1893,
+      "step": 33245
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00012522393255637293,
+      "loss": 0.9279,
+      "step": 33250
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00012519970511300826,
+      "loss": 0.8892,
+      "step": 33255
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00012517547609009738,
+      "loss": 0.9727,
+      "step": 33260
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00012515124548915905,
+      "loss": 0.9415,
+      "step": 33265
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00012512701331171195,
+      "loss": 0.9745,
+      "step": 33270
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00012510277955927505,
+      "loss": 0.8611,
+      "step": 33275
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00012507854423336737,
+      "loss": 1.0678,
+      "step": 33280
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001250543073355079,
+      "loss": 0.8843,
+      "step": 33285
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.625,
+      "learning_rate": 0.000125030068867216,
+      "loss": 0.8916,
+      "step": 33290
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001250058288300108,
+      "loss": 0.9249,
+      "step": 33295
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00012498158722541183,
+      "loss": 0.9577,
+      "step": 33300
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001249573440549385,
+      "loss": 1.1123,
+      "step": 33305
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012493309932011038,
+      "loss": 0.8586,
+      "step": 33310
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001249088530224473,
+      "loss": 1.082,
+      "step": 33315
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00012488460516346886,
+      "loss": 0.9669,
+      "step": 33320
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.000124860355744695,
+      "loss": 0.9882,
+      "step": 33325
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001248361047676458,
+      "loss": 0.9785,
+      "step": 33330
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00012481185223384123,
+      "loss": 0.8829,
+      "step": 33335
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00012478759814480155,
+      "loss": 1.1137,
+      "step": 33340
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00012476334250204694,
+      "loss": 1.1887,
+      "step": 33345
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00012473908530709782,
+      "loss": 0.847,
+      "step": 33350
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00012471482656147467,
+      "loss": 0.9713,
+      "step": 33355
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00012469056626669803,
+      "loss": 0.8042,
+      "step": 33360
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001246663044242886,
+      "loss": 0.937,
+      "step": 33365
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001246420410357671,
+      "loss": 0.9339,
+      "step": 33370
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00012461777610265444,
+      "loss": 0.8133,
+      "step": 33375
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00012459350962647147,
+      "loss": 0.9549,
+      "step": 33380
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00012456924160873936,
+      "loss": 0.9742,
+      "step": 33385
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00012454497205097916,
+      "loss": 1.0086,
+      "step": 33390
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001245207009547122,
+      "loss": 0.9169,
+      "step": 33395
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00012449642832145977,
+      "loss": 1.0439,
+      "step": 33400
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001244721541527433,
+      "loss": 0.9336,
+      "step": 33405
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00012444787845008432,
+      "loss": 0.972,
+      "step": 33410
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00012442360121500448,
+      "loss": 1.0084,
+      "step": 33415
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00012439932244902554,
+      "loss": 0.9818,
+      "step": 33420
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00012437504215366926,
+      "loss": 0.9483,
+      "step": 33425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00012435076033045757,
+      "loss": 0.9617,
+      "step": 33430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001243264769809125,
+      "loss": 0.9819,
+      "step": 33435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001243021921065561,
+      "loss": 1.0103,
+      "step": 33440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00012427790570891068,
+      "loss": 0.8925,
+      "step": 33445
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012425361778949846,
+      "loss": 0.8525,
+      "step": 33450
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00012422932834984187,
+      "loss": 1.038,
+      "step": 33455
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00012420503739146333,
+      "loss": 0.9946,
+      "step": 33460
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00012418074491588553,
+      "loss": 1.1663,
+      "step": 33465
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001241564509246311,
+      "loss": 0.9745,
+      "step": 33470
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00012413215541922282,
+      "loss": 0.8693,
+      "step": 33475
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00012410785840118353,
+      "loss": 1.0118,
+      "step": 33480
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001240835598720362,
+      "loss": 0.9253,
+      "step": 33485
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00012405925983330392,
+      "loss": 0.829,
+      "step": 33490
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00012403495828650985,
+      "loss": 0.8592,
+      "step": 33495
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012401065523317723,
+      "loss": 0.9773,
+      "step": 33500
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00012398635067482937,
+      "loss": 1.0826,
+      "step": 33505
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00012396204461298974,
+      "loss": 0.9736,
+      "step": 33510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00012393773704918185,
+      "loss": 0.9207,
+      "step": 33515
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001239134279849294,
+      "loss": 1.0038,
+      "step": 33520
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00012388911742175599,
+      "loss": 1.0029,
+      "step": 33525
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001238648053611855,
+      "loss": 1.0463,
+      "step": 33530
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00012384049180474182,
+      "loss": 1.1262,
+      "step": 33535
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00012381617675394897,
+      "loss": 0.8751,
+      "step": 33540
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00012379186021033105,
+      "loss": 1.022,
+      "step": 33545
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00012376754217541225,
+      "loss": 1.0283,
+      "step": 33550
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00012374322265071682,
+      "loss": 1.0613,
+      "step": 33555
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012371890163776912,
+      "loss": 0.9887,
+      "step": 33560
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001236945791380937,
+      "loss": 1.0686,
+      "step": 33565
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012367025515321503,
+      "loss": 1.0544,
+      "step": 33570
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00012364592968465784,
+      "loss": 1.1148,
+      "step": 33575
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00012362160273394685,
+      "loss": 1.0274,
+      "step": 33580
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001235972743026069,
+      "loss": 1.0751,
+      "step": 33585
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001235729443921629,
+      "loss": 1.0264,
+      "step": 33590
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001235486130041399,
+      "loss": 0.924,
+      "step": 33595
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00012352428014006302,
+      "loss": 1.1058,
+      "step": 33600
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001234999458014575,
+      "loss": 0.944,
+      "step": 33605
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00012347560998984857,
+      "loss": 1.0496,
+      "step": 33610
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001234512727067617,
+      "loss": 1.0869,
+      "step": 33615
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00012342693395372232,
+      "loss": 0.9073,
+      "step": 33620
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00012340259373225604,
+      "loss": 0.9367,
+      "step": 33625
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00012337825204388858,
+      "loss": 0.9548,
+      "step": 33630
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001233539088901456,
+      "loss": 0.902,
+      "step": 33635
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.000123329564272553,
+      "loss": 1.0901,
+      "step": 33640
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001233052181926368,
+      "loss": 0.9495,
+      "step": 33645
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00012328087065192295,
+      "loss": 1.1589,
+      "step": 33650
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00012325652165193763,
+      "loss": 0.8439,
+      "step": 33655
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00012323217119420706,
+      "loss": 0.9891,
+      "step": 33660
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00012320781928025747,
+      "loss": 0.8962,
+      "step": 33665
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0001231834659116154,
+      "loss": 0.9551,
+      "step": 33670
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00012315911108980727,
+      "loss": 0.9165,
+      "step": 33675
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00012313475481635965,
+      "loss": 0.8704,
+      "step": 33680
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001231103970927993,
+      "loss": 1.014,
+      "step": 33685
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001230860379206529,
+      "loss": 0.9021,
+      "step": 33690
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00012306167730144737,
+      "loss": 0.9223,
+      "step": 33695
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00012303731523670964,
+      "loss": 0.8964,
+      "step": 33700
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00012301295172796673,
+      "loss": 1.0116,
+      "step": 33705
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012298858677674585,
+      "loss": 0.9617,
+      "step": 33710
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.84765625,
+      "learning_rate": 0.00012296422038457413,
+      "loss": 0.9698,
+      "step": 33715
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001229398525529789,
+      "loss": 0.9268,
+      "step": 33720
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00012291548328348764,
+      "loss": 1.0052,
+      "step": 33725
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012289111257762775,
+      "loss": 1.0781,
+      "step": 33730
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00012286674043692687,
+      "loss": 1.0105,
+      "step": 33735
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00012284236686291265,
+      "loss": 0.9892,
+      "step": 33740
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012281799185711285,
+      "loss": 0.9825,
+      "step": 33745
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00012279361542105537,
+      "loss": 0.9173,
+      "step": 33750
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001227692375562681,
+      "loss": 0.956,
+      "step": 33755
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00012274485826427905,
+      "loss": 1.0413,
+      "step": 33760
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00012272047754661642,
+      "loss": 0.9672,
+      "step": 33765
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00012269609540480834,
+      "loss": 0.9924,
+      "step": 33770
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001226717118403832,
+      "loss": 0.915,
+      "step": 33775
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00012264732685486932,
+      "loss": 0.9189,
+      "step": 33780
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001226229404497952,
+      "loss": 0.9149,
+      "step": 33785
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001225985526266894,
+      "loss": 0.9591,
+      "step": 33790
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001225741633870806,
+      "loss": 0.9421,
+      "step": 33795
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012254977273249752,
+      "loss": 0.8335,
+      "step": 33800
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.000122525380664469,
+      "loss": 0.8758,
+      "step": 33805
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00012250098718452398,
+      "loss": 0.9467,
+      "step": 33810
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00012247659229419147,
+      "loss": 0.9301,
+      "step": 33815
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001224521959950005,
+      "loss": 0.9686,
+      "step": 33820
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012242779828848033,
+      "loss": 0.8002,
+      "step": 33825
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00012240339917616027,
+      "loss": 0.9443,
+      "step": 33830
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001223789986595696,
+      "loss": 1.0118,
+      "step": 33835
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001223545967402378,
+      "loss": 0.9694,
+      "step": 33840
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012233019341969443,
+      "loss": 0.9285,
+      "step": 33845
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00012230578869946909,
+      "loss": 1.1095,
+      "step": 33850
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.73046875,
+      "learning_rate": 0.00012228138258109153,
+      "loss": 1.1028,
+      "step": 33855
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001222569750660915,
+      "loss": 0.9773,
+      "step": 33860
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00012223256615599896,
+      "loss": 0.9878,
+      "step": 33865
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00012220815585234384,
+      "loss": 1.0119,
+      "step": 33870
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00012218374415665624,
+      "loss": 1.0379,
+      "step": 33875
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00012215933107046626,
+      "loss": 1.0315,
+      "step": 33880
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00012213491659530417,
+      "loss": 0.8895,
+      "step": 33885
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001221105007327003,
+      "loss": 0.8463,
+      "step": 33890
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001220860834841851,
+      "loss": 0.8936,
+      "step": 33895
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00012206166485128898,
+      "loss": 0.9165,
+      "step": 33900
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00012203724483554262,
+      "loss": 1.2033,
+      "step": 33905
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00012201282343847662,
+      "loss": 1.0308,
+      "step": 33910
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00012198840066162178,
+      "loss": 0.9276,
+      "step": 33915
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012196397650650897,
+      "loss": 0.8148,
+      "step": 33920
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00012193955097466909,
+      "loss": 1.0166,
+      "step": 33925
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00012191512406763319,
+      "loss": 0.8964,
+      "step": 33930
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001218906957869323,
+      "loss": 0.7973,
+      "step": 33935
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00012186626613409771,
+      "loss": 0.8976,
+      "step": 33940
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012184183511066065,
+      "loss": 0.8737,
+      "step": 33945
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00012181740271815248,
+      "loss": 0.8809,
+      "step": 33950
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00012179296895810466,
+      "loss": 0.875,
+      "step": 33955
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00012176853383204873,
+      "loss": 0.8637,
+      "step": 33960
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00012174409734151628,
+      "loss": 0.9441,
+      "step": 33965
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001217196594880391,
+      "loss": 0.9632,
+      "step": 33970
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00012169522027314888,
+      "loss": 1.0033,
+      "step": 33975
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00012167077969837755,
+      "loss": 0.9105,
+      "step": 33980
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001216463377652571,
+      "loss": 1.0741,
+      "step": 33985
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00012162189447531949,
+      "loss": 1.0443,
+      "step": 33990
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00012159744983009695,
+      "loss": 1.0444,
+      "step": 33995
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00012157300383112167,
+      "loss": 1.0718,
+      "step": 34000
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00012154855647992591,
+      "loss": 0.9506,
+      "step": 34005
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00012152410777804209,
+      "loss": 0.9411,
+      "step": 34010
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00012149965772700269,
+      "loss": 0.9822,
+      "step": 34015
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00012147520632834023,
+      "loss": 0.9988,
+      "step": 34020
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00012145075358358744,
+      "loss": 0.9612,
+      "step": 34025
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00012142629949427693,
+      "loss": 0.9243,
+      "step": 34030
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001214018440619416,
+      "loss": 1.0716,
+      "step": 34035
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001213773872881143,
+      "loss": 0.996,
+      "step": 34040
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.765625,
+      "learning_rate": 0.00012135292917432799,
+      "loss": 0.842,
+      "step": 34045
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001213284697221158,
+      "loss": 0.915,
+      "step": 34050
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00012130400893301081,
+      "loss": 0.9155,
+      "step": 34055
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00012127954680854628,
+      "loss": 0.8287,
+      "step": 34060
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00012125508335025552,
+      "loss": 0.829,
+      "step": 34065
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012123061855967195,
+      "loss": 0.9948,
+      "step": 34070
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.72265625,
+      "learning_rate": 0.00012120615243832903,
+      "loss": 0.9894,
+      "step": 34075
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001211816849877603,
+      "loss": 1.0433,
+      "step": 34080
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00012115721620949942,
+      "loss": 0.9365,
+      "step": 34085
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00012113274610508013,
+      "loss": 1.0574,
+      "step": 34090
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00012110827467603629,
+      "loss": 0.8886,
+      "step": 34095
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00012108380192390172,
+      "loss": 0.8493,
+      "step": 34100
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00012105932785021046,
+      "loss": 0.9578,
+      "step": 34105
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00012103485245649651,
+      "loss": 1.0486,
+      "step": 34110
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00012101037574429409,
+      "loss": 0.9805,
+      "step": 34115
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012098589771513736,
+      "loss": 0.9243,
+      "step": 34120
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00012096141837056067,
+      "loss": 0.8967,
+      "step": 34125
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001209369377120984,
+      "loss": 0.902,
+      "step": 34130
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00012091245574128505,
+      "loss": 1.0998,
+      "step": 34135
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001208879724596551,
+      "loss": 0.9651,
+      "step": 34140
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00012086348786874331,
+      "loss": 0.9981,
+      "step": 34145
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001208390019700843,
+      "loss": 1.0658,
+      "step": 34150
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00012081451476521293,
+      "loss": 0.9321,
+      "step": 34155
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00012079002625566409,
+      "loss": 0.9644,
+      "step": 34160
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00012076553644297268,
+      "loss": 1.0264,
+      "step": 34165
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00012074104532867381,
+      "loss": 0.9996,
+      "step": 34170
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00012071655291430261,
+      "loss": 0.9504,
+      "step": 34175
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00012069205920139428,
+      "loss": 0.9919,
+      "step": 34180
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001206675641914841,
+      "loss": 1.0253,
+      "step": 34185
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00012064306788610749,
+      "loss": 1.0116,
+      "step": 34190
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00012061857028679982,
+      "loss": 0.9081,
+      "step": 34195
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00012059407139509671,
+      "loss": 0.9639,
+      "step": 34200
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00012056957121253377,
+      "loss": 0.9585,
+      "step": 34205
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00012054506974064665,
+      "loss": 0.8973,
+      "step": 34210
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00012052056698097118,
+      "loss": 0.9475,
+      "step": 34215
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00012049606293504317,
+      "loss": 0.9987,
+      "step": 34220
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8671875,
+      "learning_rate": 0.00012047155760439861,
+      "loss": 0.9729,
+      "step": 34225
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00012044705099057352,
+      "loss": 0.8508,
+      "step": 34230
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00012042254309510398,
+      "loss": 1.0183,
+      "step": 34235
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012039803391952617,
+      "loss": 0.856,
+      "step": 34240
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.91796875,
+      "learning_rate": 0.00012037352346537639,
+      "loss": 0.7685,
+      "step": 34245
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00012034901173419091,
+      "loss": 0.9855,
+      "step": 34250
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00012032449872750621,
+      "loss": 0.9725,
+      "step": 34255
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00012029998444685881,
+      "loss": 0.9326,
+      "step": 34260
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.00012027546889378525,
+      "loss": 0.9455,
+      "step": 34265
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001202509520698222,
+      "loss": 0.9497,
+      "step": 34270
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.73828125,
+      "learning_rate": 0.00012022643397650642,
+      "loss": 1.266,
+      "step": 34275
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00012020191461537471,
+      "loss": 0.9617,
+      "step": 34280
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.76171875,
+      "learning_rate": 0.00012017739398796401,
+      "loss": 0.9301,
+      "step": 34285
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00012015287209581125,
+      "loss": 0.9156,
+      "step": 34290
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00012012834894045353,
+      "loss": 0.9515,
+      "step": 34295
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00012010382452342797,
+      "loss": 0.8186,
+      "step": 34300
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00012007929884627176,
+      "loss": 0.9397,
+      "step": 34305
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00012005477191052228,
+      "loss": 0.9877,
+      "step": 34310
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00012003024371771683,
+      "loss": 1.0068,
+      "step": 34315
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00012000571426939289,
+      "loss": 0.913,
+      "step": 34320
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.000119981183567088,
+      "loss": 0.9533,
+      "step": 34325
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011995665161233977,
+      "loss": 0.9048,
+      "step": 34330
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00011993211840668588,
+      "loss": 0.9978,
+      "step": 34335
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00011990758395166415,
+      "loss": 0.9513,
+      "step": 34340
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00011988304824881234,
+      "loss": 0.9389,
+      "step": 34345
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00011985851129966843,
+      "loss": 0.9301,
+      "step": 34350
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001198339731057704,
+      "loss": 1.1964,
+      "step": 34355
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011980943366865636,
+      "loss": 1.0127,
+      "step": 34360
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00011978489298986448,
+      "loss": 0.8845,
+      "step": 34365
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00011976035107093294,
+      "loss": 0.9555,
+      "step": 34370
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00011973580791340011,
+      "loss": 0.903,
+      "step": 34375
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00011971126351880435,
+      "loss": 0.9846,
+      "step": 34380
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011968671788868413,
+      "loss": 0.9165,
+      "step": 34385
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011966217102457807,
+      "loss": 1.0022,
+      "step": 34390
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0001196376229280247,
+      "loss": 1.0029,
+      "step": 34395
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011961307360056273,
+      "loss": 0.9662,
+      "step": 34400
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00011958852304373099,
+      "loss": 1.0455,
+      "step": 34405
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00011956397125906834,
+      "loss": 0.8823,
+      "step": 34410
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011953941824811363,
+      "loss": 0.8869,
+      "step": 34415
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011951486401240601,
+      "loss": 0.8601,
+      "step": 34420
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00011949030855348445,
+      "loss": 0.9119,
+      "step": 34425
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011946575187288815,
+      "loss": 0.937,
+      "step": 34430
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00011944119397215634,
+      "loss": 0.9595,
+      "step": 34435
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00011941663485282837,
+      "loss": 0.9202,
+      "step": 34440
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011939207451644363,
+      "loss": 0.9054,
+      "step": 34445
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011936751296454155,
+      "loss": 1.0402,
+      "step": 34450
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00011934295019866168,
+      "loss": 1.0998,
+      "step": 34455
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00011931838622034371,
+      "loss": 0.9521,
+      "step": 34460
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00011929382103112725,
+      "loss": 0.9183,
+      "step": 34465
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011926925463255214,
+      "loss": 1.0728,
+      "step": 34470
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00011924468702615818,
+      "loss": 0.8879,
+      "step": 34475
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.74609375,
+      "learning_rate": 0.00011922011821348533,
+      "loss": 0.9262,
+      "step": 34480
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00011919554819607359,
+      "loss": 0.9469,
+      "step": 34485
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00011917097697546303,
+      "loss": 1.0009,
+      "step": 34490
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00011914640455319377,
+      "loss": 0.8685,
+      "step": 34495
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.71875,
+      "learning_rate": 0.00011912183093080611,
+      "loss": 0.942,
+      "step": 34500
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00011909725610984026,
+      "loss": 0.9429,
+      "step": 34505
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00011907268009183668,
+      "loss": 0.9046,
+      "step": 34510
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00011904810287833579,
+      "loss": 0.9142,
+      "step": 34515
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001190235244708781,
+      "loss": 1.062,
+      "step": 34520
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011899894487100425,
+      "loss": 0.8536,
+      "step": 34525
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011897436408025488,
+      "loss": 1.0806,
+      "step": 34530
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00011894978210017076,
+      "loss": 1.1766,
+      "step": 34535
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00011892519893229272,
+      "loss": 0.9887,
+      "step": 34540
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00011890061457816167,
+      "loss": 0.9299,
+      "step": 34545
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011887602903931856,
+      "loss": 0.9877,
+      "step": 34550
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00011885144231730445,
+      "loss": 0.8862,
+      "step": 34555
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00011882685441366046,
+      "loss": 0.9471,
+      "step": 34560
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001188022653299278,
+      "loss": 0.9956,
+      "step": 34565
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011877767506764774,
+      "loss": 1.0488,
+      "step": 34570
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00011875308362836163,
+      "loss": 0.964,
+      "step": 34575
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00011872849101361088,
+      "loss": 0.8103,
+      "step": 34580
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00011870389722493698,
+      "loss": 0.9446,
+      "step": 34585
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00011867930226388147,
+      "loss": 0.8547,
+      "step": 34590
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00011865470613198603,
+      "loss": 0.8854,
+      "step": 34595
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001186301088307924,
+      "loss": 1.0066,
+      "step": 34600
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00011860551036184229,
+      "loss": 0.8916,
+      "step": 34605
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00011858091072667763,
+      "loss": 1.0072,
+      "step": 34610
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00011855630992684028,
+      "loss": 0.8621,
+      "step": 34615
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8828125,
+      "learning_rate": 0.00011853170796387233,
+      "loss": 0.9904,
+      "step": 34620
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00011850710483931581,
+      "loss": 1.0208,
+      "step": 34625
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00011848250055471288,
+      "loss": 0.8806,
+      "step": 34630
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00011845789511160579,
+      "loss": 0.9947,
+      "step": 34635
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00011843328851153679,
+      "loss": 0.974,
+      "step": 34640
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00011840868075604825,
+      "loss": 0.9877,
+      "step": 34645
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011838407184668265,
+      "loss": 0.9903,
+      "step": 34650
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001183594617849825,
+      "loss": 0.9617,
+      "step": 34655
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011833485057249038,
+      "loss": 0.9091,
+      "step": 34660
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00011831023821074893,
+      "loss": 0.9464,
+      "step": 34665
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00011828562470130088,
+      "loss": 0.8344,
+      "step": 34670
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00011826101004568908,
+      "loss": 0.9523,
+      "step": 34675
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011823639424545639,
+      "loss": 0.9397,
+      "step": 34680
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001182117773021457,
+      "loss": 0.9879,
+      "step": 34685
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00011818715921730006,
+      "loss": 0.9295,
+      "step": 34690
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00011816253999246258,
+      "loss": 1.0233,
+      "step": 34695
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00011813791962917642,
+      "loss": 1.0116,
+      "step": 34700
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00011811329812898482,
+      "loss": 0.9946,
+      "step": 34705
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.000118088675493431,
+      "loss": 1.0791,
+      "step": 34710
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00011806405172405845,
+      "loss": 1.0352,
+      "step": 34715
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011803942682241053,
+      "loss": 0.8494,
+      "step": 34720
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001180148007900308,
+      "loss": 0.955,
+      "step": 34725
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00011799017362846287,
+      "loss": 1.0439,
+      "step": 34730
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00011796554533925037,
+      "loss": 0.924,
+      "step": 34735
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.000117940915923937,
+      "loss": 0.8782,
+      "step": 34740
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011791628538406659,
+      "loss": 0.923,
+      "step": 34745
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00011789165372118301,
+      "loss": 1.0874,
+      "step": 34750
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00011786702093683018,
+      "loss": 0.7821,
+      "step": 34755
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011784238703255217,
+      "loss": 0.8877,
+      "step": 34760
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011781775200989298,
+      "loss": 0.8183,
+      "step": 34765
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0001177931158703968,
+      "loss": 0.933,
+      "step": 34770
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001177684786156079,
+      "loss": 1.0486,
+      "step": 34775
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011774384024707046,
+      "loss": 1.1156,
+      "step": 34780
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011771920076632897,
+      "loss": 1.1137,
+      "step": 34785
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00011769456017492777,
+      "loss": 0.8434,
+      "step": 34790
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00011766991847441136,
+      "loss": 1.0149,
+      "step": 34795
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00011764527566632435,
+      "loss": 0.9178,
+      "step": 34800
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00011762063175221139,
+      "loss": 0.9654,
+      "step": 34805
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00011759598673361714,
+      "loss": 0.9076,
+      "step": 34810
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00011757134061208642,
+      "loss": 1.0185,
+      "step": 34815
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011754669338916401,
+      "loss": 0.9442,
+      "step": 34820
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00011752204506639493,
+      "loss": 0.9538,
+      "step": 34825
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00011749739564532407,
+      "loss": 1.0216,
+      "step": 34830
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00011747274512749653,
+      "loss": 0.9645,
+      "step": 34835
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00011744809351445747,
+      "loss": 0.9073,
+      "step": 34840
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00011742344080775198,
+      "loss": 0.9201,
+      "step": 34845
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011739878700892541,
+      "loss": 1.1989,
+      "step": 34850
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00011737413211952304,
+      "loss": 0.861,
+      "step": 34855
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00011734947614109029,
+      "loss": 1.0942,
+      "step": 34860
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00011732481907517261,
+      "loss": 1.0245,
+      "step": 34865
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.83203125,
+      "learning_rate": 0.00011730016092331554,
+      "loss": 0.9959,
+      "step": 34870
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011727550168706467,
+      "loss": 0.9585,
+      "step": 34875
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00011725084136796569,
+      "loss": 1.0257,
+      "step": 34880
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00011722617996756433,
+      "loss": 0.9892,
+      "step": 34885
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00011720151748740639,
+      "loss": 0.8525,
+      "step": 34890
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00011717685392903774,
+      "loss": 1.065,
+      "step": 34895
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001171521892940043,
+      "loss": 0.9966,
+      "step": 34900
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00011712752358385216,
+      "loss": 0.9208,
+      "step": 34905
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001171028568001273,
+      "loss": 0.8902,
+      "step": 34910
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00011707818894437587,
+      "loss": 0.8528,
+      "step": 34915
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011705352001814415,
+      "loss": 1.0698,
+      "step": 34920
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011702885002297836,
+      "loss": 0.8962,
+      "step": 34925
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00011700417896042484,
+      "loss": 1.0051,
+      "step": 34930
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011697950683203003,
+      "loss": 0.8504,
+      "step": 34935
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011695483363934038,
+      "loss": 0.8656,
+      "step": 34940
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011693015938390249,
+      "loss": 0.8445,
+      "step": 34945
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001169054840672629,
+      "loss": 1.068,
+      "step": 34950
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00011688080769096827,
+      "loss": 1.0567,
+      "step": 34955
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00011685613025656543,
+      "loss": 1.0828,
+      "step": 34960
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011683145176560117,
+      "loss": 0.9981,
+      "step": 34965
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00011680677221962233,
+      "loss": 0.963,
+      "step": 34970
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011678209162017586,
+      "loss": 0.9921,
+      "step": 34975
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00011675740996880877,
+      "loss": 0.9473,
+      "step": 34980
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00011673272726706812,
+      "loss": 0.9033,
+      "step": 34985
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00011670804351650109,
+      "loss": 0.9988,
+      "step": 34990
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011668335871865487,
+      "loss": 0.7865,
+      "step": 34995
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00011665867287507672,
+      "loss": 0.9627,
+      "step": 35000
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00011663398598731399,
+      "loss": 0.9706,
+      "step": 35005
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011660929805691402,
+      "loss": 0.8479,
+      "step": 35010
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00011658460908542438,
+      "loss": 0.9575,
+      "step": 35015
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00011655991907439257,
+      "loss": 1.0464,
+      "step": 35020
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00011653522802536613,
+      "loss": 0.9209,
+      "step": 35025
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0001165105359398928,
+      "loss": 0.855,
+      "step": 35030
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00011648584281952026,
+      "loss": 0.9487,
+      "step": 35035
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011646114866579632,
+      "loss": 0.9703,
+      "step": 35040
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00011643645348026882,
+      "loss": 0.929,
+      "step": 35045
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001164117572644857,
+      "loss": 0.9992,
+      "step": 35050
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00011638706001999495,
+      "loss": 0.8706,
+      "step": 35055
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00011636236174834463,
+      "loss": 0.8983,
+      "step": 35060
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001163376624510828,
+      "loss": 1.0189,
+      "step": 35065
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00011631296212975771,
+      "loss": 0.9888,
+      "step": 35070
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0001162882607859176,
+      "loss": 0.9474,
+      "step": 35075
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001162635584211107,
+      "loss": 0.9624,
+      "step": 35080
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00011623885503688546,
+      "loss": 0.8918,
+      "step": 35085
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00011621415063479028,
+      "loss": 0.8406,
+      "step": 35090
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00011618944521637368,
+      "loss": 0.9026,
+      "step": 35095
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00011616473878318424,
+      "loss": 1.0991,
+      "step": 35100
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011614003133677052,
+      "loss": 0.9176,
+      "step": 35105
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011611532287868128,
+      "loss": 0.9866,
+      "step": 35110
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00011609061341046523,
+      "loss": 0.9598,
+      "step": 35115
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011606590293367121,
+      "loss": 0.94,
+      "step": 35120
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001160411914498481,
+      "loss": 0.9376,
+      "step": 35125
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00011601647896054486,
+      "loss": 1.0,
+      "step": 35130
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011599176546731045,
+      "loss": 0.9325,
+      "step": 35135
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.000115967050971694,
+      "loss": 1.0462,
+      "step": 35140
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011594233547524458,
+      "loss": 1.0102,
+      "step": 35145
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00011591761897951141,
+      "loss": 0.9795,
+      "step": 35150
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.67578125,
+      "learning_rate": 0.0001158929014860438,
+      "loss": 0.9715,
+      "step": 35155
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00011586818299639097,
+      "loss": 0.9373,
+      "step": 35160
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011584346351210238,
+      "loss": 0.8551,
+      "step": 35165
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011581874303472746,
+      "loss": 1.0983,
+      "step": 35170
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011579402156581568,
+      "loss": 0.9248,
+      "step": 35175
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0001157692991069167,
+      "loss": 0.7816,
+      "step": 35180
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00011574457565958003,
+      "loss": 0.9678,
+      "step": 35185
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011571985122535547,
+      "loss": 1.0563,
+      "step": 35190
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00011569512580579271,
+      "loss": 0.9766,
+      "step": 35195
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001156703994024416,
+      "loss": 1.0294,
+      "step": 35200
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00011564567201685202,
+      "loss": 0.908,
+      "step": 35205
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00011562094365057388,
+      "loss": 0.9647,
+      "step": 35210
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00011559621430515717,
+      "loss": 0.7352,
+      "step": 35215
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00011557148398215203,
+      "loss": 0.8482,
+      "step": 35220
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011554675268310853,
+      "loss": 0.8756,
+      "step": 35225
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00011552202040957684,
+      "loss": 0.926,
+      "step": 35230
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00011549728716310728,
+      "loss": 0.7878,
+      "step": 35235
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011547255294525006,
+      "loss": 1.1045,
+      "step": 35240
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011544781775755565,
+      "loss": 0.9171,
+      "step": 35245
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001154230816015744,
+      "loss": 0.8814,
+      "step": 35250
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001153983444788568,
+      "loss": 0.7096,
+      "step": 35255
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9296875,
+      "learning_rate": 0.00011537360639095349,
+      "loss": 1.052,
+      "step": 35260
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00011534886733941502,
+      "loss": 0.8847,
+      "step": 35265
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00011532412732579201,
+      "loss": 0.9098,
+      "step": 35270
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011529938635163529,
+      "loss": 0.9207,
+      "step": 35275
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001152746444184956,
+      "loss": 0.8311,
+      "step": 35280
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00011524990152792381,
+      "loss": 0.9728,
+      "step": 35285
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011522515768147082,
+      "loss": 0.846,
+      "step": 35290
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00011520041288068757,
+      "loss": 0.919,
+      "step": 35295
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00011517566712712516,
+      "loss": 1.0759,
+      "step": 35300
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00011515092042233466,
+      "loss": 0.9652,
+      "step": 35305
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00011512617276786719,
+      "loss": 1.0368,
+      "step": 35310
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00011510142416527401,
+      "loss": 1.1348,
+      "step": 35315
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00011507667461610637,
+      "loss": 0.8869,
+      "step": 35320
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011505192412191554,
+      "loss": 0.9564,
+      "step": 35325
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.000115027172684253,
+      "loss": 1.0288,
+      "step": 35330
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00011500242030467017,
+      "loss": 0.8114,
+      "step": 35335
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00011497766698471852,
+      "loss": 1.0091,
+      "step": 35340
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011495291272594968,
+      "loss": 0.9346,
+      "step": 35345
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00011492815752991521,
+      "loss": 1.0135,
+      "step": 35350
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011490340139816684,
+      "loss": 0.8858,
+      "step": 35355
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6875,
+      "learning_rate": 0.0001148786443322563,
+      "loss": 0.9391,
+      "step": 35360
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00011485388633373538,
+      "loss": 0.9083,
+      "step": 35365
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00011482912740415595,
+      "loss": 0.86,
+      "step": 35370
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00011480436754506993,
+      "loss": 0.9986,
+      "step": 35375
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00011477960675802926,
+      "loss": 0.8895,
+      "step": 35380
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011475484504458604,
+      "loss": 1.0366,
+      "step": 35385
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00011473008240629233,
+      "loss": 0.8077,
+      "step": 35390
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00011470531884470026,
+      "loss": 0.9371,
+      "step": 35395
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00011468055436136209,
+      "loss": 0.892,
+      "step": 35400
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.71484375,
+      "learning_rate": 0.00011465578895783001,
+      "loss": 0.9811,
+      "step": 35405
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011463102263565639,
+      "loss": 0.9964,
+      "step": 35410
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00011460625539639367,
+      "loss": 0.9848,
+      "step": 35415
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011458148724159414,
+      "loss": 0.9357,
+      "step": 35420
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00011455671817281044,
+      "loss": 1.0116,
+      "step": 35425
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00011453194819159506,
+      "loss": 0.9715,
+      "step": 35430
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00011450717729950058,
+      "loss": 0.901,
+      "step": 35435
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011448240549807974,
+      "loss": 0.9623,
+      "step": 35440
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00011445763278888522,
+      "loss": 0.9551,
+      "step": 35445
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00011443285917346981,
+      "loss": 1.0564,
+      "step": 35450
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00011440808465338634,
+      "loss": 0.9243,
+      "step": 35455
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011438330923018771,
+      "loss": 0.8432,
+      "step": 35460
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001143585329054269,
+      "loss": 0.925,
+      "step": 35465
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00011433375568065692,
+      "loss": 0.9936,
+      "step": 35470
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011430897755743075,
+      "loss": 0.8261,
+      "step": 35475
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0001142841985373016,
+      "loss": 1.0874,
+      "step": 35480
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00011425941862182261,
+      "loss": 0.8031,
+      "step": 35485
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00011423463781254702,
+      "loss": 0.9591,
+      "step": 35490
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00011420985611102814,
+      "loss": 0.8532,
+      "step": 35495
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011418507351881927,
+      "loss": 0.8923,
+      "step": 35500
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7109375,
+      "learning_rate": 0.00011416029003747383,
+      "loss": 1.0846,
+      "step": 35505
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00011413550566854532,
+      "loss": 0.9509,
+      "step": 35510
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00011411072041358717,
+      "loss": 0.8973,
+      "step": 35515
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00011408593427415304,
+      "loss": 1.0902,
+      "step": 35520
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011406114725179647,
+      "loss": 0.8178,
+      "step": 35525
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001140363593480712,
+      "loss": 1.0575,
+      "step": 35530
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011401157056453091,
+      "loss": 0.9071,
+      "step": 35535
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00011398678090272945,
+      "loss": 1.016,
+      "step": 35540
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011396199036422062,
+      "loss": 0.9777,
+      "step": 35545
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00011393719895055834,
+      "loss": 0.8702,
+      "step": 35550
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00011391240666329655,
+      "loss": 0.83,
+      "step": 35555
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00011388761350398927,
+      "loss": 1.0824,
+      "step": 35560
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011386281947419055,
+      "loss": 0.9833,
+      "step": 35565
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00011383802457545452,
+      "loss": 1.0281,
+      "step": 35570
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.828125,
+      "learning_rate": 0.00011381322880933536,
+      "loss": 0.8782,
+      "step": 35575
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00011378843217738726,
+      "loss": 1.0211,
+      "step": 35580
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011376363468116456,
+      "loss": 0.956,
+      "step": 35585
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00011373883632222156,
+      "loss": 0.8512,
+      "step": 35590
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00011371403710211262,
+      "loss": 0.9951,
+      "step": 35595
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00011368923702239225,
+      "loss": 0.9283,
+      "step": 35600
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001136644360846149,
+      "loss": 0.8853,
+      "step": 35605
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011363963429033513,
+      "loss": 0.865,
+      "step": 35610
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011361483164110756,
+      "loss": 0.962,
+      "step": 35615
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00011359002813848682,
+      "loss": 0.9415,
+      "step": 35620
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00011356522378402765,
+      "loss": 1.0479,
+      "step": 35625
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00011354041857928481,
+      "loss": 0.9286,
+      "step": 35630
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011351561252581308,
+      "loss": 0.978,
+      "step": 35635
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00011349080562516738,
+      "loss": 0.9409,
+      "step": 35640
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00011346599787890264,
+      "loss": 0.8311,
+      "step": 35645
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00011344118928857379,
+      "loss": 0.9399,
+      "step": 35650
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00011341637985573592,
+      "loss": 0.987,
+      "step": 35655
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00011339156958194405,
+      "loss": 1.0669,
+      "step": 35660
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00011336675846875335,
+      "loss": 1.0025,
+      "step": 35665
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.000113341946517719,
+      "loss": 0.9605,
+      "step": 35670
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00011331713373039628,
+      "loss": 0.9151,
+      "step": 35675
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00011329232010834043,
+      "loss": 0.8997,
+      "step": 35680
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011326750565310681,
+      "loss": 0.9927,
+      "step": 35685
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00011324269036625082,
+      "loss": 1.0114,
+      "step": 35690
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00011321787424932793,
+      "loss": 0.9672,
+      "step": 35695
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00011319305730389363,
+      "loss": 0.9736,
+      "step": 35700
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011316823953150349,
+      "loss": 0.8902,
+      "step": 35705
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00011314342093371307,
+      "loss": 0.8646,
+      "step": 35710
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00011311860151207807,
+      "loss": 0.9136,
+      "step": 35715
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00011309378126815416,
+      "loss": 1.0051,
+      "step": 35720
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011306896020349715,
+      "loss": 0.8496,
+      "step": 35725
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00011304413831966282,
+      "loss": 0.9563,
+      "step": 35730
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011301931561820706,
+      "loss": 1.019,
+      "step": 35735
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00011299449210068577,
+      "loss": 1.0119,
+      "step": 35740
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001129696677686549,
+      "loss": 0.8935,
+      "step": 35745
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00011294484262367049,
+      "loss": 1.0451,
+      "step": 35750
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001129200166672886,
+      "loss": 0.9686,
+      "step": 35755
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00011289518990106535,
+      "loss": 0.7722,
+      "step": 35760
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011287036232655692,
+      "loss": 0.9714,
+      "step": 35765
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00011284553394531951,
+      "loss": 1.0167,
+      "step": 35770
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001128207047589094,
+      "loss": 0.8774,
+      "step": 35775
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00011279587476888292,
+      "loss": 1.0096,
+      "step": 35780
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00011277104397679646,
+      "loss": 1.0358,
+      "step": 35785
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00011274621238420637,
+      "loss": 0.9213,
+      "step": 35790
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001127213799926692,
+      "loss": 0.9913,
+      "step": 35795
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00011269654680374144,
+      "loss": 0.8914,
+      "step": 35800
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011267171281897966,
+      "loss": 0.8998,
+      "step": 35805
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00011264687803994052,
+      "loss": 0.8822,
+      "step": 35810
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011262204246818062,
+      "loss": 0.8956,
+      "step": 35815
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011259720610525674,
+      "loss": 0.895,
+      "step": 35820
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00011257236895272565,
+      "loss": 0.9545,
+      "step": 35825
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00011254753101214412,
+      "loss": 1.0074,
+      "step": 35830
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011252269228506909,
+      "loss": 0.8765,
+      "step": 35835
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7890625,
+      "learning_rate": 0.00011249785277305743,
+      "loss": 0.9155,
+      "step": 35840
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011247301247766611,
+      "loss": 0.9397,
+      "step": 35845
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011244817140045219,
+      "loss": 1.038,
+      "step": 35850
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011242332954297268,
+      "loss": 0.9239,
+      "step": 35855
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00011239848690678474,
+      "loss": 0.9346,
+      "step": 35860
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00011237364349344553,
+      "loss": 1.0483,
+      "step": 35865
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00011234879930451223,
+      "loss": 0.8625,
+      "step": 35870
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00011232395434154215,
+      "loss": 0.8454,
+      "step": 35875
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011229910860609256,
+      "loss": 0.85,
+      "step": 35880
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00011227426209972082,
+      "loss": 0.9926,
+      "step": 35885
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.00011224941482398441,
+      "loss": 1.1048,
+      "step": 35890
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00011222456678044067,
+      "loss": 0.8748,
+      "step": 35895
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011219971797064718,
+      "loss": 1.0172,
+      "step": 35900
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001121748683961615,
+      "loss": 1.0032,
+      "step": 35905
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00011215001805854116,
+      "loss": 0.9065,
+      "step": 35910
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00011212516695934388,
+      "loss": 0.9394,
+      "step": 35915
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001121003151001273,
+      "loss": 0.9367,
+      "step": 35920
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.0001120754624824492,
+      "loss": 0.9802,
+      "step": 35925
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00011205060910786737,
+      "loss": 0.8837,
+      "step": 35930
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00011202575497793961,
+      "loss": 1.0204,
+      "step": 35935
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00011200090009422388,
+      "loss": 0.9781,
+      "step": 35940
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011197604445827803,
+      "loss": 0.9268,
+      "step": 35945
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00011195118807166008,
+      "loss": 0.7944,
+      "step": 35950
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00011192633093592803,
+      "loss": 0.8921,
+      "step": 35955
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00011190147305264,
+      "loss": 1.0664,
+      "step": 35960
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00011187661442335407,
+      "loss": 0.9301,
+      "step": 35965
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00011185175504962846,
+      "loss": 0.833,
+      "step": 35970
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00011182689493302128,
+      "loss": 0.8505,
+      "step": 35975
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001118020340750909,
+      "loss": 0.9573,
+      "step": 35980
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00011177717247739559,
+      "loss": 0.9987,
+      "step": 35985
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00011175231014149366,
+      "loss": 0.8967,
+      "step": 35990
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00011172744706894357,
+      "loss": 0.9154,
+      "step": 35995
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00011170258326130374,
+      "loss": 0.8346,
+      "step": 36000
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00011167771872013264,
+      "loss": 1.0448,
+      "step": 36005
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00011165285344698886,
+      "loss": 0.912,
+      "step": 36010
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00011162798744343094,
+      "loss": 0.8778,
+      "step": 36015
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00011160312071101755,
+      "loss": 0.9186,
+      "step": 36020
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001115782532513073,
+      "loss": 0.8453,
+      "step": 36025
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00011155338506585895,
+      "loss": 1.0438,
+      "step": 36030
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001115285161562313,
+      "loss": 1.0465,
+      "step": 36035
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001115036465239831,
+      "loss": 0.6984,
+      "step": 36040
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00011147877617067326,
+      "loss": 0.9824,
+      "step": 36045
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011145390509786065,
+      "loss": 1.0159,
+      "step": 36050
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00011142903330710422,
+      "loss": 0.8533,
+      "step": 36055
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00011140416079996294,
+      "loss": 0.8495,
+      "step": 36060
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001113792875779959,
+      "loss": 0.9059,
+      "step": 36065
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.95703125,
+      "learning_rate": 0.00011135441364276214,
+      "loss": 0.9416,
+      "step": 36070
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011132953899582081,
+      "loss": 1.1856,
+      "step": 36075
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011130466363873109,
+      "loss": 0.9599,
+      "step": 36080
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00011127978757305213,
+      "loss": 0.8942,
+      "step": 36085
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00011125491080034327,
+      "loss": 0.9854,
+      "step": 36090
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001112300333221638,
+      "loss": 0.9776,
+      "step": 36095
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011120515514007302,
+      "loss": 0.9301,
+      "step": 36100
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011118027625563037,
+      "loss": 1.0208,
+      "step": 36105
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00011115539667039528,
+      "loss": 0.8565,
+      "step": 36110
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00011113051638592717,
+      "loss": 0.9049,
+      "step": 36115
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011110563540378567,
+      "loss": 0.9627,
+      "step": 36120
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011108075372553029,
+      "loss": 0.9632,
+      "step": 36125
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00011105587135272064,
+      "loss": 0.9762,
+      "step": 36130
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00011103098828691639,
+      "loss": 0.9035,
+      "step": 36135
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00011100610452967719,
+      "loss": 0.9729,
+      "step": 36140
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00011098122008256286,
+      "loss": 0.9325,
+      "step": 36145
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00011095633494713315,
+      "loss": 0.9717,
+      "step": 36150
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00011093144912494788,
+      "loss": 1.0258,
+      "step": 36155
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00011090656261756694,
+      "loss": 0.946,
+      "step": 36160
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00011088167542655022,
+      "loss": 0.8508,
+      "step": 36165
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00011085678755345768,
+      "loss": 0.9069,
+      "step": 36170
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011083189899984933,
+      "loss": 0.8887,
+      "step": 36175
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00011080700976728525,
+      "loss": 1.1359,
+      "step": 36180
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00011078211985732545,
+      "loss": 0.9425,
+      "step": 36185
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00011075722927153012,
+      "loss": 0.8805,
+      "step": 36190
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001107323380114594,
+      "loss": 1.0638,
+      "step": 36195
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00011070744607867352,
+      "loss": 0.8841,
+      "step": 36200
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00011068255347473271,
+      "loss": 0.8163,
+      "step": 36205
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001106576602011973,
+      "loss": 0.9477,
+      "step": 36210
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.0001106327662596276,
+      "loss": 0.9684,
+      "step": 36215
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.000110607871651584,
+      "loss": 0.9523,
+      "step": 36220
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00011058297637862691,
+      "loss": 0.8194,
+      "step": 36225
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00011055808044231686,
+      "loss": 1.1832,
+      "step": 36230
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00011053318384421423,
+      "loss": 1.05,
+      "step": 36235
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00011050828658587968,
+      "loss": 0.959,
+      "step": 36240
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00011048338866887376,
+      "loss": 0.8375,
+      "step": 36245
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00011045849009475709,
+      "loss": 1.179,
+      "step": 36250
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00011043359086509038,
+      "loss": 0.9546,
+      "step": 36255
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00011040869098143425,
+      "loss": 0.9842,
+      "step": 36260
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00011038379044534957,
+      "loss": 0.865,
+      "step": 36265
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00011035888925839705,
+      "loss": 0.958,
+      "step": 36270
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00011033398742213755,
+      "loss": 0.9291,
+      "step": 36275
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011030908493813197,
+      "loss": 0.9853,
+      "step": 36280
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00011028418180794122,
+      "loss": 0.982,
+      "step": 36285
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00011025927803312619,
+      "loss": 0.9879,
+      "step": 36290
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6796875,
+      "learning_rate": 0.00011023437361524795,
+      "loss": 0.9501,
+      "step": 36295
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00011020946855586753,
+      "loss": 1.0892,
+      "step": 36300
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00011018456285654597,
+      "loss": 0.9984,
+      "step": 36305
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00011015965651884446,
+      "loss": 1.0251,
+      "step": 36310
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00011013474954432405,
+      "loss": 0.9086,
+      "step": 36315
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00011010984193454603,
+      "loss": 0.9331,
+      "step": 36320
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001100849336910716,
+      "loss": 0.7747,
+      "step": 36325
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00011006002481546201,
+      "loss": 1.2171,
+      "step": 36330
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00011003511530927866,
+      "loss": 1.0336,
+      "step": 36335
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00011001020517408283,
+      "loss": 0.9015,
+      "step": 36340
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010998529441143593,
+      "loss": 0.9482,
+      "step": 36345
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00010996038302289939,
+      "loss": 0.9363,
+      "step": 36350
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00010993547101003474,
+      "loss": 1.0001,
+      "step": 36355
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010991055837440343,
+      "loss": 1.002,
+      "step": 36360
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00010988564511756703,
+      "loss": 0.8675,
+      "step": 36365
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010986073124108711,
+      "loss": 1.0402,
+      "step": 36370
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00010983581674652536,
+      "loss": 0.952,
+      "step": 36375
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00010981090163544341,
+      "loss": 0.9668,
+      "step": 36380
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010978598590940294,
+      "loss": 0.9341,
+      "step": 36385
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00010976106956996579,
+      "loss": 0.7929,
+      "step": 36390
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00010973615261869365,
+      "loss": 0.9433,
+      "step": 36395
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00010971123505714835,
+      "loss": 0.843,
+      "step": 36400
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00010968631688689181,
+      "loss": 1.108,
+      "step": 36405
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010966139810948591,
+      "loss": 0.8542,
+      "step": 36410
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010963647872649255,
+      "loss": 1.1927,
+      "step": 36415
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00010961155873947373,
+      "loss": 0.9521,
+      "step": 36420
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00010958663814999145,
+      "loss": 0.8526,
+      "step": 36425
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00010956171695960778,
+      "loss": 0.9811,
+      "step": 36430
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001095367951698848,
+      "loss": 0.8424,
+      "step": 36435
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010951187278238464,
+      "loss": 0.8957,
+      "step": 36440
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00010948694979866947,
+      "loss": 0.9119,
+      "step": 36445
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010946202622030149,
+      "loss": 0.9794,
+      "step": 36450
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010943710204884288,
+      "loss": 0.8748,
+      "step": 36455
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.000109412177285856,
+      "loss": 0.9861,
+      "step": 36460
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.75390625,
+      "learning_rate": 0.00010938725193290313,
+      "loss": 1.0779,
+      "step": 36465
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00010936232599154664,
+      "loss": 0.8525,
+      "step": 36470
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00010933739946334886,
+      "loss": 0.953,
+      "step": 36475
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00010931247234987223,
+      "loss": 1.0332,
+      "step": 36480
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00010928754465267925,
+      "loss": 0.9057,
+      "step": 36485
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00010926261637333242,
+      "loss": 1.0217,
+      "step": 36490
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00010923768751339422,
+      "loss": 0.9859,
+      "step": 36495
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010921275807442726,
+      "loss": 0.9719,
+      "step": 36500
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00010918782805799412,
+      "loss": 0.9264,
+      "step": 36505
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00010916289746565742,
+      "loss": 0.9057,
+      "step": 36510
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00010913796629897993,
+      "loss": 0.9249,
+      "step": 36515
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00010911303455952427,
+      "loss": 0.9619,
+      "step": 36520
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00010908810224885325,
+      "loss": 0.9483,
+      "step": 36525
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001090631693685296,
+      "loss": 1.0105,
+      "step": 36530
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010903823592011617,
+      "loss": 1.0082,
+      "step": 36535
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010901330190517585,
+      "loss": 0.9012,
+      "step": 36540
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010898836732527148,
+      "loss": 0.8343,
+      "step": 36545
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00010896343218196603,
+      "loss": 0.8575,
+      "step": 36550
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010893849647682244,
+      "loss": 0.9455,
+      "step": 36555
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.0001089135602114037,
+      "loss": 0.9434,
+      "step": 36560
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00010888862338727284,
+      "loss": 0.8242,
+      "step": 36565
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00010886368600599297,
+      "loss": 0.8579,
+      "step": 36570
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010883874806912715,
+      "loss": 0.9151,
+      "step": 36575
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00010881380957823853,
+      "loss": 0.8738,
+      "step": 36580
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00010878887053489031,
+      "loss": 0.9275,
+      "step": 36585
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00010876393094064563,
+      "loss": 0.957,
+      "step": 36590
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00010873899079706782,
+      "loss": 0.9025,
+      "step": 36595
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00010871405010572009,
+      "loss": 1.0292,
+      "step": 36600
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00010868910886816579,
+      "loss": 1.007,
+      "step": 36605
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00010866416708596827,
+      "loss": 0.8085,
+      "step": 36610
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00010863922476069084,
+      "loss": 0.9712,
+      "step": 36615
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00010861428189389699,
+      "loss": 0.8855,
+      "step": 36620
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00010858933848715016,
+      "loss": 1.0071,
+      "step": 36625
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00010856439454201376,
+      "loss": 1.0599,
+      "step": 36630
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001085394500600514,
+      "loss": 1.0069,
+      "step": 36635
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00010851450504282658,
+      "loss": 1.013,
+      "step": 36640
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00010848955949190286,
+      "loss": 0.9143,
+      "step": 36645
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00010846461340884393,
+      "loss": 0.9171,
+      "step": 36650
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00010843966679521333,
+      "loss": 1.0172,
+      "step": 36655
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00010841471965257483,
+      "loss": 0.8871,
+      "step": 36660
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010838977198249214,
+      "loss": 0.9314,
+      "step": 36665
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00010836482378652895,
+      "loss": 0.9426,
+      "step": 36670
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010833987506624912,
+      "loss": 0.9492,
+      "step": 36675
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.92578125,
+      "learning_rate": 0.00010831492582321641,
+      "loss": 1.0501,
+      "step": 36680
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00010828997605899465,
+      "loss": 0.898,
+      "step": 36685
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00010826502577514777,
+      "loss": 0.8869,
+      "step": 36690
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00010824007497323967,
+      "loss": 1.022,
+      "step": 36695
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010821512365483426,
+      "loss": 0.9217,
+      "step": 36700
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010819017182149558,
+      "loss": 0.9814,
+      "step": 36705
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00010816521947478757,
+      "loss": 1.0245,
+      "step": 36710
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00010814026661627432,
+      "loss": 0.909,
+      "step": 36715
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00010811531324751989,
+      "loss": 1.0503,
+      "step": 36720
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00010809035937008835,
+      "loss": 1.059,
+      "step": 36725
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00010806540498554392,
+      "loss": 0.9477,
+      "step": 36730
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0001080404500954507,
+      "loss": 0.8944,
+      "step": 36735
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00010801549470137289,
+      "loss": 1.0347,
+      "step": 36740
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010799053880487477,
+      "loss": 0.997,
+      "step": 36745
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010796558240752057,
+      "loss": 1.064,
+      "step": 36750
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.76953125,
+      "learning_rate": 0.00010794062551087461,
+      "loss": 1.0009,
+      "step": 36755
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7421875,
+      "learning_rate": 0.00010791566811650118,
+      "loss": 1.0184,
+      "step": 36760
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00010789071022596461,
+      "loss": 0.9217,
+      "step": 36765
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001078657518408294,
+      "loss": 0.9953,
+      "step": 36770
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001078407929626599,
+      "loss": 0.9915,
+      "step": 36775
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010781583359302056,
+      "loss": 0.8871,
+      "step": 36780
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010779087373347587,
+      "loss": 1.0015,
+      "step": 36785
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.70703125,
+      "learning_rate": 0.00010776591338559034,
+      "loss": 1.1607,
+      "step": 36790
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00010774095255092848,
+      "loss": 0.9473,
+      "step": 36795
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00010771599123105495,
+      "loss": 1.0485,
+      "step": 36800
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010769102942753429,
+      "loss": 1.1098,
+      "step": 36805
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010766606714193113,
+      "loss": 0.9416,
+      "step": 36810
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00010764110437581017,
+      "loss": 0.9518,
+      "step": 36815
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00010761614113073605,
+      "loss": 0.9433,
+      "step": 36820
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00010759117740827356,
+      "loss": 0.9649,
+      "step": 36825
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00010756621320998743,
+      "loss": 0.9854,
+      "step": 36830
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00010754124853744242,
+      "loss": 1.0369,
+      "step": 36835
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6484375,
+      "learning_rate": 0.00010751628339220336,
+      "loss": 0.9787,
+      "step": 36840
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00010749131777583512,
+      "loss": 0.8924,
+      "step": 36845
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0001074663516899025,
+      "loss": 0.8231,
+      "step": 36850
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00010744138513597051,
+      "loss": 1.0391,
+      "step": 36855
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.000107416418115604,
+      "loss": 0.983,
+      "step": 36860
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00010739145063036797,
+      "loss": 0.9169,
+      "step": 36865
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00010736648268182738,
+      "loss": 0.9787,
+      "step": 36870
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00010734151427154726,
+      "loss": 0.7626,
+      "step": 36875
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00010731654540109268,
+      "loss": 0.9184,
+      "step": 36880
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00010729157607202873,
+      "loss": 1.0985,
+      "step": 36885
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00010726660628592045,
+      "loss": 0.9564,
+      "step": 36890
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00010724163604433302,
+      "loss": 0.9134,
+      "step": 36895
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001072166653488316,
+      "loss": 0.8505,
+      "step": 36900
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00010719169420098134,
+      "loss": 0.9296,
+      "step": 36905
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00010716672260234753,
+      "loss": 0.9877,
+      "step": 36910
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010714175055449538,
+      "loss": 0.8699,
+      "step": 36915
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00010711677805899017,
+      "loss": 0.9097,
+      "step": 36920
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0001070918051173972,
+      "loss": 0.9882,
+      "step": 36925
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0001070668317312818,
+      "loss": 0.9172,
+      "step": 36930
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00010704185790220938,
+      "loss": 0.9566,
+      "step": 36935
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010701688363174524,
+      "loss": 1.0493,
+      "step": 36940
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00010699190892145487,
+      "loss": 0.897,
+      "step": 36945
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001069669337729037,
+      "loss": 0.9825,
+      "step": 36950
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010694195818765713,
+      "loss": 0.893,
+      "step": 36955
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00010691698216728076,
+      "loss": 0.9239,
+      "step": 36960
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010689200571334009,
+      "loss": 0.8994,
+      "step": 36965
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00010686702882740061,
+      "loss": 0.8999,
+      "step": 36970
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00010684205151102795,
+      "loss": 0.9204,
+      "step": 36975
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00010681707376578771,
+      "loss": 0.9035,
+      "step": 36980
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00010679209559324552,
+      "loss": 1.1007,
+      "step": 36985
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00010676711699496706,
+      "loss": 1.1832,
+      "step": 36990
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00010674213797251798,
+      "loss": 0.9332,
+      "step": 36995
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00010671715852746403,
+      "loss": 0.9942,
+      "step": 37000
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00010669217866137094,
+      "loss": 0.9642,
+      "step": 37005
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010666719837580443,
+      "loss": 0.8863,
+      "step": 37010
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010664221767233036,
+      "loss": 0.8984,
+      "step": 37015
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00010661723655251454,
+      "loss": 0.9529,
+      "step": 37020
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00010659225501792277,
+      "loss": 0.8908,
+      "step": 37025
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00010656727307012096,
+      "loss": 0.9531,
+      "step": 37030
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00010654229071067499,
+      "loss": 0.8834,
+      "step": 37035
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00010651730794115075,
+      "loss": 0.9505,
+      "step": 37040
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00010649232476311428,
+      "loss": 0.9008,
+      "step": 37045
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00010646734117813146,
+      "loss": 0.8971,
+      "step": 37050
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00010644235718776836,
+      "loss": 1.0166,
+      "step": 37055
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00010641737279359097,
+      "loss": 1.0054,
+      "step": 37060
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00010639238799716533,
+      "loss": 0.9391,
+      "step": 37065
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00010636740280005754,
+      "loss": 0.9439,
+      "step": 37070
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00010634241720383372,
+      "loss": 1.0102,
+      "step": 37075
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0001063174312100599,
+      "loss": 0.9331,
+      "step": 37080
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00010629244482030234,
+      "loss": 1.0283,
+      "step": 37085
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010626745803612717,
+      "loss": 0.8927,
+      "step": 37090
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001062424708591006,
+      "loss": 0.9256,
+      "step": 37095
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010621748329078882,
+      "loss": 0.9316,
+      "step": 37100
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0001061924953327581,
+      "loss": 0.7805,
+      "step": 37105
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00010616750698657474,
+      "loss": 0.9389,
+      "step": 37110
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00010614251825380504,
+      "loss": 0.9466,
+      "step": 37115
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00010611752913601525,
+      "loss": 1.0314,
+      "step": 37120
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00010609253963477182,
+      "loss": 0.903,
+      "step": 37125
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010606754975164103,
+      "loss": 0.8865,
+      "step": 37130
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00010604255948818931,
+      "loss": 0.8133,
+      "step": 37135
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0001060175688459831,
+      "loss": 0.925,
+      "step": 37140
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010599257782658882,
+      "loss": 1.0755,
+      "step": 37145
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.00010596758643157294,
+      "loss": 1.0568,
+      "step": 37150
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010594259466250194,
+      "loss": 1.0214,
+      "step": 37155
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001059176025209423,
+      "loss": 0.9462,
+      "step": 37160
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00010589261000846065,
+      "loss": 0.9693,
+      "step": 37165
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00010586761712662345,
+      "loss": 1.0498,
+      "step": 37170
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00010584262387699737,
+      "loss": 1.0542,
+      "step": 37175
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00010581763026114894,
+      "loss": 0.9571,
+      "step": 37180
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010579263628064484,
+      "loss": 1.0827,
+      "step": 37185
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00010576764193705167,
+      "loss": 0.9859,
+      "step": 37190
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00010574264723193617,
+      "loss": 1.0388,
+      "step": 37195
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010571765216686498,
+      "loss": 0.9622,
+      "step": 37200
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00010569265674340485,
+      "loss": 0.8824,
+      "step": 37205
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010566766096312251,
+      "loss": 0.8486,
+      "step": 37210
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0001056426648275847,
+      "loss": 1.0735,
+      "step": 37215
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00010561766833835826,
+      "loss": 0.9757,
+      "step": 37220
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010559267149700998,
+      "loss": 0.9932,
+      "step": 37225
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00010556767430510665,
+      "loss": 0.9202,
+      "step": 37230
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00010554267676421518,
+      "loss": 0.8795,
+      "step": 37235
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0001055176788759024,
+      "loss": 0.8951,
+      "step": 37240
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00010549268064173523,
+      "loss": 0.9054,
+      "step": 37245
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001054676820632806,
+      "loss": 0.8105,
+      "step": 37250
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010544268314210541,
+      "loss": 1.041,
+      "step": 37255
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00010541768387977664,
+      "loss": 1.0327,
+      "step": 37260
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00010539268427786129,
+      "loss": 0.8874,
+      "step": 37265
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00010536768433792632,
+      "loss": 0.9032,
+      "step": 37270
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001053426840615388,
+      "loss": 0.9239,
+      "step": 37275
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00010531768345026576,
+      "loss": 0.9904,
+      "step": 37280
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00010529268250567427,
+      "loss": 0.8399,
+      "step": 37285
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00010526768122933142,
+      "loss": 0.9121,
+      "step": 37290
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001052426796228043,
+      "loss": 0.9694,
+      "step": 37295
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.69140625,
+      "learning_rate": 0.00010521767768766002,
+      "loss": 0.8676,
+      "step": 37300
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.0001051926754254658,
+      "loss": 0.9611,
+      "step": 37305
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010516767283778877,
+      "loss": 1.1587,
+      "step": 37310
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001051426699261961,
+      "loss": 0.9311,
+      "step": 37315
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00010511766669225505,
+      "loss": 0.9896,
+      "step": 37320
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00010509266313753279,
+      "loss": 0.9138,
+      "step": 37325
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010506765926359661,
+      "loss": 0.8587,
+      "step": 37330
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00010504265507201378,
+      "loss": 0.999,
+      "step": 37335
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001050176505643516,
+      "loss": 0.8369,
+      "step": 37340
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00010499264574217738,
+      "loss": 0.9698,
+      "step": 37345
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0001049676406070584,
+      "loss": 1.0164,
+      "step": 37350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00010494263516056206,
+      "loss": 1.0847,
+      "step": 37355
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010491762940425576,
+      "loss": 0.9465,
+      "step": 37360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001048926233397068,
+      "loss": 0.9101,
+      "step": 37365
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00010486761696848263,
+      "loss": 0.9026,
+      "step": 37370
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00010484261029215073,
+      "loss": 1.0973,
+      "step": 37375
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010481760331227845,
+      "loss": 1.0309,
+      "step": 37380
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00010479259603043336,
+      "loss": 0.9052,
+      "step": 37385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010476758844818286,
+      "loss": 0.9943,
+      "step": 37390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00010474258056709449,
+      "loss": 0.9287,
+      "step": 37395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00010471757238873578,
+      "loss": 1.0231,
+      "step": 37400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010469256391467424,
+      "loss": 1.0086,
+      "step": 37405
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00010466755514647749,
+      "loss": 0.9275,
+      "step": 37410
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00010464254608571304,
+      "loss": 1.0473,
+      "step": 37415
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0001046175367339485,
+      "loss": 0.9569,
+      "step": 37420
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010459252709275152,
+      "loss": 1.0617,
+      "step": 37425
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00010456751716368971,
+      "loss": 0.9215,
+      "step": 37430
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.68359375,
+      "learning_rate": 0.0001045425069483307,
+      "loss": 0.9922,
+      "step": 37435
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010451749644824222,
+      "loss": 0.9558,
+      "step": 37440
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00010449248566499188,
+      "loss": 0.8627,
+      "step": 37445
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00010446747460014743,
+      "loss": 1.0908,
+      "step": 37450
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0001044424632552766,
+      "loss": 0.9257,
+      "step": 37455
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00010441745163194709,
+      "loss": 0.8122,
+      "step": 37460
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010439243973172673,
+      "loss": 1.0586,
+      "step": 37465
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0001043674275561832,
+      "loss": 0.9671,
+      "step": 37470
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.65234375,
+      "learning_rate": 0.00010434241510688434,
+      "loss": 0.9491,
+      "step": 37475
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00010431740238539796,
+      "loss": 0.9244,
+      "step": 37480
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00010429238939329189,
+      "loss": 0.8859,
+      "step": 37485
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010426737613213395,
+      "loss": 0.9897,
+      "step": 37490
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00010424236260349203,
+      "loss": 0.9511,
+      "step": 37495
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00010421734880893396,
+      "loss": 0.8938,
+      "step": 37500
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001041923347500277,
+      "loss": 0.9602,
+      "step": 37505
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00010416732042834112,
+      "loss": 0.9406,
+      "step": 37510
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001041423058454421,
+      "loss": 0.9569,
+      "step": 37515
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0001041172910028987,
+      "loss": 0.8252,
+      "step": 37520
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00010409227590227879,
+      "loss": 1.0166,
+      "step": 37525
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00010406726054515035,
+      "loss": 0.9833,
+      "step": 37530
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00010404224493308139,
+      "loss": 1.1919,
+      "step": 37535
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00010401722906763993,
+      "loss": 0.8367,
+      "step": 37540
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00010399221295039396,
+      "loss": 1.0228,
+      "step": 37545
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8125,
+      "learning_rate": 0.00010396719658291155,
+      "loss": 1.0634,
+      "step": 37550
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001039421799667607,
+      "loss": 0.9437,
+      "step": 37555
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010391716310350957,
+      "loss": 0.9024,
+      "step": 37560
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00010389214599472617,
+      "loss": 0.9552,
+      "step": 37565
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00010386712864197863,
+      "loss": 1.1085,
+      "step": 37570
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00010384211104683508,
+      "loss": 0.8691,
+      "step": 37575
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00010381709321086361,
+      "loss": 1.0081,
+      "step": 37580
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00010379207513563239,
+      "loss": 0.951,
+      "step": 37585
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00010376705682270958,
+      "loss": 1.0732,
+      "step": 37590
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00010374203827366338,
+      "loss": 1.0202,
+      "step": 37595
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00010371701949006195,
+      "loss": 0.9479,
+      "step": 37600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0001036920004734735,
+      "loss": 0.8367,
+      "step": 37605
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00010366698122546623,
+      "loss": 0.891,
+      "step": 37610
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010364196174760845,
+      "loss": 0.8879,
+      "step": 37615
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00010361694204146833,
+      "loss": 1.0175,
+      "step": 37620
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00010359192210861417,
+      "loss": 0.9633,
+      "step": 37625
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00010356690195061424,
+      "loss": 1.1454,
+      "step": 37630
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00010354188156903686,
+      "loss": 0.8291,
+      "step": 37635
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00010351686096545026,
+      "loss": 0.9924,
+      "step": 37640
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010349184014142284,
+      "loss": 0.8741,
+      "step": 37645
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0001034668190985229,
+      "loss": 0.8813,
+      "step": 37650
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0001034417978383188,
+      "loss": 0.822,
+      "step": 37655
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00010341677636237887,
+      "loss": 0.9509,
+      "step": 37660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001033917546722715,
+      "loss": 0.81,
+      "step": 37665
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0001033667327695651,
+      "loss": 0.9395,
+      "step": 37670
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00010334171065582807,
+      "loss": 0.8975,
+      "step": 37675
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0001033166883326288,
+      "loss": 0.9838,
+      "step": 37680
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00010329166580153573,
+      "loss": 0.7565,
+      "step": 37685
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00010326664306411727,
+      "loss": 0.9554,
+      "step": 37690
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00010324162012194194,
+      "loss": 1.0028,
+      "step": 37695
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00010321659697657818,
+      "loss": 0.9018,
+      "step": 37700
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0001031915736295944,
+      "loss": 0.943,
+      "step": 37705
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0001031665500825592,
+      "loss": 0.8955,
+      "step": 37710
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00010314152633704102,
+      "loss": 1.1076,
+      "step": 37715
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010311650239460834,
+      "loss": 1.0408,
+      "step": 37720
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00010309147825682982,
+      "loss": 0.8527,
+      "step": 37725
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010306645392527388,
+      "loss": 0.9127,
+      "step": 37730
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00010304142940150913,
+      "loss": 0.8388,
+      "step": 37735
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00010301640468710412,
+      "loss": 1.0603,
+      "step": 37740
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001029913797836274,
+      "loss": 1.1031,
+      "step": 37745
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00010296635469264764,
+      "loss": 0.9181,
+      "step": 37750
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00010294132941573338,
+      "loss": 0.8225,
+      "step": 37755
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001029163039544532,
+      "loss": 1.0889,
+      "step": 37760
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00010289127831037579,
+      "loss": 0.8887,
+      "step": 37765
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010286625248506979,
+      "loss": 1.0263,
+      "step": 37770
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00010284122648010377,
+      "loss": 0.9872,
+      "step": 37775
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010281620029704649,
+      "loss": 0.977,
+      "step": 37780
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0001027911739374665,
+      "loss": 1.1302,
+      "step": 37785
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00010276614740293262,
+      "loss": 1.0363,
+      "step": 37790
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00010274112069501344,
+      "loss": 0.9864,
+      "step": 37795
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00010271609381527767,
+      "loss": 0.8938,
+      "step": 37800
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001026910667652941,
+      "loss": 0.9837,
+      "step": 37805
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00010266603954663136,
+      "loss": 1.0605,
+      "step": 37810
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00010264101216085821,
+      "loss": 0.9841,
+      "step": 37815
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00010261598460954345,
+      "loss": 0.8508,
+      "step": 37820
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7578125,
+      "learning_rate": 0.0001025909568942558,
+      "loss": 0.9525,
+      "step": 37825
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00010256592901656397,
+      "loss": 0.9262,
+      "step": 37830
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010254090097803685,
+      "loss": 0.94,
+      "step": 37835
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001025158727802431,
+      "loss": 0.8998,
+      "step": 37840
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010249084442475163,
+      "loss": 0.9269,
+      "step": 37845
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00010246581591313118,
+      "loss": 0.8688,
+      "step": 37850
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010244078724695055,
+      "loss": 1.0001,
+      "step": 37855
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00010241575842777864,
+      "loss": 0.8315,
+      "step": 37860
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00010239072945718422,
+      "loss": 0.882,
+      "step": 37865
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00010236570033673614,
+      "loss": 1.1545,
+      "step": 37870
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00010234067106800329,
+      "loss": 0.9951,
+      "step": 37875
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010231564165255452,
+      "loss": 0.9681,
+      "step": 37880
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00010229061209195867,
+      "loss": 0.9191,
+      "step": 37885
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010226558238778466,
+      "loss": 0.9133,
+      "step": 37890
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010224055254160134,
+      "loss": 0.9152,
+      "step": 37895
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00010221552255497767,
+      "loss": 1.0462,
+      "step": 37900
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010219049242948251,
+      "loss": 1.047,
+      "step": 37905
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00010216546216668479,
+      "loss": 1.0363,
+      "step": 37910
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00010214043176815344,
+      "loss": 1.0124,
+      "step": 37915
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010211540123545737,
+      "loss": 1.0344,
+      "step": 37920
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00010209037057016556,
+      "loss": 0.9466,
+      "step": 37925
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00010206533977384694,
+      "loss": 1.0074,
+      "step": 37930
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00010204030884807046,
+      "loss": 0.982,
+      "step": 37935
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00010201527779440509,
+      "loss": 0.8657,
+      "step": 37940
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00010199024661441985,
+      "loss": 1.0115,
+      "step": 37945
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010196521530968363,
+      "loss": 0.7979,
+      "step": 37950
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00010194018388176552,
+      "loss": 0.8499,
+      "step": 37955
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00010191515233223448,
+      "loss": 1.0006,
+      "step": 37960
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00010189012066265949,
+      "loss": 0.8995,
+      "step": 37965
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00010186508887460959,
+      "loss": 1.0087,
+      "step": 37970
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0001018400569696538,
+      "loss": 0.8561,
+      "step": 37975
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0001018150249493611,
+      "loss": 1.0744,
+      "step": 37980
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00010178999281530062,
+      "loss": 0.9463,
+      "step": 37985
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00010176496056904135,
+      "loss": 0.8571,
+      "step": 37990
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00010173992821215232,
+      "loss": 0.7871,
+      "step": 37995
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.64453125,
+      "learning_rate": 0.00010171489574620263,
+      "loss": 1.1391,
+      "step": 38000
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010168986317276128,
+      "loss": 1.0766,
+      "step": 38005
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00010166483049339741,
+      "loss": 0.9119,
+      "step": 38010
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00010163979770968008,
+      "loss": 0.8999,
+      "step": 38015
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010161476482317835,
+      "loss": 1.039,
+      "step": 38020
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00010158973183546132,
+      "loss": 1.0054,
+      "step": 38025
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010156469874809808,
+      "loss": 0.891,
+      "step": 38030
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00010153966556265775,
+      "loss": 1.0695,
+      "step": 38035
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010151463228070943,
+      "loss": 0.8359,
+      "step": 38040
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00010148959890382224,
+      "loss": 0.9787,
+      "step": 38045
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001014645654335653,
+      "loss": 0.9187,
+      "step": 38050
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00010143953187150772,
+      "loss": 0.9577,
+      "step": 38055
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00010141449821921862,
+      "loss": 1.0231,
+      "step": 38060
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00010138946447826718,
+      "loss": 0.936,
+      "step": 38065
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00010136443065022254,
+      "loss": 0.8882,
+      "step": 38070
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.00010133939673665382,
+      "loss": 0.8248,
+      "step": 38075
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0001013143627391302,
+      "loss": 0.9949,
+      "step": 38080
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00010128932865922078,
+      "loss": 1.0271,
+      "step": 38085
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.66796875,
+      "learning_rate": 0.00010126429449849482,
+      "loss": 0.9705,
+      "step": 38090
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010123926025852144,
+      "loss": 0.8858,
+      "step": 38095
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.00010121422594086978,
+      "loss": 0.9708,
+      "step": 38100
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.00010118919154710909,
+      "loss": 0.9252,
+      "step": 38105
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00010116415707880848,
+      "loss": 0.9082,
+      "step": 38110
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00010113912253753719,
+      "loss": 0.9346,
+      "step": 38115
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00010111408792486446,
+      "loss": 0.8487,
+      "step": 38120
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00010108905324235935,
+      "loss": 0.8382,
+      "step": 38125
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0001010640184915912,
+      "loss": 0.9948,
+      "step": 38130
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6953125,
+      "learning_rate": 0.00010103898367412913,
+      "loss": 1.1,
+      "step": 38135
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00010101394879154238,
+      "loss": 1.0851,
+      "step": 38140
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00010098891384540017,
+      "loss": 1.0185,
+      "step": 38145
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00010096387883727174,
+      "loss": 1.0252,
+      "step": 38150
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00010093884376872625,
+      "loss": 1.0436,
+      "step": 38155
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00010091380864133297,
+      "loss": 0.9197,
+      "step": 38160
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00010088877345666112,
+      "loss": 0.9213,
+      "step": 38165
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010086373821627995,
+      "loss": 1.0176,
+      "step": 38170
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0001008387029217587,
+      "loss": 0.9032,
+      "step": 38175
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00010081366757466655,
+      "loss": 0.9536,
+      "step": 38180
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010078863217657282,
+      "loss": 0.8605,
+      "step": 38185
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010076359672904673,
+      "loss": 1.0488,
+      "step": 38190
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0001007385612336575,
+      "loss": 1.0442,
+      "step": 38195
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00010071352569197446,
+      "loss": 1.0319,
+      "step": 38200
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0001006884901055668,
+      "loss": 0.9562,
+      "step": 38205
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00010066345447600375,
+      "loss": 0.9968,
+      "step": 38210
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00010063841880485467,
+      "loss": 0.9077,
+      "step": 38215
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00010061338309368875,
+      "loss": 1.001,
+      "step": 38220
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00010058834734407529,
+      "loss": 0.966,
+      "step": 38225
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00010056331155758356,
+      "loss": 0.8654,
+      "step": 38230
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010053827573578278,
+      "loss": 1.0531,
+      "step": 38235
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010051323988024229,
+      "loss": 0.9787,
+      "step": 38240
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00010048820399253134,
+      "loss": 0.9843,
+      "step": 38245
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00010046316807421918,
+      "loss": 0.9878,
+      "step": 38250
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00010043813212687516,
+      "loss": 0.8344,
+      "step": 38255
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00010041309615206851,
+      "loss": 0.8606,
+      "step": 38260
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.625,
+      "learning_rate": 0.00010038806015136851,
+      "loss": 1.0538,
+      "step": 38265
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00010036302412634446,
+      "loss": 0.8955,
+      "step": 38270
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00010033798807856565,
+      "loss": 1.0416,
+      "step": 38275
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00010031295200960136,
+      "loss": 0.88,
+      "step": 38280
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00010028791592102087,
+      "loss": 0.9125,
+      "step": 38285
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00010026287981439348,
+      "loss": 0.9274,
+      "step": 38290
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0001002378436912885,
+      "loss": 0.9349,
+      "step": 38295
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0001002128075532752,
+      "loss": 0.834,
+      "step": 38300
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00010018777140192288,
+      "loss": 0.9952,
+      "step": 38305
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00010016273523880084,
+      "loss": 0.9502,
+      "step": 38310
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.00010013769906547839,
+      "loss": 0.9224,
+      "step": 38315
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00010011266288352477,
+      "loss": 1.001,
+      "step": 38320
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00010008762669450931,
+      "loss": 0.9386,
+      "step": 38325
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00010006259050000133,
+      "loss": 0.8958,
+      "step": 38330
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00010003755430157012,
+      "loss": 0.8092,
+      "step": 38335
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00010001251810078493,
+      "loss": 0.9878,
+      "step": 38340
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.998748189921509e-05,
+      "loss": 1.0938,
+      "step": 38345
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.996244569842992e-05,
+      "loss": 0.9138,
+      "step": 38350
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.993740949999869e-05,
+      "loss": 0.9054,
+      "step": 38355
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.991237330549067e-05,
+      "loss": 0.9356,
+      "step": 38360
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.988733711647524e-05,
+      "loss": 0.8874,
+      "step": 38365
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.986230093452166e-05,
+      "loss": 0.9673,
+      "step": 38370
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.983726476119918e-05,
+      "loss": 0.9727,
+      "step": 38375
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.981222859807715e-05,
+      "loss": 0.9835,
+      "step": 38380
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.978719244672481e-05,
+      "loss": 0.9609,
+      "step": 38385
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.976215630871152e-05,
+      "loss": 0.9405,
+      "step": 38390
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.973712018560654e-05,
+      "loss": 0.8642,
+      "step": 38395
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.971208407897914e-05,
+      "loss": 1.0028,
+      "step": 38400
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.968704799039867e-05,
+      "loss": 0.9833,
+      "step": 38405
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.966201192143439e-05,
+      "loss": 0.991,
+      "step": 38410
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.963697587365555e-05,
+      "loss": 0.8559,
+      "step": 38415
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.96119398486315e-05,
+      "loss": 1.0733,
+      "step": 38420
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.958690384793154e-05,
+      "loss": 0.8843,
+      "step": 38425
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.956186787312488e-05,
+      "loss": 0.9691,
+      "step": 38430
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.953683192578083e-05,
+      "loss": 0.9404,
+      "step": 38435
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.951179600746868e-05,
+      "loss": 1.0152,
+      "step": 38440
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.948676011975773e-05,
+      "loss": 1.011,
+      "step": 38445
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.946172426421725e-05,
+      "loss": 1.028,
+      "step": 38450
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.49609375,
+      "learning_rate": 9.943668844241647e-05,
+      "loss": 1.0303,
+      "step": 38455
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.941165265592472e-05,
+      "loss": 0.9126,
+      "step": 38460
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.938661690631127e-05,
+      "loss": 0.991,
+      "step": 38465
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.936158119514533e-05,
+      "loss": 0.9697,
+      "step": 38470
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5,
+      "learning_rate": 9.933654552399628e-05,
+      "loss": 0.8526,
+      "step": 38475
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.931150989443325e-05,
+      "loss": 0.9171,
+      "step": 38480
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.928647430802558e-05,
+      "loss": 1.0696,
+      "step": 38485
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.926143876634252e-05,
+      "loss": 0.9272,
+      "step": 38490
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.923640327095329e-05,
+      "loss": 0.9579,
+      "step": 38495
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.921136782342719e-05,
+      "loss": 0.9735,
+      "step": 38500
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.918633242533347e-05,
+      "loss": 0.8839,
+      "step": 38505
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.916129707824133e-05,
+      "loss": 1.0197,
+      "step": 38510
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.913626178372006e-05,
+      "loss": 1.0029,
+      "step": 38515
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.911122654333889e-05,
+      "loss": 0.9344,
+      "step": 38520
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.908619135866704e-05,
+      "loss": 0.8733,
+      "step": 38525
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.906115623127381e-05,
+      "loss": 1.0772,
+      "step": 38530
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.90361211627283e-05,
+      "loss": 1.0062,
+      "step": 38535
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.901108615459986e-05,
+      "loss": 0.8779,
+      "step": 38540
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.898605120845766e-05,
+      "loss": 0.9713,
+      "step": 38545
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.896101632587089e-05,
+      "loss": 0.971,
+      "step": 38550
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.893598150840884e-05,
+      "loss": 1.0255,
+      "step": 38555
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.891094675764067e-05,
+      "loss": 0.9138,
+      "step": 38560
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.888591207513556e-05,
+      "loss": 0.9943,
+      "step": 38565
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.88608774624628e-05,
+      "loss": 0.8789,
+      "step": 38570
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.883584292119153e-05,
+      "loss": 1.0116,
+      "step": 38575
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.881080845289097e-05,
+      "loss": 0.8831,
+      "step": 38580
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.878577405913027e-05,
+      "loss": 1.0177,
+      "step": 38585
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.87607397414786e-05,
+      "loss": 1.1244,
+      "step": 38590
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.873570550150522e-05,
+      "loss": 0.9846,
+      "step": 38595
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.871067134077924e-05,
+      "loss": 0.9154,
+      "step": 38600
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.868563726086983e-05,
+      "loss": 1.0878,
+      "step": 38605
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.866060326334621e-05,
+      "loss": 0.8965,
+      "step": 38610
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.86355693497775e-05,
+      "loss": 0.8671,
+      "step": 38615
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.861053552173281e-05,
+      "loss": 0.8291,
+      "step": 38620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.858550178078137e-05,
+      "loss": 0.9788,
+      "step": 38625
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.85604681284923e-05,
+      "loss": 1.0503,
+      "step": 38630
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.853543456643475e-05,
+      "loss": 1.1138,
+      "step": 38635
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.851040109617777e-05,
+      "loss": 1.1198,
+      "step": 38640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.84853677192906e-05,
+      "loss": 0.9442,
+      "step": 38645
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.846033443734227e-05,
+      "loss": 0.863,
+      "step": 38650
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.843530125190194e-05,
+      "loss": 1.0133,
+      "step": 38655
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.841026816453869e-05,
+      "loss": 1.0073,
+      "step": 38660
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.838523517682166e-05,
+      "loss": 1.006,
+      "step": 38665
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.836020229031995e-05,
+      "loss": 0.9199,
+      "step": 38670
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.833516950660259e-05,
+      "loss": 1.0166,
+      "step": 38675
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.831013682723872e-05,
+      "loss": 0.9378,
+      "step": 38680
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.828510425379742e-05,
+      "loss": 1.1848,
+      "step": 38685
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.826007178784772e-05,
+      "loss": 0.9421,
+      "step": 38690
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.823503943095869e-05,
+      "loss": 0.9063,
+      "step": 38695
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.82100071846994e-05,
+      "loss": 0.9038,
+      "step": 38700
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.818497505063891e-05,
+      "loss": 1.0482,
+      "step": 38705
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.815994303034623e-05,
+      "loss": 1.0177,
+      "step": 38710
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.813491112539043e-05,
+      "loss": 0.9322,
+      "step": 38715
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.810987933734054e-05,
+      "loss": 0.85,
+      "step": 38720
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.808484766776556e-05,
+      "loss": 1.0295,
+      "step": 38725
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.805981611823448e-05,
+      "loss": 1.066,
+      "step": 38730
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.803478469031636e-05,
+      "loss": 1.0679,
+      "step": 38735
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.80097533855802e-05,
+      "loss": 0.9112,
+      "step": 38740
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.798472220559493e-05,
+      "loss": 1.1377,
+      "step": 38745
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.795969115192957e-05,
+      "loss": 0.955,
+      "step": 38750
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.48828125,
+      "learning_rate": 9.79346602261531e-05,
+      "loss": 0.9294,
+      "step": 38755
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.49609375,
+      "learning_rate": 9.790962942983447e-05,
+      "loss": 0.827,
+      "step": 38760
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.788459876454264e-05,
+      "loss": 0.9328,
+      "step": 38765
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.785956823184659e-05,
+      "loss": 0.9699,
+      "step": 38770
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.498046875,
+      "learning_rate": 9.783453783331524e-05,
+      "loss": 0.8915,
+      "step": 38775
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.780950757051749e-05,
+      "loss": 0.9012,
+      "step": 38780
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.778447744502234e-05,
+      "loss": 0.8871,
+      "step": 38785
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.775944745839867e-05,
+      "loss": 1.0527,
+      "step": 38790
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.773441761221538e-05,
+      "loss": 0.9303,
+      "step": 38795
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.770938790804138e-05,
+      "loss": 0.8257,
+      "step": 38800
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.768435834744552e-05,
+      "loss": 0.9429,
+      "step": 38805
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.765932893199673e-05,
+      "loss": 0.9351,
+      "step": 38810
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.763429966326387e-05,
+      "loss": 0.9168,
+      "step": 38815
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.760927054281579e-05,
+      "loss": 0.9941,
+      "step": 38820
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.758424157222138e-05,
+      "loss": 0.9234,
+      "step": 38825
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.755921275304945e-05,
+      "loss": 0.8665,
+      "step": 38830
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.753418408686883e-05,
+      "loss": 1.0219,
+      "step": 38835
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.750915557524838e-05,
+      "loss": 0.9246,
+      "step": 38840
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.46484375,
+      "learning_rate": 9.748412721975691e-05,
+      "loss": 1.0578,
+      "step": 38845
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.74590990219632e-05,
+      "loss": 0.8982,
+      "step": 38850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.743407098343604e-05,
+      "loss": 0.8119,
+      "step": 38855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.740904310574424e-05,
+      "loss": 1.0587,
+      "step": 38860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.738401539045656e-05,
+      "loss": 0.9207,
+      "step": 38865
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.73589878391418e-05,
+      "loss": 0.9301,
+      "step": 38870
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.733396045336865e-05,
+      "loss": 0.8845,
+      "step": 38875
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.730893323470593e-05,
+      "loss": 1.0785,
+      "step": 38880
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.728390618472232e-05,
+      "loss": 0.9223,
+      "step": 38885
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.725887930498657e-05,
+      "loss": 0.9278,
+      "step": 38890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.723385259706743e-05,
+      "loss": 1.0404,
+      "step": 38895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.72088260625335e-05,
+      "loss": 1.114,
+      "step": 38900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.718379970295356e-05,
+      "loss": 0.9147,
+      "step": 38905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.715877351989625e-05,
+      "loss": 0.9504,
+      "step": 38910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.713374751493024e-05,
+      "loss": 0.9138,
+      "step": 38915
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.710872168962422e-05,
+      "loss": 1.0339,
+      "step": 38920
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.708369604554681e-05,
+      "loss": 0.977,
+      "step": 38925
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.705867058426664e-05,
+      "loss": 0.9473,
+      "step": 38930
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.703364530735237e-05,
+      "loss": 0.8984,
+      "step": 38935
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.70086202163726e-05,
+      "loss": 0.961,
+      "step": 38940
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.698359531289588e-05,
+      "loss": 0.9424,
+      "step": 38945
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.695857059849092e-05,
+      "loss": 0.9191,
+      "step": 38950
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.693354607472613e-05,
+      "loss": 0.8682,
+      "step": 38955
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.690852174317021e-05,
+      "loss": 1.0405,
+      "step": 38960
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.486328125,
+      "learning_rate": 9.688349760539167e-05,
+      "loss": 1.0359,
+      "step": 38965
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.685847366295902e-05,
+      "loss": 0.9881,
+      "step": 38970
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5,
+      "learning_rate": 9.683344991744083e-05,
+      "loss": 0.9328,
+      "step": 38975
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.680842637040561e-05,
+      "loss": 0.8895,
+      "step": 38980
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.678340302342184e-05,
+      "loss": 0.9235,
+      "step": 38985
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.675837987805807e-05,
+      "loss": 1.0094,
+      "step": 38990
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.673335693588273e-05,
+      "loss": 0.9143,
+      "step": 38995
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.670833419846432e-05,
+      "loss": 0.9175,
+      "step": 39000
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.668331166737124e-05,
+      "loss": 0.8666,
+      "step": 39005
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.665828934417196e-05,
+      "loss": 0.9682,
+      "step": 39010
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.66332672304349e-05,
+      "loss": 0.9945,
+      "step": 39015
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.660824532772852e-05,
+      "loss": 0.9521,
+      "step": 39020
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.658322363762115e-05,
+      "loss": 0.969,
+      "step": 39025
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.625,
+      "learning_rate": 9.655820216168123e-05,
+      "loss": 0.9418,
+      "step": 39030
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.653318090147711e-05,
+      "loss": 0.9502,
+      "step": 39035
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.650815985857716e-05,
+      "loss": 0.9308,
+      "step": 39040
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.648313903454975e-05,
+      "loss": 1.0048,
+      "step": 39045
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.645811843096316e-05,
+      "loss": 1.0241,
+      "step": 39050
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.643309804938578e-05,
+      "loss": 0.8488,
+      "step": 39055
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.640807789138586e-05,
+      "loss": 0.9414,
+      "step": 39060
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.638305795853168e-05,
+      "loss": 0.888,
+      "step": 39065
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.635803825239158e-05,
+      "loss": 1.0104,
+      "step": 39070
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.633301877453378e-05,
+      "loss": 1.0704,
+      "step": 39075
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.630799952652651e-05,
+      "loss": 0.8919,
+      "step": 39080
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.628298050993806e-05,
+      "loss": 1.0273,
+      "step": 39085
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.625796172633664e-05,
+      "loss": 1.0297,
+      "step": 39090
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.623294317729042e-05,
+      "loss": 0.7903,
+      "step": 39095
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.620792486436762e-05,
+      "loss": 0.9657,
+      "step": 39100
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.618290678913642e-05,
+      "loss": 0.932,
+      "step": 39105
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.615788895316498e-05,
+      "loss": 1.0416,
+      "step": 39110
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.613287135802142e-05,
+      "loss": 0.8526,
+      "step": 39115
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.610785400527385e-05,
+      "loss": 0.9282,
+      "step": 39120
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.608283689649047e-05,
+      "loss": 0.9398,
+      "step": 39125
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.605782003323932e-05,
+      "loss": 0.8503,
+      "step": 39130
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.603280341708848e-05,
+      "loss": 0.8561,
+      "step": 39135
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.600778704960606e-05,
+      "loss": 1.0332,
+      "step": 39140
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.59827709323601e-05,
+      "loss": 0.9378,
+      "step": 39145
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.59577550669186e-05,
+      "loss": 0.9466,
+      "step": 39150
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.593273945484966e-05,
+      "loss": 0.9579,
+      "step": 39155
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.590772409772125e-05,
+      "loss": 0.9372,
+      "step": 39160
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.588270899710133e-05,
+      "loss": 0.8365,
+      "step": 39165
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.494140625,
+      "learning_rate": 9.58576941545579e-05,
+      "loss": 0.8629,
+      "step": 39170
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.583267957165891e-05,
+      "loss": 0.9038,
+      "step": 39175
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.580766524997232e-05,
+      "loss": 0.9896,
+      "step": 39180
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.578265119106605e-05,
+      "loss": 0.9699,
+      "step": 39185
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.575763739650798e-05,
+      "loss": 0.9685,
+      "step": 39190
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.573262386786607e-05,
+      "loss": 0.9458,
+      "step": 39195
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.570761060670814e-05,
+      "loss": 0.9869,
+      "step": 39200
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.568259761460205e-05,
+      "loss": 0.9797,
+      "step": 39205
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.565758489311572e-05,
+      "loss": 1.0223,
+      "step": 39210
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.563257244381683e-05,
+      "loss": 0.9996,
+      "step": 39215
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.560756026827333e-05,
+      "loss": 0.9554,
+      "step": 39220
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.558254836805293e-05,
+      "loss": 0.8866,
+      "step": 39225
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.555753674472342e-05,
+      "loss": 0.8414,
+      "step": 39230
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.553252539985258e-05,
+      "loss": 0.9681,
+      "step": 39235
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.550751433500814e-05,
+      "loss": 1.0373,
+      "step": 39240
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.54825035517578e-05,
+      "loss": 0.9632,
+      "step": 39245
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.54574930516693e-05,
+      "loss": 0.9035,
+      "step": 39250
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.543248283631031e-05,
+      "loss": 0.8037,
+      "step": 39255
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.540747290724848e-05,
+      "loss": 1.0928,
+      "step": 39260
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.538246326605154e-05,
+      "loss": 1.0113,
+      "step": 39265
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.535745391428699e-05,
+      "loss": 0.8858,
+      "step": 39270
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.533244485352255e-05,
+      "loss": 0.9432,
+      "step": 39275
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.530743608532577e-05,
+      "loss": 0.859,
+      "step": 39280
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.528242761126424e-05,
+      "loss": 0.9839,
+      "step": 39285
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.525741943290552e-05,
+      "loss": 1.0186,
+      "step": 39290
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.523241155181716e-05,
+      "loss": 0.8606,
+      "step": 39295
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.520740396956665e-05,
+      "loss": 0.9145,
+      "step": 39300
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.518239668772154e-05,
+      "loss": 0.9839,
+      "step": 39305
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.515738970784928e-05,
+      "loss": 0.8016,
+      "step": 39310
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.513238303151739e-05,
+      "loss": 0.9934,
+      "step": 39315
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.510737666029323e-05,
+      "loss": 0.8347,
+      "step": 39320
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.508237059574429e-05,
+      "loss": 0.9688,
+      "step": 39325
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.498046875,
+      "learning_rate": 9.505736483943795e-05,
+      "loss": 1.0757,
+      "step": 39330
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.503235939294163e-05,
+      "loss": 0.9708,
+      "step": 39335
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.500735425782266e-05,
+      "loss": 0.8411,
+      "step": 39340
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.49823494356484e-05,
+      "loss": 1.0257,
+      "step": 39345
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.474609375,
+      "learning_rate": 9.495734492798623e-05,
+      "loss": 0.885,
+      "step": 39350
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.493234073640339e-05,
+      "loss": 1.0719,
+      "step": 39355
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.490733686246722e-05,
+      "loss": 0.9743,
+      "step": 39360
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.488233330774497e-05,
+      "loss": 0.8661,
+      "step": 39365
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.485733007380395e-05,
+      "loss": 0.955,
+      "step": 39370
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.483232716221127e-05,
+      "loss": 0.8181,
+      "step": 39375
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.480732457453422e-05,
+      "loss": 0.9923,
+      "step": 39380
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.478232231234e-05,
+      "loss": 1.033,
+      "step": 39385
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.475732037719572e-05,
+      "loss": 0.8905,
+      "step": 39390
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.625,
+      "learning_rate": 9.473231877066861e-05,
+      "loss": 0.9827,
+      "step": 39395
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.470731749432574e-05,
+      "loss": 0.9783,
+      "step": 39400
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.468231654973425e-05,
+      "loss": 0.9423,
+      "step": 39405
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.46573159384612e-05,
+      "loss": 0.9276,
+      "step": 39410
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.463231566207368e-05,
+      "loss": 0.992,
+      "step": 39415
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.460731572213875e-05,
+      "loss": 0.9216,
+      "step": 39420
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.45823161202234e-05,
+      "loss": 0.95,
+      "step": 39425
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.455731685789461e-05,
+      "loss": 0.9028,
+      "step": 39430
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.453231793671944e-05,
+      "loss": 0.9488,
+      "step": 39435
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.450731935826479e-05,
+      "loss": 1.054,
+      "step": 39440
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.494140625,
+      "learning_rate": 9.44823211240976e-05,
+      "loss": 0.9122,
+      "step": 39445
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.445732323578484e-05,
+      "loss": 0.877,
+      "step": 39450
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.443232569489337e-05,
+      "loss": 0.9831,
+      "step": 39455
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.440732850299003e-05,
+      "loss": 1.0377,
+      "step": 39460
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.439453125,
+      "learning_rate": 9.438233166164175e-05,
+      "loss": 1.0571,
+      "step": 39465
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.43573351724153e-05,
+      "loss": 1.04,
+      "step": 39470
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.433233903687754e-05,
+      "loss": 1.0514,
+      "step": 39475
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.43073432565952e-05,
+      "loss": 0.8964,
+      "step": 39480
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.428234783313505e-05,
+      "loss": 1.0734,
+      "step": 39485
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.425735276806387e-05,
+      "loss": 0.8907,
+      "step": 39490
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.423235806294835e-05,
+      "loss": 0.9568,
+      "step": 39495
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.420736371935519e-05,
+      "loss": 0.9042,
+      "step": 39500
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.418236973885107e-05,
+      "loss": 0.8978,
+      "step": 39505
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.415737612300266e-05,
+      "loss": 0.9227,
+      "step": 39510
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.413238287337653e-05,
+      "loss": 1.0013,
+      "step": 39515
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.410738999153936e-05,
+      "loss": 0.8893,
+      "step": 39520
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.408239747905771e-05,
+      "loss": 0.8983,
+      "step": 39525
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.498046875,
+      "learning_rate": 9.405740533749811e-05,
+      "loss": 0.9171,
+      "step": 39530
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.403241356842711e-05,
+      "loss": 0.9745,
+      "step": 39535
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.40074221734112e-05,
+      "loss": 1.0574,
+      "step": 39540
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.398243115401693e-05,
+      "loss": 1.023,
+      "step": 39545
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.39574405118107e-05,
+      "loss": 0.9277,
+      "step": 39550
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.393245024835898e-05,
+      "loss": 1.0421,
+      "step": 39555
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.39074603652282e-05,
+      "loss": 0.939,
+      "step": 39560
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.388247086398475e-05,
+      "loss": 0.826,
+      "step": 39565
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.385748174619497e-05,
+      "loss": 0.9839,
+      "step": 39570
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.46484375,
+      "learning_rate": 9.383249301342524e-05,
+      "loss": 0.9497,
+      "step": 39575
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.380750466724192e-05,
+      "loss": 0.9422,
+      "step": 39580
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.378251670921122e-05,
+      "loss": 0.7498,
+      "step": 39585
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.375752914089946e-05,
+      "loss": 1.0139,
+      "step": 39590
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.490234375,
+      "learning_rate": 9.373254196387286e-05,
+      "loss": 0.8377,
+      "step": 39595
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.370755517969768e-05,
+      "loss": 0.8605,
+      "step": 39600
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.368256878994012e-05,
+      "loss": 0.8669,
+      "step": 39605
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.365758279616631e-05,
+      "loss": 0.9177,
+      "step": 39610
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.363259719994247e-05,
+      "loss": 0.9787,
+      "step": 39615
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.360761200283468e-05,
+      "loss": 0.948,
+      "step": 39620
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.358262720640903e-05,
+      "loss": 1.0305,
+      "step": 39625
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.355764281223168e-05,
+      "loss": 0.9977,
+      "step": 39630
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.353265882186855e-05,
+      "loss": 0.7942,
+      "step": 39635
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.625,
+      "learning_rate": 9.350767523688574e-05,
+      "loss": 0.902,
+      "step": 39640
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.348269205884926e-05,
+      "loss": 0.8388,
+      "step": 39645
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.345770928932505e-05,
+      "loss": 1.0024,
+      "step": 39650
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.343272692987908e-05,
+      "loss": 0.8947,
+      "step": 39655
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.458984375,
+      "learning_rate": 9.340774498207726e-05,
+      "loss": 1.0304,
+      "step": 39660
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.338276344748548e-05,
+      "loss": 0.9795,
+      "step": 39665
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.335778232766964e-05,
+      "loss": 1.0649,
+      "step": 39670
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.333280162419558e-05,
+      "loss": 0.9476,
+      "step": 39675
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.330782133862907e-05,
+      "loss": 1.0906,
+      "step": 39680
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.328284147253601e-05,
+      "loss": 0.9704,
+      "step": 39685
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.325786202748203e-05,
+      "loss": 0.9293,
+      "step": 39690
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.323288300503296e-05,
+      "loss": 1.004,
+      "step": 39695
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.32079044067545e-05,
+      "loss": 0.9423,
+      "step": 39700
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.31829262342123e-05,
+      "loss": 0.9661,
+      "step": 39705
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.315794848897207e-05,
+      "loss": 0.8995,
+      "step": 39710
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.313297117259941e-05,
+      "loss": 0.9261,
+      "step": 39715
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.310799428665992e-05,
+      "loss": 0.9131,
+      "step": 39720
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.308301783271923e-05,
+      "loss": 1.0272,
+      "step": 39725
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.305804181234286e-05,
+      "loss": 0.9359,
+      "step": 39730
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.484375,
+      "learning_rate": 9.303306622709636e-05,
+      "loss": 1.0216,
+      "step": 39735
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8828125,
+      "learning_rate": 9.300809107854517e-05,
+      "loss": 0.9969,
+      "step": 39740
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.298311636825477e-05,
+      "loss": 1.057,
+      "step": 39745
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4921875,
+      "learning_rate": 9.295814209779066e-05,
+      "loss": 0.9669,
+      "step": 39750
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.293316826871821e-05,
+      "loss": 0.7833,
+      "step": 39755
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.290819488260281e-05,
+      "loss": 0.8582,
+      "step": 39760
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.288322194100985e-05,
+      "loss": 0.994,
+      "step": 39765
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.285824944550465e-05,
+      "loss": 1.081,
+      "step": 39770
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.283327739765248e-05,
+      "loss": 0.9136,
+      "step": 39775
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.280830579901867e-05,
+      "loss": 1.1627,
+      "step": 39780
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.278333465116844e-05,
+      "loss": 1.0138,
+      "step": 39785
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.275836395566703e-05,
+      "loss": 0.9436,
+      "step": 39790
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.27333937140796e-05,
+      "loss": 0.8221,
+      "step": 39795
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.270842392797131e-05,
+      "loss": 0.9202,
+      "step": 39800
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.268345459890734e-05,
+      "loss": 0.9205,
+      "step": 39805
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.265848572845275e-05,
+      "loss": 0.8917,
+      "step": 39810
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.263351731817263e-05,
+      "loss": 0.9382,
+      "step": 39815
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.828125,
+      "learning_rate": 9.260854936963205e-05,
+      "loss": 0.9602,
+      "step": 39820
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.2583581884396e-05,
+      "loss": 0.9374,
+      "step": 39825
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.255861486402949e-05,
+      "loss": 1.0302,
+      "step": 39830
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.253364831009748e-05,
+      "loss": 0.7649,
+      "step": 39835
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.250868222416493e-05,
+      "loss": 0.8398,
+      "step": 39840
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.248371660779666e-05,
+      "loss": 0.9781,
+      "step": 39845
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.245875146255763e-05,
+      "loss": 1.0139,
+      "step": 39850
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.24337867900126e-05,
+      "loss": 0.8359,
+      "step": 39855
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.625,
+      "learning_rate": 9.240882259172647e-05,
+      "loss": 1.1231,
+      "step": 39860
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.238385886926397e-05,
+      "loss": 0.8595,
+      "step": 39865
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.235889562418986e-05,
+      "loss": 1.133,
+      "step": 39870
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.48828125,
+      "learning_rate": 9.233393285806888e-05,
+      "loss": 0.9109,
+      "step": 39875
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.230897057246574e-05,
+      "loss": 0.924,
+      "step": 39880
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.48828125,
+      "learning_rate": 9.228400876894506e-05,
+      "loss": 0.8275,
+      "step": 39885
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.22590474490715e-05,
+      "loss": 0.8016,
+      "step": 39890
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.22340866144097e-05,
+      "loss": 0.9906,
+      "step": 39895
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.220912626652417e-05,
+      "loss": 0.9459,
+      "step": 39900
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.218416640697947e-05,
+      "loss": 0.92,
+      "step": 39905
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.215920703734012e-05,
+      "loss": 1.059,
+      "step": 39910
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.213424815917062e-05,
+      "loss": 1.1374,
+      "step": 39915
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.21092897740354e-05,
+      "loss": 1.0304,
+      "step": 39920
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.208433188349885e-05,
+      "loss": 0.8611,
+      "step": 39925
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.205937448912543e-05,
+      "loss": 0.9039,
+      "step": 39930
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.203441759247946e-05,
+      "loss": 0.8528,
+      "step": 39935
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.200946119512523e-05,
+      "loss": 0.764,
+      "step": 39940
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.198450529862714e-05,
+      "loss": 0.9309,
+      "step": 39945
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.195954990454934e-05,
+      "loss": 1.0455,
+      "step": 39950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.193459501445611e-05,
+      "loss": 1.173,
+      "step": 39955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.190964062991166e-05,
+      "loss": 0.7986,
+      "step": 39960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.188468675248014e-05,
+      "loss": 0.8808,
+      "step": 39965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.18597333837257e-05,
+      "loss": 0.862,
+      "step": 39970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.183478052521244e-05,
+      "loss": 0.893,
+      "step": 39975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.180982817850443e-05,
+      "loss": 1.0495,
+      "step": 39980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.178487634516573e-05,
+      "loss": 0.9471,
+      "step": 39985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.175992502676035e-05,
+      "loss": 0.9318,
+      "step": 39990
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.173497422485222e-05,
+      "loss": 0.956,
+      "step": 39995
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.171002394100539e-05,
+      "loss": 0.9119,
+      "step": 40000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.168507417678364e-05,
+      "loss": 0.9412,
+      "step": 40005
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.166012493375091e-05,
+      "loss": 1.026,
+      "step": 40010
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.163517621347106e-05,
+      "loss": 1.0485,
+      "step": 40015
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.161022801750788e-05,
+      "loss": 0.9197,
+      "step": 40020
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.158528034742518e-05,
+      "loss": 0.9262,
+      "step": 40025
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.156033320478669e-05,
+      "loss": 0.9726,
+      "step": 40030
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5,
+      "learning_rate": 9.15353865911561e-05,
+      "loss": 0.8281,
+      "step": 40035
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.151044050809715e-05,
+      "loss": 0.9731,
+      "step": 40040
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5,
+      "learning_rate": 9.148549495717344e-05,
+      "loss": 1.0655,
+      "step": 40045
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.146054993994864e-05,
+      "loss": 1.0064,
+      "step": 40050
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.143560545798625e-05,
+      "loss": 0.9116,
+      "step": 40055
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.141066151284988e-05,
+      "loss": 0.9974,
+      "step": 40060
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.138571810610303e-05,
+      "loss": 0.9558,
+      "step": 40065
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.136077523930918e-05,
+      "loss": 0.9115,
+      "step": 40070
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.133583291403176e-05,
+      "loss": 1.0983,
+      "step": 40075
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.131089113183422e-05,
+      "loss": 1.028,
+      "step": 40080
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.128594989427992e-05,
+      "loss": 0.9602,
+      "step": 40085
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.126100920293219e-05,
+      "loss": 0.9662,
+      "step": 40090
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.123606905935436e-05,
+      "loss": 0.944,
+      "step": 40095
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.121112946510973e-05,
+      "loss": 0.9506,
+      "step": 40100
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.484375,
+      "learning_rate": 9.11861904217615e-05,
+      "loss": 0.9151,
+      "step": 40105
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.73828125,
+      "learning_rate": 9.116125193087288e-05,
+      "loss": 0.8144,
+      "step": 40110
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.113631399400707e-05,
+      "loss": 0.9939,
+      "step": 40115
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.11113766127272e-05,
+      "loss": 0.9773,
+      "step": 40120
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.108643978859632e-05,
+      "loss": 0.9812,
+      "step": 40125
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.10615035231776e-05,
+      "loss": 1.0516,
+      "step": 40130
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.103656781803399e-05,
+      "loss": 1.0237,
+      "step": 40135
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.101163267472852e-05,
+      "loss": 0.969,
+      "step": 40140
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.098669809482415e-05,
+      "loss": 1.0175,
+      "step": 40145
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.096176407988382e-05,
+      "loss": 0.9175,
+      "step": 40150
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.625,
+      "learning_rate": 9.093683063147044e-05,
+      "loss": 1.1157,
+      "step": 40155
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.09118977511468e-05,
+      "loss": 1.0574,
+      "step": 40160
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.088696544047574e-05,
+      "loss": 0.8854,
+      "step": 40165
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.08620337010201e-05,
+      "loss": 0.8979,
+      "step": 40170
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.083710253434259e-05,
+      "loss": 1.1431,
+      "step": 40175
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.08121719420059e-05,
+      "loss": 0.9464,
+      "step": 40180
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.078724192557278e-05,
+      "loss": 0.9012,
+      "step": 40185
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.07623124866058e-05,
+      "loss": 0.9206,
+      "step": 40190
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.073738362666759e-05,
+      "loss": 0.9284,
+      "step": 40195
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.071245534732073e-05,
+      "loss": 0.864,
+      "step": 40200
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.068752765012777e-05,
+      "loss": 0.9126,
+      "step": 40205
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5,
+      "learning_rate": 9.066260053665119e-05,
+      "loss": 0.9227,
+      "step": 40210
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.063767400845343e-05,
+      "loss": 1.0837,
+      "step": 40215
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.061274806709688e-05,
+      "loss": 1.0174,
+      "step": 40220
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.058782271414402e-05,
+      "loss": 0.8127,
+      "step": 40225
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.056289795115715e-05,
+      "loss": 0.84,
+      "step": 40230
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.053797377969855e-05,
+      "loss": 1.0014,
+      "step": 40235
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.625,
+      "learning_rate": 9.051305020133055e-05,
+      "loss": 1.0487,
+      "step": 40240
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.048812721761538e-05,
+      "loss": 0.9631,
+      "step": 40245
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.04632048301152e-05,
+      "loss": 0.9664,
+      "step": 40250
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.043828304039222e-05,
+      "loss": 0.8728,
+      "step": 40255
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.49609375,
+      "learning_rate": 9.041336185000858e-05,
+      "loss": 0.9733,
+      "step": 40260
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.038844126052633e-05,
+      "loss": 1.0681,
+      "step": 40265
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.03635212735075e-05,
+      "loss": 0.8276,
+      "step": 40270
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.033860189051412e-05,
+      "loss": 0.979,
+      "step": 40275
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.031368311310821e-05,
+      "loss": 1.0012,
+      "step": 40280
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.028876494285166e-05,
+      "loss": 0.8916,
+      "step": 40285
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.026384738130637e-05,
+      "loss": 1.0582,
+      "step": 40290
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.023893043003423e-05,
+      "loss": 0.9211,
+      "step": 40295
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.021401409059704e-05,
+      "loss": 0.902,
+      "step": 40300
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.018909836455659e-05,
+      "loss": 0.9903,
+      "step": 40305
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.016418325347464e-05,
+      "loss": 0.8672,
+      "step": 40310
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4921875,
+      "learning_rate": 9.013926875891291e-05,
+      "loss": 0.9729,
+      "step": 40315
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.49609375,
+      "learning_rate": 9.011435488243302e-05,
+      "loss": 0.8721,
+      "step": 40320
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.008944162559663e-05,
+      "loss": 0.9681,
+      "step": 40325
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.494140625,
+      "learning_rate": 9.006452898996529e-05,
+      "loss": 0.9349,
+      "step": 40330
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.003961697710062e-05,
+      "loss": 1.0404,
+      "step": 40335
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.001470558856411e-05,
+      "loss": 0.9361,
+      "step": 40340
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.482421875,
+      "learning_rate": 8.998979482591718e-05,
+      "loss": 0.8448,
+      "step": 40345
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.996488469072136e-05,
+      "loss": 1.0374,
+      "step": 40350
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.993997518453799e-05,
+      "loss": 0.797,
+      "step": 40355
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.99150663089284e-05,
+      "loss": 0.7895,
+      "step": 40360
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.989015806545402e-05,
+      "loss": 0.9464,
+      "step": 40365
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.986525045567597e-05,
+      "loss": 0.9364,
+      "step": 40370
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.984034348115558e-05,
+      "loss": 0.9559,
+      "step": 40375
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.981543714345404e-05,
+      "loss": 1.109,
+      "step": 40380
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.979053144413248e-05,
+      "loss": 1.0065,
+      "step": 40385
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.976562638475206e-05,
+      "loss": 1.0162,
+      "step": 40390
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.974072196687384e-05,
+      "loss": 1.0205,
+      "step": 40395
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.971581819205881e-05,
+      "loss": 1.1037,
+      "step": 40400
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.969091506186804e-05,
+      "loss": 0.8065,
+      "step": 40405
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.966601257786246e-05,
+      "loss": 0.9767,
+      "step": 40410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.964111074160296e-05,
+      "loss": 1.0224,
+      "step": 40415
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.961620955465049e-05,
+      "loss": 0.9845,
+      "step": 40420
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.959130901856576e-05,
+      "loss": 1.135,
+      "step": 40425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.956640913490968e-05,
+      "loss": 1.0147,
+      "step": 40430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.954150990524294e-05,
+      "loss": 1.1181,
+      "step": 40435
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.951661133112625e-05,
+      "loss": 0.8177,
+      "step": 40440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.949171341412034e-05,
+      "loss": 0.9945,
+      "step": 40445
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.946681615578578e-05,
+      "loss": 1.0107,
+      "step": 40450
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.944191955768317e-05,
+      "loss": 0.9096,
+      "step": 40455
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.941702362137309e-05,
+      "loss": 1.0481,
+      "step": 40460
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.939212834841602e-05,
+      "loss": 0.885,
+      "step": 40465
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.936723374037245e-05,
+      "loss": 0.8906,
+      "step": 40470
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.46875,
+      "learning_rate": 8.934233979880276e-05,
+      "loss": 1.0109,
+      "step": 40475
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.931744652526731e-05,
+      "loss": 1.0306,
+      "step": 40480
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.929255392132652e-05,
+      "loss": 0.7733,
+      "step": 40485
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.926766198854063e-05,
+      "loss": 0.9697,
+      "step": 40490
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.92427707284699e-05,
+      "loss": 0.9151,
+      "step": 40495
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.625,
+      "learning_rate": 8.921788014267456e-05,
+      "loss": 1.1148,
+      "step": 40500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.919299023271479e-05,
+      "loss": 0.9216,
+      "step": 40505
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.916810100015066e-05,
+      "loss": 0.8619,
+      "step": 40510
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.914321244654233e-05,
+      "loss": 1.042,
+      "step": 40515
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.91183245734498e-05,
+      "loss": 0.9905,
+      "step": 40520
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.909343738243312e-05,
+      "loss": 0.8287,
+      "step": 40525
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.906855087505217e-05,
+      "loss": 0.9968,
+      "step": 40530
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.904366505286687e-05,
+      "loss": 0.944,
+      "step": 40535
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.901877991743717e-05,
+      "loss": 0.8778,
+      "step": 40540
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.899389547032283e-05,
+      "loss": 0.9997,
+      "step": 40545
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.896901171308364e-05,
+      "loss": 1.0573,
+      "step": 40550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.894412864727937e-05,
+      "loss": 0.8705,
+      "step": 40555
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.891924627446974e-05,
+      "loss": 0.9807,
+      "step": 40560
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4609375,
+      "learning_rate": 8.889436459621432e-05,
+      "loss": 1.0162,
+      "step": 40565
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.886948361407281e-05,
+      "loss": 0.9592,
+      "step": 40570
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.884460332960477e-05,
+      "loss": 1.0517,
+      "step": 40575
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.881972374436967e-05,
+      "loss": 1.0649,
+      "step": 40580
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.879484485992702e-05,
+      "loss": 0.85,
+      "step": 40585
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.876996667783624e-05,
+      "loss": 1.0682,
+      "step": 40590
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.874508919965675e-05,
+      "loss": 1.009,
+      "step": 40595
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.872021242694789e-05,
+      "loss": 0.9226,
+      "step": 40600
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.869533636126895e-05,
+      "loss": 0.8566,
+      "step": 40605
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5,
+      "learning_rate": 8.867046100417921e-05,
+      "loss": 0.949,
+      "step": 40610
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.864558635723787e-05,
+      "loss": 0.9686,
+      "step": 40615
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.862071242200411e-05,
+      "loss": 0.9619,
+      "step": 40620
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.859583920003707e-05,
+      "loss": 0.9039,
+      "step": 40625
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.857096669289583e-05,
+      "loss": 0.9528,
+      "step": 40630
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.85460949021394e-05,
+      "loss": 0.912,
+      "step": 40635
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.852122382932678e-05,
+      "loss": 1.0178,
+      "step": 40640
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.849635347601692e-05,
+      "loss": 0.906,
+      "step": 40645
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.847148384376872e-05,
+      "loss": 0.9614,
+      "step": 40650
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.844661493414106e-05,
+      "loss": 0.9905,
+      "step": 40655
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.84217467486927e-05,
+      "loss": 0.8828,
+      "step": 40660
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.839687928898248e-05,
+      "loss": 1.0715,
+      "step": 40665
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.837201255656907e-05,
+      "loss": 0.8497,
+      "step": 40670
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.834714655301114e-05,
+      "loss": 0.9762,
+      "step": 40675
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.832228127986734e-05,
+      "loss": 0.9128,
+      "step": 40680
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.82974167386963e-05,
+      "loss": 1.2287,
+      "step": 40685
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.827255293105646e-05,
+      "loss": 1.03,
+      "step": 40690
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.824768985850638e-05,
+      "loss": 1.0054,
+      "step": 40695
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4921875,
+      "learning_rate": 8.822282752260445e-05,
+      "loss": 0.8711,
+      "step": 40700
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.482421875,
+      "learning_rate": 8.819796592490912e-05,
+      "loss": 0.969,
+      "step": 40705
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.817310506697873e-05,
+      "loss": 1.0671,
+      "step": 40710
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.814824495037157e-05,
+      "loss": 0.9447,
+      "step": 40715
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.812338557664594e-05,
+      "loss": 1.0268,
+      "step": 40720
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.809852694736001e-05,
+      "loss": 1.0194,
+      "step": 40725
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.807366906407196e-05,
+      "loss": 0.9151,
+      "step": 40730
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9609375,
+      "learning_rate": 8.804881192833996e-05,
+      "loss": 0.7921,
+      "step": 40735
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.802395554172201e-05,
+      "loss": 1.011,
+      "step": 40740
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.799909990577617e-05,
+      "loss": 0.915,
+      "step": 40745
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.79742450220604e-05,
+      "loss": 1.0229,
+      "step": 40750
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.794939089213265e-05,
+      "loss": 0.9477,
+      "step": 40755
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.792453751755081e-05,
+      "loss": 0.7819,
+      "step": 40760
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.789968489987273e-05,
+      "loss": 0.9665,
+      "step": 40765
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4921875,
+      "learning_rate": 8.787483304065613e-05,
+      "loss": 0.7705,
+      "step": 40770
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.784998194145885e-05,
+      "loss": 0.9689,
+      "step": 40775
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.782513160383852e-05,
+      "loss": 0.8291,
+      "step": 40780
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.625,
+      "learning_rate": 8.780028202935281e-05,
+      "loss": 0.9454,
+      "step": 40785
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.777543321955934e-05,
+      "loss": 0.9423,
+      "step": 40790
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.775058517601563e-05,
+      "loss": 1.0121,
+      "step": 40795
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.772573790027919e-05,
+      "loss": 0.9232,
+      "step": 40800
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.770089139390745e-05,
+      "loss": 1.001,
+      "step": 40805
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.767604565845787e-05,
+      "loss": 0.929,
+      "step": 40810
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.76512006954878e-05,
+      "loss": 0.9996,
+      "step": 40815
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.76263565065545e-05,
+      "loss": 0.9141,
+      "step": 40820
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.760151309321527e-05,
+      "loss": 1.0298,
+      "step": 40825
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5,
+      "learning_rate": 8.757667045702733e-05,
+      "loss": 0.8709,
+      "step": 40830
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.755182859954784e-05,
+      "loss": 0.9799,
+      "step": 40835
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.752698752233394e-05,
+      "loss": 1.0642,
+      "step": 40840
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.490234375,
+      "learning_rate": 8.750214722694261e-05,
+      "loss": 0.9883,
+      "step": 40845
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.747730771493095e-05,
+      "loss": 0.9238,
+      "step": 40850
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.74524689878559e-05,
+      "loss": 0.9287,
+      "step": 40855
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.742763104727439e-05,
+      "loss": 0.9518,
+      "step": 40860
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.740279389474327e-05,
+      "loss": 0.8663,
+      "step": 40865
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.737795753181939e-05,
+      "loss": 0.9363,
+      "step": 40870
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.735312196005949e-05,
+      "loss": 1.1332,
+      "step": 40875
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.732828718102033e-05,
+      "loss": 1.0223,
+      "step": 40880
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.730345319625857e-05,
+      "loss": 1.0087,
+      "step": 40885
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.727862000733081e-05,
+      "loss": 0.9222,
+      "step": 40890
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.725378761579366e-05,
+      "loss": 1.0914,
+      "step": 40895
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.722895602320358e-05,
+      "loss": 1.0208,
+      "step": 40900
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.720412523111709e-05,
+      "loss": 0.8234,
+      "step": 40905
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.717929524109062e-05,
+      "loss": 0.8737,
+      "step": 40910
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.48828125,
+      "learning_rate": 8.71544660546805e-05,
+      "loss": 1.0277,
+      "step": 40915
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.71296376734431e-05,
+      "loss": 0.9144,
+      "step": 40920
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.710481009893467e-05,
+      "loss": 0.9951,
+      "step": 40925
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6953125,
+      "learning_rate": 8.70799833327114e-05,
+      "loss": 0.9455,
+      "step": 40930
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.705515737632952e-05,
+      "loss": 0.9778,
+      "step": 40935
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.703033223134511e-05,
+      "loss": 0.9631,
+      "step": 40940
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.47265625,
+      "learning_rate": 8.700550789931428e-05,
+      "loss": 0.8185,
+      "step": 40945
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.698068438179297e-05,
+      "loss": 1.0634,
+      "step": 40950
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.695586168033719e-05,
+      "loss": 1.0236,
+      "step": 40955
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.693103979650288e-05,
+      "loss": 1.0162,
+      "step": 40960
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.690621873184587e-05,
+      "loss": 0.9561,
+      "step": 40965
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.688139848792196e-05,
+      "loss": 1.0784,
+      "step": 40970
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.474609375,
+      "learning_rate": 8.685657906628696e-05,
+      "loss": 0.8949,
+      "step": 40975
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.683176046849655e-05,
+      "loss": 0.9578,
+      "step": 40980
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.46875,
+      "learning_rate": 8.680694269610637e-05,
+      "loss": 0.8533,
+      "step": 40985
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.678212575067207e-05,
+      "loss": 0.8894,
+      "step": 40990
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.675730963374918e-05,
+      "loss": 0.9729,
+      "step": 40995
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.673249434689322e-05,
+      "loss": 0.9695,
+      "step": 41000
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.670767989165962e-05,
+      "loss": 0.8899,
+      "step": 41005
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.668286626960375e-05,
+      "loss": 0.868,
+      "step": 41010
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.6658053482281e-05,
+      "loss": 0.9941,
+      "step": 41015
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.663324153124667e-05,
+      "loss": 0.9316,
+      "step": 41020
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.660843041805597e-05,
+      "loss": 0.8675,
+      "step": 41025
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.658362014426411e-05,
+      "loss": 0.8628,
+      "step": 41030
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.655881071142622e-05,
+      "loss": 0.8342,
+      "step": 41035
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.653400212109737e-05,
+      "loss": 0.9824,
+      "step": 41040
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.650919437483262e-05,
+      "loss": 1.1338,
+      "step": 41045
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.648438747418694e-05,
+      "loss": 0.9745,
+      "step": 41050
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.645958142071524e-05,
+      "loss": 0.9639,
+      "step": 41055
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.64347762159724e-05,
+      "loss": 1.0613,
+      "step": 41060
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.64099718615132e-05,
+      "loss": 0.9962,
+      "step": 41065
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.638516835889248e-05,
+      "loss": 0.8651,
+      "step": 41070
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.636036570966491e-05,
+      "loss": 0.8809,
+      "step": 41075
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.633556391538511e-05,
+      "loss": 1.0232,
+      "step": 41080
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.631076297760776e-05,
+      "loss": 0.9977,
+      "step": 41085
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.628596289788738e-05,
+      "loss": 1.0693,
+      "step": 41090
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.626116367777845e-05,
+      "loss": 0.9539,
+      "step": 41095
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.623636531883545e-05,
+      "loss": 0.894,
+      "step": 41100
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.621156782261275e-05,
+      "loss": 1.0753,
+      "step": 41105
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.618677119066468e-05,
+      "loss": 0.8618,
+      "step": 41110
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.616197542454551e-05,
+      "loss": 1.0041,
+      "step": 41115
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.613718052580946e-05,
+      "loss": 0.9626,
+      "step": 41120
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.611238649601075e-05,
+      "loss": 0.7566,
+      "step": 41125
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.608759333670349e-05,
+      "loss": 0.9487,
+      "step": 41130
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.606280104944168e-05,
+      "loss": 0.9074,
+      "step": 41135
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.603800963577939e-05,
+      "loss": 1.0181,
+      "step": 41140
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.601321909727056e-05,
+      "loss": 0.8415,
+      "step": 41145
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.598842943546908e-05,
+      "loss": 0.891,
+      "step": 41150
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.596364065192885e-05,
+      "loss": 0.9422,
+      "step": 41155
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.593885274820356e-05,
+      "loss": 0.9747,
+      "step": 41160
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.591406572584701e-05,
+      "loss": 0.9683,
+      "step": 41165
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.625,
+      "learning_rate": 8.588927958641286e-05,
+      "loss": 0.9289,
+      "step": 41170
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.586449433145472e-05,
+      "loss": 0.94,
+      "step": 41175
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.583970996252618e-05,
+      "loss": 1.0362,
+      "step": 41180
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.581492648118077e-05,
+      "loss": 1.1214,
+      "step": 41185
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.579014388897188e-05,
+      "loss": 0.9097,
+      "step": 41190
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.576536218745299e-05,
+      "loss": 0.9254,
+      "step": 41195
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.57405813781774e-05,
+      "loss": 1.0014,
+      "step": 41200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.57158014626984e-05,
+      "loss": 1.0573,
+      "step": 41205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.569102244256929e-05,
+      "loss": 0.91,
+      "step": 41210
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.566624431934312e-05,
+      "loss": 0.9182,
+      "step": 41215
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.564146709457312e-05,
+      "loss": 0.9758,
+      "step": 41220
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.56166907698123e-05,
+      "loss": 0.7203,
+      "step": 41225
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.559191534661367e-05,
+      "loss": 0.8752,
+      "step": 41230
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.556714082653021e-05,
+      "loss": 1.0124,
+      "step": 41235
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.55423672111148e-05,
+      "loss": 0.9686,
+      "step": 41240
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.551759450192026e-05,
+      "loss": 1.0235,
+      "step": 41245
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.549282270049941e-05,
+      "loss": 1.0041,
+      "step": 41250
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.546805180840497e-05,
+      "loss": 0.9878,
+      "step": 41255
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.54432818271896e-05,
+      "loss": 0.8734,
+      "step": 41260
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.541851275840589e-05,
+      "loss": 0.9486,
+      "step": 41265
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.539374460360638e-05,
+      "loss": 0.7125,
+      "step": 41270
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.536897736434362e-05,
+      "loss": 0.9791,
+      "step": 41275
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.72265625,
+      "learning_rate": 8.534421104217001e-05,
+      "loss": 1.0311,
+      "step": 41280
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.531944563863794e-05,
+      "loss": 0.9974,
+      "step": 41285
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.529468115529975e-05,
+      "loss": 0.9587,
+      "step": 41290
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.52699175937077e-05,
+      "loss": 0.9041,
+      "step": 41295
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.524515495541395e-05,
+      "loss": 0.8812,
+      "step": 41300
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.522039324197073e-05,
+      "loss": 0.9394,
+      "step": 41305
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.51956324549301e-05,
+      "loss": 0.9483,
+      "step": 41310
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.517087259584409e-05,
+      "loss": 0.8897,
+      "step": 41315
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.514611366626466e-05,
+      "loss": 0.9513,
+      "step": 41320
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.512135566774372e-05,
+      "loss": 0.9343,
+      "step": 41325
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.50965986018332e-05,
+      "loss": 0.8736,
+      "step": 41330
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.507184247008482e-05,
+      "loss": 1.0342,
+      "step": 41335
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.504708727405035e-05,
+      "loss": 0.7705,
+      "step": 41340
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.502233301528149e-05,
+      "loss": 1.098,
+      "step": 41345
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.499757969532986e-05,
+      "loss": 0.855,
+      "step": 41350
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.497282731574701e-05,
+      "loss": 0.9717,
+      "step": 41355
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.494807587808447e-05,
+      "loss": 0.9026,
+      "step": 41360
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.49233253838937e-05,
+      "loss": 0.8608,
+      "step": 41365
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.489857583472604e-05,
+      "loss": 0.8898,
+      "step": 41370
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.487382723213284e-05,
+      "loss": 1.023,
+      "step": 41375
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.484907957766537e-05,
+      "loss": 0.8763,
+      "step": 41380
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.482433287287485e-05,
+      "loss": 0.9427,
+      "step": 41385
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.479958711931245e-05,
+      "loss": 0.8177,
+      "step": 41390
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.477484231852921e-05,
+      "loss": 0.9392,
+      "step": 41395
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.475009847207622e-05,
+      "loss": 0.895,
+      "step": 41400
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.472535558150441e-05,
+      "loss": 0.9946,
+      "step": 41405
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.470061364836472e-05,
+      "loss": 0.9943,
+      "step": 41410
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.467587267420797e-05,
+      "loss": 0.9164,
+      "step": 41415
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.465113266058503e-05,
+      "loss": 0.9155,
+      "step": 41420
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.462639360904653e-05,
+      "loss": 0.9499,
+      "step": 41425
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.460165552114321e-05,
+      "loss": 0.8514,
+      "step": 41430
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.451171875,
+      "learning_rate": 8.457691839842563e-05,
+      "loss": 0.8172,
+      "step": 41435
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.455218224244439e-05,
+      "loss": 0.9282,
+      "step": 41440
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.452744705474995e-05,
+      "loss": 0.9488,
+      "step": 41445
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.450271283689274e-05,
+      "loss": 0.9748,
+      "step": 41450
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.447797959042317e-05,
+      "loss": 0.8702,
+      "step": 41455
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.44532473168915e-05,
+      "loss": 0.9403,
+      "step": 41460
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.442851601784798e-05,
+      "loss": 0.8392,
+      "step": 41465
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.440378569484286e-05,
+      "loss": 0.9365,
+      "step": 41470
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.437905634942618e-05,
+      "loss": 0.9095,
+      "step": 41475
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.435432798314805e-05,
+      "loss": 1.0542,
+      "step": 41480
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.432960059755842e-05,
+      "loss": 0.9465,
+      "step": 41485
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.430487419420732e-05,
+      "loss": 0.9138,
+      "step": 41490
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.428014877464457e-05,
+      "loss": 0.9988,
+      "step": 41495
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.425542434041999e-05,
+      "loss": 0.983,
+      "step": 41500
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.423070089308333e-05,
+      "loss": 0.9907,
+      "step": 41505
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.420597843418432e-05,
+      "loss": 0.8086,
+      "step": 41510
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.418125696527256e-05,
+      "loss": 1.0055,
+      "step": 41515
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.415653648789762e-05,
+      "loss": 0.8668,
+      "step": 41520
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.76171875,
+      "learning_rate": 8.413181700360905e-05,
+      "loss": 1.0207,
+      "step": 41525
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.410709851395626e-05,
+      "loss": 0.8962,
+      "step": 41530
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.408238102048861e-05,
+      "loss": 0.8318,
+      "step": 41535
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.466796875,
+      "learning_rate": 8.405766452475544e-05,
+      "loss": 0.8854,
+      "step": 41540
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.403294902830603e-05,
+      "loss": 0.9743,
+      "step": 41545
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.400823453268956e-05,
+      "loss": 0.8932,
+      "step": 41550
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.398352103945515e-05,
+      "loss": 0.9847,
+      "step": 41555
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.39588085501519e-05,
+      "loss": 0.8523,
+      "step": 41560
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.39340970663288e-05,
+      "loss": 0.9778,
+      "step": 41565
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.390938658953478e-05,
+      "loss": 0.9594,
+      "step": 41570
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.486328125,
+      "learning_rate": 8.388467712131877e-05,
+      "loss": 0.8426,
+      "step": 41575
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.38599686632295e-05,
+      "loss": 1.0202,
+      "step": 41580
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.383526121681581e-05,
+      "loss": 1.0363,
+      "step": 41585
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.486328125,
+      "learning_rate": 8.381055478362635e-05,
+      "loss": 0.9128,
+      "step": 41590
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.378584936520973e-05,
+      "loss": 0.9713,
+      "step": 41595
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.376114496311456e-05,
+      "loss": 0.9081,
+      "step": 41600
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.373644157888933e-05,
+      "loss": 0.993,
+      "step": 41605
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.371173921408243e-05,
+      "loss": 0.9647,
+      "step": 41610
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.368703787024229e-05,
+      "loss": 1.16,
+      "step": 41615
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.36623375489172e-05,
+      "loss": 0.8755,
+      "step": 41620
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.363763825165538e-05,
+      "loss": 1.1381,
+      "step": 41625
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.48046875,
+      "learning_rate": 8.361293998000509e-05,
+      "loss": 0.9973,
+      "step": 41630
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.35882427355143e-05,
+      "loss": 0.9467,
+      "step": 41635
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.35635465197312e-05,
+      "loss": 0.8742,
+      "step": 41640
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.353885133420372e-05,
+      "loss": 0.9813,
+      "step": 41645
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.351415718047976e-05,
+      "loss": 1.0443,
+      "step": 41650
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.348946406010721e-05,
+      "loss": 0.9129,
+      "step": 41655
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.346477197463387e-05,
+      "loss": 0.938,
+      "step": 41660
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.344008092560743e-05,
+      "loss": 0.9234,
+      "step": 41665
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.341539091457561e-05,
+      "loss": 1.1227,
+      "step": 41670
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.625,
+      "learning_rate": 8.339070194308597e-05,
+      "loss": 0.9144,
+      "step": 41675
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.336601401268607e-05,
+      "loss": 0.9558,
+      "step": 41680
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.334132712492333e-05,
+      "loss": 0.8646,
+      "step": 41685
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.331664128134515e-05,
+      "loss": 1.0444,
+      "step": 41690
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.329195648349892e-05,
+      "loss": 0.9609,
+      "step": 41695
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.32672727329319e-05,
+      "loss": 1.0872,
+      "step": 41700
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.324259003119126e-05,
+      "loss": 1.0415,
+      "step": 41705
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.321790837982417e-05,
+      "loss": 0.9616,
+      "step": 41710
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.31932277803777e-05,
+      "loss": 0.7862,
+      "step": 41715
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.316854823439884e-05,
+      "loss": 0.9196,
+      "step": 41720
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.314386974343455e-05,
+      "loss": 0.8995,
+      "step": 41725
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.311919230903172e-05,
+      "loss": 0.9577,
+      "step": 41730
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.462890625,
+      "learning_rate": 8.309451593273717e-05,
+      "loss": 0.8472,
+      "step": 41735
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.306984061609758e-05,
+      "loss": 0.9422,
+      "step": 41740
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.304516636065964e-05,
+      "loss": 1.1044,
+      "step": 41745
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.302049316796999e-05,
+      "loss": 0.8926,
+      "step": 41750
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.486328125,
+      "learning_rate": 8.29958210395752e-05,
+      "loss": 0.8397,
+      "step": 41755
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.297114997702166e-05,
+      "loss": 0.9482,
+      "step": 41760
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.294647998185587e-05,
+      "loss": 0.8748,
+      "step": 41765
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.292181105562414e-05,
+      "loss": 0.9345,
+      "step": 41770
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.83984375,
+      "learning_rate": 8.289714319987271e-05,
+      "loss": 1.0754,
+      "step": 41775
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.287247641614785e-05,
+      "loss": 0.9052,
+      "step": 41780
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.28478107059957e-05,
+      "loss": 0.9907,
+      "step": 41785
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.282314607096228e-05,
+      "loss": 1.0787,
+      "step": 41790
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.279848251259364e-05,
+      "loss": 0.9882,
+      "step": 41795
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.277382003243567e-05,
+      "loss": 0.8283,
+      "step": 41800
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.274915863203432e-05,
+      "loss": 0.8574,
+      "step": 41805
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.272449831293534e-05,
+      "loss": 0.902,
+      "step": 41810
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.269983907668446e-05,
+      "loss": 0.965,
+      "step": 41815
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.26751809248274e-05,
+      "loss": 1.014,
+      "step": 41820
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.265052385890972e-05,
+      "loss": 0.947,
+      "step": 41825
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.262586788047696e-05,
+      "loss": 1.0511,
+      "step": 41830
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.26012129910746e-05,
+      "loss": 0.8874,
+      "step": 41835
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.257655919224804e-05,
+      "loss": 0.8345,
+      "step": 41840
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.255190648554257e-05,
+      "loss": 0.8518,
+      "step": 41845
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.252725487250349e-05,
+      "loss": 1.0714,
+      "step": 41850
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.250260435467594e-05,
+      "loss": 0.9246,
+      "step": 41855
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.247795493360511e-05,
+      "loss": 0.978,
+      "step": 41860
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.2453306610836e-05,
+      "loss": 0.8857,
+      "step": 41865
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.24286593879136e-05,
+      "loss": 0.9305,
+      "step": 41870
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.240401326638287e-05,
+      "loss": 0.8613,
+      "step": 41875
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.498046875,
+      "learning_rate": 8.237936824778863e-05,
+      "loss": 1.0857,
+      "step": 41880
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.235472433367563e-05,
+      "loss": 0.9642,
+      "step": 41885
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.233008152558868e-05,
+      "loss": 0.8736,
+      "step": 41890
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.230543982507227e-05,
+      "loss": 0.9844,
+      "step": 41895
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.228079923367107e-05,
+      "loss": 0.9585,
+      "step": 41900
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.225615975292955e-05,
+      "loss": 0.9325,
+      "step": 41905
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.223152138439213e-05,
+      "loss": 0.8639,
+      "step": 41910
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.220688412960321e-05,
+      "loss": 0.987,
+      "step": 41915
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.498046875,
+      "learning_rate": 8.218224799010704e-05,
+      "loss": 0.8088,
+      "step": 41920
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.215761296744786e-05,
+      "loss": 1.0435,
+      "step": 41925
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.213297906316983e-05,
+      "loss": 0.8773,
+      "step": 41930
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.210834627881702e-05,
+      "loss": 0.9031,
+      "step": 41935
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.208371461593341e-05,
+      "loss": 0.8389,
+      "step": 41940
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.205908407606306e-05,
+      "loss": 0.9602,
+      "step": 41945
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.203445466074968e-05,
+      "loss": 0.9552,
+      "step": 41950
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.200982637153715e-05,
+      "loss": 0.9129,
+      "step": 41955
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.198519920996921e-05,
+      "loss": 0.8813,
+      "step": 41960
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.625,
+      "learning_rate": 8.196057317758948e-05,
+      "loss": 1.0221,
+      "step": 41965
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.193594827594158e-05,
+      "loss": 0.9442,
+      "step": 41970
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5,
+      "learning_rate": 8.191132450656902e-05,
+      "loss": 0.9139,
+      "step": 41975
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.188670187101521e-05,
+      "loss": 1.1236,
+      "step": 41980
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.186208037082359e-05,
+      "loss": 0.8923,
+      "step": 41985
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.183746000753743e-05,
+      "loss": 1.0116,
+      "step": 41990
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.181284078269997e-05,
+      "loss": 0.8802,
+      "step": 41995
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.178822269785435e-05,
+      "loss": 1.0867,
+      "step": 42000
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.484375,
+      "learning_rate": 8.176360575454366e-05,
+      "loss": 0.9187,
+      "step": 42005
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.173898995431095e-05,
+      "loss": 1.1618,
+      "step": 42010
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.171437529869912e-05,
+      "loss": 0.993,
+      "step": 42015
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.168976178925109e-05,
+      "loss": 1.0539,
+      "step": 42020
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.166514942750964e-05,
+      "loss": 0.9547,
+      "step": 42025
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.45703125,
+      "learning_rate": 8.164053821501751e-05,
+      "loss": 0.7925,
+      "step": 42030
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.161592815331734e-05,
+      "loss": 0.8804,
+      "step": 42035
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.159131924395176e-05,
+      "loss": 1.0156,
+      "step": 42040
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.156671148846325e-05,
+      "loss": 0.914,
+      "step": 42045
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.154210488839426e-05,
+      "loss": 0.9318,
+      "step": 42050
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.151749944528714e-05,
+      "loss": 1.0271,
+      "step": 42055
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.14928951606842e-05,
+      "loss": 1.0947,
+      "step": 42060
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.146829203612769e-05,
+      "loss": 0.7624,
+      "step": 42065
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.144369007315973e-05,
+      "loss": 0.7011,
+      "step": 42070
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.141908927332239e-05,
+      "loss": 1.0747,
+      "step": 42075
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.139448963815772e-05,
+      "loss": 0.968,
+      "step": 42080
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.136989116920763e-05,
+      "loss": 1.0052,
+      "step": 42085
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.134529386801396e-05,
+      "loss": 1.0768,
+      "step": 42090
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.132069773611854e-05,
+      "loss": 1.1377,
+      "step": 42095
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.484375,
+      "learning_rate": 8.129610277506308e-05,
+      "loss": 0.9914,
+      "step": 42100
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.127150898638918e-05,
+      "loss": 1.1432,
+      "step": 42105
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.124691637163842e-05,
+      "loss": 0.873,
+      "step": 42110
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.122232493235228e-05,
+      "loss": 1.0254,
+      "step": 42115
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.119773467007222e-05,
+      "loss": 1.0591,
+      "step": 42120
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.117314558633956e-05,
+      "loss": 0.8329,
+      "step": 42125
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.114855768269556e-05,
+      "loss": 0.981,
+      "step": 42130
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.112397096068146e-05,
+      "loss": 0.9654,
+      "step": 42135
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.109938542183836e-05,
+      "loss": 0.9152,
+      "step": 42140
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.107480106770729e-05,
+      "loss": 1.0351,
+      "step": 42145
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.105021789982925e-05,
+      "loss": 1.0923,
+      "step": 42150
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.102563591974516e-05,
+      "loss": 0.9168,
+      "step": 42155
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.100105512899579e-05,
+      "loss": 0.9979,
+      "step": 42160
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.097647552912192e-05,
+      "loss": 0.9022,
+      "step": 42165
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.095189712166425e-05,
+      "loss": 1.0565,
+      "step": 42170
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.092731990816335e-05,
+      "loss": 0.8553,
+      "step": 42175
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.484375,
+      "learning_rate": 8.090274389015978e-05,
+      "loss": 0.9753,
+      "step": 42180
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.087816906919391e-05,
+      "loss": 0.9659,
+      "step": 42185
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.085359544680623e-05,
+      "loss": 0.9577,
+      "step": 42190
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.490234375,
+      "learning_rate": 8.082902302453701e-05,
+      "loss": 0.9535,
+      "step": 42195
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.080445180392642e-05,
+      "loss": 1.0207,
+      "step": 42200
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.07798817865147e-05,
+      "loss": 0.9898,
+      "step": 42205
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.075531297384185e-05,
+      "loss": 0.9406,
+      "step": 42210
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.07307453674479e-05,
+      "loss": 0.9575,
+      "step": 42215
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.070617896887277e-05,
+      "loss": 0.9358,
+      "step": 42220
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.068161377965633e-05,
+      "loss": 0.9085,
+      "step": 42225
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.065704980133833e-05,
+      "loss": 0.8574,
+      "step": 42230
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.063248703545847e-05,
+      "loss": 1.0217,
+      "step": 42235
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.06079254835564e-05,
+      "loss": 0.8214,
+      "step": 42240
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.058336514717164e-05,
+      "loss": 0.9639,
+      "step": 42245
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.055880602784367e-05,
+      "loss": 0.9374,
+      "step": 42250
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.053424812711186e-05,
+      "loss": 1.2145,
+      "step": 42255
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.050969144651559e-05,
+      "loss": 0.9234,
+      "step": 42260
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.048513598759403e-05,
+      "loss": 0.9074,
+      "step": 42265
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.046058175188638e-05,
+      "loss": 1.0246,
+      "step": 42270
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.043602874093169e-05,
+      "loss": 0.8926,
+      "step": 42275
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.041147695626902e-05,
+      "loss": 0.9339,
+      "step": 42280
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.49609375,
+      "learning_rate": 8.03869263994373e-05,
+      "loss": 0.9105,
+      "step": 42285
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.036237707197534e-05,
+      "loss": 0.8557,
+      "step": 42290
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.033782897542197e-05,
+      "loss": 0.9911,
+      "step": 42295
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5,
+      "learning_rate": 8.031328211131586e-05,
+      "loss": 0.8827,
+      "step": 42300
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.028873648119565e-05,
+      "loss": 0.8477,
+      "step": 42305
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.026419208659994e-05,
+      "loss": 0.8995,
+      "step": 42310
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.023964892906709e-05,
+      "loss": 0.9163,
+      "step": 42315
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.021510701013557e-05,
+      "loss": 1.0234,
+      "step": 42320
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.019056633134367e-05,
+      "loss": 0.9321,
+      "step": 42325
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.01660268942296e-05,
+      "loss": 0.9225,
+      "step": 42330
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.01414887003316e-05,
+      "loss": 0.9481,
+      "step": 42335
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.01169517511877e-05,
+      "loss": 0.9872,
+      "step": 42340
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.009241604833588e-05,
+      "loss": 0.9599,
+      "step": 42345
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.006788159331412e-05,
+      "loss": 0.9175,
+      "step": 42350
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.004334838766024e-05,
+      "loss": 0.9903,
+      "step": 42355
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.001881643291198e-05,
+      "loss": 0.9758,
+      "step": 42360
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.486328125,
+      "learning_rate": 7.999428573060715e-05,
+      "loss": 1.0384,
+      "step": 42365
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.99697562822832e-05,
+      "loss": 0.9855,
+      "step": 42370
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.994522808947775e-05,
+      "loss": 0.9536,
+      "step": 42375
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.992070115372825e-05,
+      "loss": 1.2132,
+      "step": 42380
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.989617547657206e-05,
+      "loss": 0.9809,
+      "step": 42385
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.98716510595465e-05,
+      "loss": 0.9123,
+      "step": 42390
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.984712790418878e-05,
+      "loss": 0.9471,
+      "step": 42395
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.412109375,
+      "learning_rate": 7.9822606012036e-05,
+      "loss": 0.8363,
+      "step": 42400
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.979808538462528e-05,
+      "loss": 0.961,
+      "step": 42405
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.977356602349359e-05,
+      "loss": 0.8457,
+      "step": 42410
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.974904793017784e-05,
+      "loss": 1.0536,
+      "step": 42415
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.972453110621479e-05,
+      "loss": 0.9591,
+      "step": 42420
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.970001555314121e-05,
+      "loss": 0.8937,
+      "step": 42425
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.967550127249381e-05,
+      "loss": 0.8458,
+      "step": 42430
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.965098826580912e-05,
+      "loss": 0.906,
+      "step": 42435
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.962647653462364e-05,
+      "loss": 1.0945,
+      "step": 42440
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.455078125,
+      "learning_rate": 7.960196608047385e-05,
+      "loss": 0.9566,
+      "step": 42445
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.498046875,
+      "learning_rate": 7.957745690489604e-05,
+      "loss": 0.898,
+      "step": 42450
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.955294900942646e-05,
+      "loss": 0.8699,
+      "step": 42455
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.952844239560138e-05,
+      "loss": 0.9584,
+      "step": 42460
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.950393706495683e-05,
+      "loss": 0.9459,
+      "step": 42465
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.947943301902885e-05,
+      "loss": 0.9533,
+      "step": 42470
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.945493025935338e-05,
+      "loss": 0.93,
+      "step": 42475
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.943042878746626e-05,
+      "loss": 0.8268,
+      "step": 42480
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.94059286049033e-05,
+      "loss": 0.8896,
+      "step": 42485
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.93814297132002e-05,
+      "loss": 1.0771,
+      "step": 42490
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.935693211389254e-05,
+      "loss": 0.8995,
+      "step": 42495
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.933243580851591e-05,
+      "loss": 0.9314,
+      "step": 42500
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.930794079860575e-05,
+      "loss": 0.8779,
+      "step": 42505
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.625,
+      "learning_rate": 7.92834470856974e-05,
+      "loss": 1.006,
+      "step": 42510
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.925895467132618e-05,
+      "loss": 0.8862,
+      "step": 42515
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.923446355702735e-05,
+      "loss": 0.9694,
+      "step": 42520
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.920997374433596e-05,
+      "loss": 0.8995,
+      "step": 42525
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.494140625,
+      "learning_rate": 7.91854852347871e-05,
+      "loss": 0.8734,
+      "step": 42530
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.916099802991572e-05,
+      "loss": 0.9556,
+      "step": 42535
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.913651213125672e-05,
+      "loss": 0.9399,
+      "step": 42540
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.90625,
+      "learning_rate": 7.911202754034491e-05,
+      "loss": 0.9976,
+      "step": 42545
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.908754425871498e-05,
+      "loss": 0.9188,
+      "step": 42550
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.906306228790161e-05,
+      "loss": 0.8477,
+      "step": 42555
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.903858162943935e-05,
+      "loss": 1.0648,
+      "step": 42560
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.901410228486264e-05,
+      "loss": 0.9438,
+      "step": 42565
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.898962425570592e-05,
+      "loss": 0.9425,
+      "step": 42570
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.89651475435035e-05,
+      "loss": 0.8806,
+      "step": 42575
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.498046875,
+      "learning_rate": 7.894067214978959e-05,
+      "loss": 0.9518,
+      "step": 42580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.891619807609832e-05,
+      "loss": 0.9417,
+      "step": 42585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.889172532396373e-05,
+      "loss": 0.8483,
+      "step": 42590
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.886725389491987e-05,
+      "loss": 0.8405,
+      "step": 42595
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.88427837905006e-05,
+      "loss": 0.9486,
+      "step": 42600
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.881831501223973e-05,
+      "loss": 1.03,
+      "step": 42605
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.879384756167101e-05,
+      "loss": 0.9571,
+      "step": 42610
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.876938144032807e-05,
+      "loss": 1.0589,
+      "step": 42615
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.874491664974448e-05,
+      "loss": 0.9998,
+      "step": 42620
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.872045319145376e-05,
+      "loss": 1.0316,
+      "step": 42625
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.869599106698922e-05,
+      "loss": 0.9804,
+      "step": 42630
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.867153027788424e-05,
+      "loss": 0.8879,
+      "step": 42635
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.864707082567204e-05,
+      "loss": 1.0132,
+      "step": 42640
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.862261271188574e-05,
+      "loss": 1.0564,
+      "step": 42645
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.859815593805844e-05,
+      "loss": 1.0339,
+      "step": 42650
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.85737005057231e-05,
+      "loss": 1.0806,
+      "step": 42655
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.854924641641258e-05,
+      "loss": 0.9391,
+      "step": 42660
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.498046875,
+      "learning_rate": 7.852479367165976e-05,
+      "loss": 0.8207,
+      "step": 42665
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.850034227299734e-05,
+      "loss": 0.9667,
+      "step": 42670
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.84758922219579e-05,
+      "loss": 0.9414,
+      "step": 42675
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4921875,
+      "learning_rate": 7.845144352007414e-05,
+      "loss": 0.9857,
+      "step": 42680
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.842699616887837e-05,
+      "loss": 0.956,
+      "step": 42685
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.840255016990308e-05,
+      "loss": 0.8266,
+      "step": 42690
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.837810552468053e-05,
+      "loss": 0.866,
+      "step": 42695
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.835366223474293e-05,
+      "loss": 0.9463,
+      "step": 42700
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.832922030162246e-05,
+      "loss": 0.8329,
+      "step": 42705
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.830477972685115e-05,
+      "loss": 0.9353,
+      "step": 42710
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.828034051196093e-05,
+      "loss": 0.9475,
+      "step": 42715
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.825590265848371e-05,
+      "loss": 0.9154,
+      "step": 42720
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.823146616795129e-05,
+      "loss": 1.0003,
+      "step": 42725
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.820703104189538e-05,
+      "loss": 0.8323,
+      "step": 42730
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.818259728184757e-05,
+      "loss": 0.8252,
+      "step": 42735
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.815816488933938e-05,
+      "loss": 1.0588,
+      "step": 42740
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.813373386590232e-05,
+      "loss": 0.917,
+      "step": 42745
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.810930421306772e-05,
+      "loss": 0.9622,
+      "step": 42750
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.486328125,
+      "learning_rate": 7.808487593236683e-05,
+      "loss": 0.8535,
+      "step": 42755
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.806044902533092e-05,
+      "loss": 1.0698,
+      "step": 42760
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.803602349349104e-05,
+      "loss": 0.9576,
+      "step": 42765
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.801159933837821e-05,
+      "loss": 1.0837,
+      "step": 42770
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.798717656152339e-05,
+      "loss": 0.8904,
+      "step": 42775
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.796275516445741e-05,
+      "loss": 0.9422,
+      "step": 42780
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.458984375,
+      "learning_rate": 7.793833514871106e-05,
+      "loss": 0.9295,
+      "step": 42785
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.791391651581497e-05,
+      "loss": 0.9939,
+      "step": 42790
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.788949926729972e-05,
+      "loss": 0.9126,
+      "step": 42795
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.786508340469586e-05,
+      "loss": 0.8445,
+      "step": 42800
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.78406689295338e-05,
+      "loss": 0.9759,
+      "step": 42805
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.462890625,
+      "learning_rate": 7.78162558433438e-05,
+      "loss": 1.0567,
+      "step": 42810
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.779184414765618e-05,
+      "loss": 0.862,
+      "step": 42815
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.776743384400106e-05,
+      "loss": 1.0773,
+      "step": 42820
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.77430249339085e-05,
+      "loss": 0.9642,
+      "step": 42825
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.771861741890848e-05,
+      "loss": 0.8702,
+      "step": 42830
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.769421130053094e-05,
+      "loss": 0.8947,
+      "step": 42835
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.766980658030562e-05,
+      "loss": 0.9315,
+      "step": 42840
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.764540325976225e-05,
+      "loss": 0.9786,
+      "step": 42845
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.762100134043043e-05,
+      "loss": 0.9207,
+      "step": 42850
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.759660082383977e-05,
+      "loss": 1.0167,
+      "step": 42855
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.757220171151967e-05,
+      "loss": 0.9736,
+      "step": 42860
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.47265625,
+      "learning_rate": 7.754780400499951e-05,
+      "loss": 1.0015,
+      "step": 42865
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5,
+      "learning_rate": 7.752340770580858e-05,
+      "loss": 0.9226,
+      "step": 42870
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.749901281547604e-05,
+      "loss": 0.9444,
+      "step": 42875
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.747461933553099e-05,
+      "loss": 1.0346,
+      "step": 42880
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.462890625,
+      "learning_rate": 7.745022726750248e-05,
+      "loss": 0.9532,
+      "step": 42885
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.742583661291943e-05,
+      "loss": 0.8445,
+      "step": 42890
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.740144737331062e-05,
+      "loss": 0.9205,
+      "step": 42895
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.48046875,
+      "learning_rate": 7.737705955020482e-05,
+      "loss": 0.9847,
+      "step": 42900
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.735267314513069e-05,
+      "loss": 0.9144,
+      "step": 42905
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.732828815961683e-05,
+      "loss": 1.0651,
+      "step": 42910
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.730390459519167e-05,
+      "loss": 0.9732,
+      "step": 42915
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.727952245338359e-05,
+      "loss": 0.9412,
+      "step": 42920
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.725514173572096e-05,
+      "loss": 1.1264,
+      "step": 42925
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.723076244373195e-05,
+      "loss": 0.9375,
+      "step": 42930
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.720638457894465e-05,
+      "loss": 0.9551,
+      "step": 42935
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.48046875,
+      "learning_rate": 7.718200814288717e-05,
+      "loss": 1.0087,
+      "step": 42940
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.715763313708739e-05,
+      "loss": 1.0652,
+      "step": 42945
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.713325956307316e-05,
+      "loss": 0.963,
+      "step": 42950
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.710888742237226e-05,
+      "loss": 0.8619,
+      "step": 42955
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.708451671651239e-05,
+      "loss": 0.9818,
+      "step": 42960
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.70601474470211e-05,
+      "loss": 0.9524,
+      "step": 42965
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.703577961542588e-05,
+      "loss": 1.0222,
+      "step": 42970
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.701141322325418e-05,
+      "loss": 1.0135,
+      "step": 42975
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.698704827203326e-05,
+      "loss": 0.8863,
+      "step": 42980
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.696268476329038e-05,
+      "loss": 0.9803,
+      "step": 42985
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.693832269855264e-05,
+      "loss": 1.1359,
+      "step": 42990
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.691396207934711e-05,
+      "loss": 0.9429,
+      "step": 42995
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.688960290720073e-05,
+      "loss": 0.996,
+      "step": 43000
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.686524518364036e-05,
+      "loss": 0.9694,
+      "step": 43005
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.828125,
+      "learning_rate": 7.684088891019276e-05,
+      "loss": 1.0616,
+      "step": 43010
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.681653408838462e-05,
+      "loss": 0.8888,
+      "step": 43015
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.679218071974254e-05,
+      "loss": 1.0106,
+      "step": 43020
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.676782880579298e-05,
+      "loss": 0.9764,
+      "step": 43025
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.674347834806238e-05,
+      "loss": 0.9293,
+      "step": 43030
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.671912934807706e-05,
+      "loss": 1.0437,
+      "step": 43035
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.669478180736319e-05,
+      "loss": 0.8764,
+      "step": 43040
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.76953125,
+      "learning_rate": 7.667043572744701e-05,
+      "loss": 1.0054,
+      "step": 43045
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.664609110985444e-05,
+      "loss": 0.9659,
+      "step": 43050
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.662174795611147e-05,
+      "loss": 0.8469,
+      "step": 43055
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.659740626774399e-05,
+      "loss": 1.0004,
+      "step": 43060
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.482421875,
+      "learning_rate": 7.65730660462777e-05,
+      "loss": 0.8755,
+      "step": 43065
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.654872729323834e-05,
+      "loss": 0.9929,
+      "step": 43070
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.652439001015145e-05,
+      "loss": 0.9484,
+      "step": 43075
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.650005419854252e-05,
+      "loss": 1.0401,
+      "step": 43080
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.647571985993697e-05,
+      "loss": 1.1313,
+      "step": 43085
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.645138699586013e-05,
+      "loss": 0.8608,
+      "step": 43090
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.64270556078371e-05,
+      "loss": 1.0649,
+      "step": 43095
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.640272569739316e-05,
+      "loss": 1.0124,
+      "step": 43100
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.478515625,
+      "learning_rate": 7.637839726605318e-05,
+      "loss": 0.8626,
+      "step": 43105
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.635407031534218e-05,
+      "loss": 0.8512,
+      "step": 43110
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.632974484678499e-05,
+      "loss": 1.1224,
+      "step": 43115
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.630542086190633e-05,
+      "loss": 0.9001,
+      "step": 43120
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.62810983622309e-05,
+      "loss": 0.9177,
+      "step": 43125
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.625677734928322e-05,
+      "loss": 0.932,
+      "step": 43130
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.703125,
+      "learning_rate": 7.623245782458777e-05,
+      "loss": 1.0339,
+      "step": 43135
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.620813978966895e-05,
+      "loss": 0.9261,
+      "step": 43140
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.618382324605104e-05,
+      "loss": 0.9913,
+      "step": 43145
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.615950819525821e-05,
+      "loss": 0.9563,
+      "step": 43150
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.613519463881456e-05,
+      "loss": 0.8773,
+      "step": 43155
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.611088257824405e-05,
+      "loss": 0.8426,
+      "step": 43160
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.608657201507066e-05,
+      "loss": 1.0284,
+      "step": 43165
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.606226295081815e-05,
+      "loss": 0.9378,
+      "step": 43170
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.603795538701026e-05,
+      "loss": 0.983,
+      "step": 43175
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.601364932517065e-05,
+      "loss": 0.9846,
+      "step": 43180
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.598934476682279e-05,
+      "loss": 0.8697,
+      "step": 43185
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.596504171349013e-05,
+      "loss": 1.0234,
+      "step": 43190
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.625,
+      "learning_rate": 7.594074016669606e-05,
+      "loss": 0.9989,
+      "step": 43195
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.59164401279638e-05,
+      "loss": 0.9878,
+      "step": 43200
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.589214159881652e-05,
+      "loss": 0.8622,
+      "step": 43205
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.586784458077723e-05,
+      "loss": 1.1644,
+      "step": 43210
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.584354907536892e-05,
+      "loss": 1.0065,
+      "step": 43215
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.581925508411448e-05,
+      "loss": 1.0772,
+      "step": 43220
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.579496260853666e-05,
+      "loss": 0.8406,
+      "step": 43225
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.577067165015815e-05,
+      "loss": 0.8243,
+      "step": 43230
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.574638221050156e-05,
+      "loss": 0.8884,
+      "step": 43235
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.572209429108935e-05,
+      "loss": 0.9756,
+      "step": 43240
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.569780789344388e-05,
+      "loss": 0.9726,
+      "step": 43245
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.567352301908752e-05,
+      "loss": 0.9453,
+      "step": 43250
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.564923966954247e-05,
+      "loss": 0.8655,
+      "step": 43255
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.562495784633078e-05,
+      "loss": 0.8646,
+      "step": 43260
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.56006775509745e-05,
+      "loss": 1.0515,
+      "step": 43265
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.557639878499551e-05,
+      "loss": 0.932,
+      "step": 43270
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.555212154991569e-05,
+      "loss": 0.9745,
+      "step": 43275
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.552784584725674e-05,
+      "loss": 1.104,
+      "step": 43280
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.550357167854025e-05,
+      "loss": 1.0603,
+      "step": 43285
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.547929904528783e-05,
+      "loss": 0.8669,
+      "step": 43290
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.545502794902084e-05,
+      "loss": 0.8064,
+      "step": 43295
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.543075839126065e-05,
+      "loss": 1.0797,
+      "step": 43300
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.540649037352853e-05,
+      "loss": 0.9345,
+      "step": 43305
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.538222389734561e-05,
+      "loss": 0.9949,
+      "step": 43310
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.535795896423292e-05,
+      "loss": 0.9763,
+      "step": 43315
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.533369557571144e-05,
+      "loss": 0.9526,
+      "step": 43320
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.486328125,
+      "learning_rate": 7.530943373330197e-05,
+      "loss": 0.9384,
+      "step": 43325
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.528517343852535e-05,
+      "loss": 1.0181,
+      "step": 43330
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.526091469290221e-05,
+      "loss": 0.8762,
+      "step": 43335
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.470703125,
+      "learning_rate": 7.523665749795308e-05,
+      "loss": 0.8089,
+      "step": 43340
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.521240185519849e-05,
+      "loss": 0.9557,
+      "step": 43345
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.518814776615878e-05,
+      "loss": 0.8498,
+      "step": 43350
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.498046875,
+      "learning_rate": 7.51638952323542e-05,
+      "loss": 1.0413,
+      "step": 43355
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.513964425530502e-05,
+      "loss": 0.8996,
+      "step": 43360
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.511539483653119e-05,
+      "loss": 1.0167,
+      "step": 43365
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.478515625,
+      "learning_rate": 7.509114697755277e-05,
+      "loss": 0.9066,
+      "step": 43370
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.506690067988963e-05,
+      "loss": 0.9654,
+      "step": 43375
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.504265594506152e-05,
+      "loss": 1.0211,
+      "step": 43380
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.50184127745882e-05,
+      "loss": 0.9962,
+      "step": 43385
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.499417116998921e-05,
+      "loss": 1.0111,
+      "step": 43390
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.496993113278403e-05,
+      "loss": 0.9227,
+      "step": 43395
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.49456926644921e-05,
+      "loss": 0.8432,
+      "step": 43400
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.492145576663267e-05,
+      "loss": 0.9982,
+      "step": 43405
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.489722044072493e-05,
+      "loss": 0.9431,
+      "step": 43410
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.487298668828809e-05,
+      "loss": 1.0562,
+      "step": 43415
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.484875451084098e-05,
+      "loss": 1.0369,
+      "step": 43420
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.482452390990262e-05,
+      "loss": 1.0191,
+      "step": 43425
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.480029488699177e-05,
+      "loss": 1.0395,
+      "step": 43430
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.47760674436271e-05,
+      "loss": 0.9695,
+      "step": 43435
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.475184158132728e-05,
+      "loss": 0.9514,
+      "step": 43440
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.472761730161079e-05,
+      "loss": 1.0904,
+      "step": 43445
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.470339460599601e-05,
+      "loss": 0.8637,
+      "step": 43450
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.467917349600129e-05,
+      "loss": 0.7954,
+      "step": 43455
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.46549539731448e-05,
+      "loss": 0.9869,
+      "step": 43460
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.463073603894469e-05,
+      "loss": 0.8432,
+      "step": 43465
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.460651969491892e-05,
+      "loss": 0.8657,
+      "step": 43470
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.45823049425854e-05,
+      "loss": 0.9686,
+      "step": 43475
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.455809178346196e-05,
+      "loss": 1.0432,
+      "step": 43480
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.453388021906631e-05,
+      "loss": 0.7601,
+      "step": 43485
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.450967025091604e-05,
+      "loss": 0.9924,
+      "step": 43490
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.448546188052867e-05,
+      "loss": 0.9843,
+      "step": 43495
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.458984375,
+      "learning_rate": 7.446125510942162e-05,
+      "loss": 0.9573,
+      "step": 43500
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.443704993911216e-05,
+      "loss": 0.9863,
+      "step": 43505
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.441284637111755e-05,
+      "loss": 0.9481,
+      "step": 43510
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.438864440695487e-05,
+      "loss": 0.9117,
+      "step": 43515
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.436444404814115e-05,
+      "loss": 0.9144,
+      "step": 43520
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.434024529619325e-05,
+      "loss": 0.9627,
+      "step": 43525
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.431604815262799e-05,
+      "loss": 1.0474,
+      "step": 43530
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.470703125,
+      "learning_rate": 7.429185261896207e-05,
+      "loss": 1.0196,
+      "step": 43535
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.426765869671214e-05,
+      "loss": 0.8849,
+      "step": 43540
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.424346638739463e-05,
+      "loss": 0.9278,
+      "step": 43545
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.421927569252601e-05,
+      "loss": 0.9472,
+      "step": 43550
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.494140625,
+      "learning_rate": 7.419508661362255e-05,
+      "loss": 0.9385,
+      "step": 43555
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.417089915220044e-05,
+      "loss": 0.8907,
+      "step": 43560
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.41467133097758e-05,
+      "loss": 1.055,
+      "step": 43565
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.412252908786463e-05,
+      "loss": 0.9292,
+      "step": 43570
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.409834648798279e-05,
+      "loss": 1.0102,
+      "step": 43575
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.407416551164608e-05,
+      "loss": 0.9795,
+      "step": 43580
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.478515625,
+      "learning_rate": 7.404998616037022e-05,
+      "loss": 0.8821,
+      "step": 43585
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.402580843567078e-05,
+      "loss": 0.8465,
+      "step": 43590
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.400163233906324e-05,
+      "loss": 1.0351,
+      "step": 43595
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.397745787206298e-05,
+      "loss": 0.9298,
+      "step": 43600
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.395328503618533e-05,
+      "loss": 0.982,
+      "step": 43605
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.392911383294543e-05,
+      "loss": 0.9462,
+      "step": 43610
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.390494426385835e-05,
+      "loss": 1.0016,
+      "step": 43615
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.388077633043908e-05,
+      "loss": 0.8682,
+      "step": 43620
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.385661003420255e-05,
+      "loss": 0.9146,
+      "step": 43625
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.383244537666345e-05,
+      "loss": 0.896,
+      "step": 43630
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.380828235933644e-05,
+      "loss": 1.0661,
+      "step": 43635
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.378412098373616e-05,
+      "loss": 0.909,
+      "step": 43640
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.375996125137704e-05,
+      "loss": 1.0145,
+      "step": 43645
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.373580316377337e-05,
+      "loss": 0.8261,
+      "step": 43650
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.371164672243953e-05,
+      "loss": 0.9722,
+      "step": 43655
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.36874919288896e-05,
+      "loss": 0.856,
+      "step": 43660
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.366333878463765e-05,
+      "loss": 1.0439,
+      "step": 43665
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.36391872911976e-05,
+      "loss": 1.074,
+      "step": 43670
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.361503745008335e-05,
+      "loss": 0.8275,
+      "step": 43675
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.359088926280858e-05,
+      "loss": 0.9063,
+      "step": 43680
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.356674273088695e-05,
+      "loss": 0.9664,
+      "step": 43685
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.354259785583197e-05,
+      "loss": 0.9558,
+      "step": 43690
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.351845463915711e-05,
+      "loss": 0.9554,
+      "step": 43695
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.349431308237568e-05,
+      "loss": 1.1306,
+      "step": 43700
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.347017318700087e-05,
+      "loss": 0.8899,
+      "step": 43705
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.344603495454582e-05,
+      "loss": 0.9215,
+      "step": 43710
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.342189838652357e-05,
+      "loss": 0.9706,
+      "step": 43715
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.339776348444696e-05,
+      "loss": 0.8483,
+      "step": 43720
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.337363024982886e-05,
+      "loss": 0.8448,
+      "step": 43725
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.334949868418197e-05,
+      "loss": 0.8894,
+      "step": 43730
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.332536878901882e-05,
+      "loss": 0.9876,
+      "step": 43735
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.330124056585194e-05,
+      "loss": 0.983,
+      "step": 43740
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.327711401619369e-05,
+      "loss": 1.0151,
+      "step": 43745
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.325298914155638e-05,
+      "loss": 1.0221,
+      "step": 43750
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.322886594345218e-05,
+      "loss": 0.854,
+      "step": 43755
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.320474442339314e-05,
+      "loss": 0.8749,
+      "step": 43760
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.490234375,
+      "learning_rate": 7.318062458289124e-05,
+      "loss": 0.8302,
+      "step": 43765
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.315650642345835e-05,
+      "loss": 1.0153,
+      "step": 43770
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.313238994660618e-05,
+      "loss": 1.0044,
+      "step": 43775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.310827515384648e-05,
+      "loss": 0.9339,
+      "step": 43780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.308416204669063e-05,
+      "loss": 0.8673,
+      "step": 43785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.30600506266502e-05,
+      "loss": 0.9973,
+      "step": 43790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.303594089523649e-05,
+      "loss": 0.818,
+      "step": 43795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.301183285396068e-05,
+      "loss": 0.8573,
+      "step": 43800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.298772650433394e-05,
+      "loss": 0.9827,
+      "step": 43805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.296362184786728e-05,
+      "loss": 0.9497,
+      "step": 43810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.293951888607156e-05,
+      "loss": 1.0426,
+      "step": 43815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.49609375,
+      "learning_rate": 7.291541762045765e-05,
+      "loss": 0.9381,
+      "step": 43820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.28913180525362e-05,
+      "loss": 0.8794,
+      "step": 43825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.28672201838178e-05,
+      "loss": 0.9819,
+      "step": 43830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.2843124015813e-05,
+      "loss": 1.1377,
+      "step": 43835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.281902955003204e-05,
+      "loss": 1.0289,
+      "step": 43840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.279493678798529e-05,
+      "loss": 0.9459,
+      "step": 43845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.277084573118289e-05,
+      "loss": 0.9653,
+      "step": 43850
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.274675638113486e-05,
+      "loss": 0.8437,
+      "step": 43855
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.272266873935122e-05,
+      "loss": 0.974,
+      "step": 43860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.269858280734176e-05,
+      "loss": 1.1957,
+      "step": 43865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.26744985866162e-05,
+      "loss": 0.9526,
+      "step": 43870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.265041607868422e-05,
+      "loss": 0.8111,
+      "step": 43875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.262633528505529e-05,
+      "loss": 0.8836,
+      "step": 43880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.260225620723888e-05,
+      "loss": 0.8641,
+      "step": 43885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.257817884674421e-05,
+      "loss": 0.82,
+      "step": 43890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.255410320508052e-05,
+      "loss": 1.0119,
+      "step": 43895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.253002928375692e-05,
+      "loss": 0.8987,
+      "step": 43900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.250595708428236e-05,
+      "loss": 1.0118,
+      "step": 43905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.248188660816571e-05,
+      "loss": 1.0069,
+      "step": 43910
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.245781785691576e-05,
+      "loss": 0.9783,
+      "step": 43915
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.243375083204116e-05,
+      "loss": 0.9433,
+      "step": 43920
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.240968553505043e-05,
+      "loss": 0.9561,
+      "step": 43925
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.238562196745206e-05,
+      "loss": 0.9336,
+      "step": 43930
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.236156013075435e-05,
+      "loss": 1.0196,
+      "step": 43935
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.233750002646555e-05,
+      "loss": 0.9104,
+      "step": 43940
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.231344165609375e-05,
+      "loss": 0.9293,
+      "step": 43945
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.228938502114692e-05,
+      "loss": 0.9552,
+      "step": 43950
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.226533012313301e-05,
+      "loss": 0.8964,
+      "step": 43955
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.224127696355981e-05,
+      "loss": 0.889,
+      "step": 43960
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.221722554393496e-05,
+      "loss": 1.1623,
+      "step": 43965
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.219317586576609e-05,
+      "loss": 0.8779,
+      "step": 43970
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.21691279305606e-05,
+      "loss": 0.8894,
+      "step": 43975
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.214508173982585e-05,
+      "loss": 1.0329,
+      "step": 43980
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.212103729506914e-05,
+      "loss": 0.8058,
+      "step": 43985
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.209699459779758e-05,
+      "loss": 0.8913,
+      "step": 43990
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.625,
+      "learning_rate": 7.207295364951814e-05,
+      "loss": 0.9224,
+      "step": 43995
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.204891445173779e-05,
+      "loss": 1.0408,
+      "step": 44000
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.202487700596328e-05,
+      "loss": 1.0718,
+      "step": 44005
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.200084131370138e-05,
+      "loss": 0.91,
+      "step": 44010
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.197680737645861e-05,
+      "loss": 0.8467,
+      "step": 44015
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.195277519574147e-05,
+      "loss": 0.9021,
+      "step": 44020
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.192874477305633e-05,
+      "loss": 0.8741,
+      "step": 44025
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.190471610990944e-05,
+      "loss": 1.0427,
+      "step": 44030
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.188068920780692e-05,
+      "loss": 0.9584,
+      "step": 44035
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.185666406825486e-05,
+      "loss": 0.9775,
+      "step": 44040
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.183264069275915e-05,
+      "loss": 0.9982,
+      "step": 44045
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.18086190828256e-05,
+      "loss": 0.9059,
+      "step": 44050
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.17845992399599e-05,
+      "loss": 0.9274,
+      "step": 44055
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.176058116566764e-05,
+      "loss": 1.0291,
+      "step": 44060
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.173656486145434e-05,
+      "loss": 0.8917,
+      "step": 44065
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.171255032882534e-05,
+      "loss": 0.9251,
+      "step": 44070
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.168853756928587e-05,
+      "loss": 1.0542,
+      "step": 44075
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.166452658434115e-05,
+      "loss": 1.0037,
+      "step": 44080
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7890625,
+      "learning_rate": 7.164051737549615e-05,
+      "loss": 0.9898,
+      "step": 44085
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.161650994425582e-05,
+      "loss": 0.8652,
+      "step": 44090
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.159250429212503e-05,
+      "loss": 1.0474,
+      "step": 44095
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.156850042060837e-05,
+      "loss": 1.0262,
+      "step": 44100
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.154449833121049e-05,
+      "loss": 0.8672,
+      "step": 44105
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.152049802543587e-05,
+      "loss": 0.8591,
+      "step": 44110
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.498046875,
+      "learning_rate": 7.149649950478884e-05,
+      "loss": 0.9944,
+      "step": 44115
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.147250277077371e-05,
+      "loss": 0.9759,
+      "step": 44120
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.14485078248946e-05,
+      "loss": 0.9551,
+      "step": 44125
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.142451466865551e-05,
+      "loss": 0.9274,
+      "step": 44130
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.140052330356042e-05,
+      "loss": 0.9259,
+      "step": 44135
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.137653373111309e-05,
+      "loss": 1.0829,
+      "step": 44140
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.135254595281719e-05,
+      "loss": 0.982,
+      "step": 44145
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.132855997017642e-05,
+      "loss": 0.8594,
+      "step": 44150
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.13045757846941e-05,
+      "loss": 1.0339,
+      "step": 44155
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.128059339787368e-05,
+      "loss": 0.974,
+      "step": 44160
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.125661281121837e-05,
+      "loss": 0.8284,
+      "step": 44165
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.123263402623125e-05,
+      "loss": 0.9314,
+      "step": 44170
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.120865704441546e-05,
+      "loss": 0.9241,
+      "step": 44175
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.11846818672738e-05,
+      "loss": 0.9662,
+      "step": 44180
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.116070849630911e-05,
+      "loss": 1.0086,
+      "step": 44185
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.113673693302406e-05,
+      "loss": 0.9713,
+      "step": 44190
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.111276717892121e-05,
+      "loss": 0.9079,
+      "step": 44195
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.108879923550305e-05,
+      "loss": 1.0538,
+      "step": 44200
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.106483310427184e-05,
+      "loss": 0.9012,
+      "step": 44205
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.48828125,
+      "learning_rate": 7.104086878672984e-05,
+      "loss": 0.8179,
+      "step": 44210
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.101690628437918e-05,
+      "loss": 0.8541,
+      "step": 44215
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.099294559872184e-05,
+      "loss": 1.0969,
+      "step": 44220
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.096898673125969e-05,
+      "loss": 1.0144,
+      "step": 44225
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.094502968349453e-05,
+      "loss": 0.9345,
+      "step": 44230
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.092107445692802e-05,
+      "loss": 1.0356,
+      "step": 44235
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.089712105306163e-05,
+      "loss": 1.0366,
+      "step": 44240
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.087316947339689e-05,
+      "loss": 0.9069,
+      "step": 44245
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.084921971943503e-05,
+      "loss": 0.9973,
+      "step": 44250
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.082527179267731e-05,
+      "loss": 0.9333,
+      "step": 44255
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.080132569462474e-05,
+      "loss": 1.0377,
+      "step": 44260
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.077738142677836e-05,
+      "loss": 0.8374,
+      "step": 44265
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.0753438990639e-05,
+      "loss": 0.9651,
+      "step": 44270
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.072949838770737e-05,
+      "loss": 1.0695,
+      "step": 44275
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.07055596194841e-05,
+      "loss": 0.9352,
+      "step": 44280
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.068162268746975e-05,
+      "loss": 0.8394,
+      "step": 44285
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.065768759316468e-05,
+      "loss": 1.0907,
+      "step": 44290
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5,
+      "learning_rate": 7.063375433806914e-05,
+      "loss": 0.9587,
+      "step": 44295
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.060982292368334e-05,
+      "loss": 1.0745,
+      "step": 44300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.058589335150734e-05,
+      "loss": 0.9421,
+      "step": 44305
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.056196562304103e-05,
+      "loss": 0.9171,
+      "step": 44310
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.053803973978423e-05,
+      "loss": 0.9543,
+      "step": 44315
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.051411570323665e-05,
+      "loss": 0.8534,
+      "step": 44320
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.04901935148979e-05,
+      "loss": 0.7847,
+      "step": 44325
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.625,
+      "learning_rate": 7.04662731762674e-05,
+      "loss": 0.9292,
+      "step": 44330
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.044235468884455e-05,
+      "loss": 0.9307,
+      "step": 44335
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.041843805412855e-05,
+      "loss": 0.9345,
+      "step": 44340
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.039452327361857e-05,
+      "loss": 0.8205,
+      "step": 44345
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.037061034881358e-05,
+      "loss": 0.9699,
+      "step": 44350
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.034669928121248e-05,
+      "loss": 0.9605,
+      "step": 44355
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.625,
+      "learning_rate": 7.032279007231406e-05,
+      "loss": 1.0062,
+      "step": 44360
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.029888272361695e-05,
+      "loss": 0.8797,
+      "step": 44365
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.027497723661967e-05,
+      "loss": 0.9759,
+      "step": 44370
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.025107361282069e-05,
+      "loss": 0.9113,
+      "step": 44375
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.02271718537183e-05,
+      "loss": 1.0109,
+      "step": 44380
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.020327196081067e-05,
+      "loss": 0.8384,
+      "step": 44385
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.01793739355959e-05,
+      "loss": 0.9772,
+      "step": 44390
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.015547777957194e-05,
+      "loss": 1.0573,
+      "step": 44395
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.01315834942366e-05,
+      "loss": 1.0351,
+      "step": 44400
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.010769108108764e-05,
+      "loss": 0.8791,
+      "step": 44405
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.008380054162268e-05,
+      "loss": 0.9083,
+      "step": 44410
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.005991187733914e-05,
+      "loss": 0.9679,
+      "step": 44415
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.003602508973444e-05,
+      "loss": 0.8742,
+      "step": 44420
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.001214018030578e-05,
+      "loss": 0.8526,
+      "step": 44425
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.998825715055035e-05,
+      "loss": 0.8964,
+      "step": 44430
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.996437600196514e-05,
+      "loss": 1.1129,
+      "step": 44435
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.994049673604703e-05,
+      "loss": 0.8891,
+      "step": 44440
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.991661935429284e-05,
+      "loss": 0.8465,
+      "step": 44445
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.989274385819921e-05,
+      "loss": 1.0433,
+      "step": 44450
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.986887024926267e-05,
+      "loss": 0.8986,
+      "step": 44455
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.984499852897968e-05,
+      "loss": 1.0319,
+      "step": 44460
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.982112869884654e-05,
+      "loss": 0.9768,
+      "step": 44465
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.734375,
+      "learning_rate": 6.97972607603594e-05,
+      "loss": 0.8718,
+      "step": 44470
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.977339471501436e-05,
+      "loss": 0.969,
+      "step": 44475
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.974953056430736e-05,
+      "loss": 0.9752,
+      "step": 44480
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.972566830973423e-05,
+      "loss": 0.8427,
+      "step": 44485
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.970180795279069e-05,
+      "loss": 0.909,
+      "step": 44490
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.967794949497233e-05,
+      "loss": 0.9467,
+      "step": 44495
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.965409293777464e-05,
+      "loss": 1.0362,
+      "step": 44500
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.963023828269297e-05,
+      "loss": 0.8744,
+      "step": 44505
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.960638553122254e-05,
+      "loss": 0.9793,
+      "step": 44510
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4765625,
+      "learning_rate": 6.958253468485853e-05,
+      "loss": 0.9228,
+      "step": 44515
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.955868574509583e-05,
+      "loss": 0.9583,
+      "step": 44520
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.953483871342941e-05,
+      "loss": 1.0902,
+      "step": 44525
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.951099359135399e-05,
+      "loss": 1.0439,
+      "step": 44530
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.948715038036417e-05,
+      "loss": 1.0237,
+      "step": 44535
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.946330908195457e-05,
+      "loss": 1.0154,
+      "step": 44540
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.943946969761951e-05,
+      "loss": 0.9348,
+      "step": 44545
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.94156322288533e-05,
+      "loss": 0.9672,
+      "step": 44550
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.93917966771501e-05,
+      "loss": 0.9573,
+      "step": 44555
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.936796304400395e-05,
+      "loss": 1.1258,
+      "step": 44560
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.46875,
+      "learning_rate": 6.934413133090872e-05,
+      "loss": 0.9212,
+      "step": 44565
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.932030153935834e-05,
+      "loss": 0.9884,
+      "step": 44570
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.929647367084633e-05,
+      "loss": 0.9147,
+      "step": 44575
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.927264772686635e-05,
+      "loss": 1.0419,
+      "step": 44580
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.924882370891179e-05,
+      "loss": 0.9673,
+      "step": 44585
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.922500161847596e-05,
+      "loss": 0.9509,
+      "step": 44590
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.92011814570521e-05,
+      "loss": 1.022,
+      "step": 44595
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.917736322613329e-05,
+      "loss": 1.0838,
+      "step": 44600
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.915354692721242e-05,
+      "loss": 1.0785,
+      "step": 44605
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.912973256178236e-05,
+      "loss": 1.054,
+      "step": 44610
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.910592013133584e-05,
+      "loss": 0.811,
+      "step": 44615
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.908210963736546e-05,
+      "loss": 0.9427,
+      "step": 44620
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.905830108136362e-05,
+      "loss": 0.8712,
+      "step": 44625
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.903449446482271e-05,
+      "loss": 0.8157,
+      "step": 44630
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.901068978923495e-05,
+      "loss": 1.0176,
+      "step": 44635
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.898688705609246e-05,
+      "loss": 0.875,
+      "step": 44640
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.896308626688719e-05,
+      "loss": 1.0577,
+      "step": 44645
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.893928742311104e-05,
+      "loss": 0.9952,
+      "step": 44650
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.494140625,
+      "learning_rate": 6.891549052625574e-05,
+      "loss": 0.9785,
+      "step": 44655
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.889169557781285e-05,
+      "loss": 0.9298,
+      "step": 44660
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.886790257927395e-05,
+      "loss": 0.9041,
+      "step": 44665
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.884411153213037e-05,
+      "loss": 1.0325,
+      "step": 44670
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.88203224378734e-05,
+      "loss": 0.9072,
+      "step": 44675
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.879653529799408e-05,
+      "loss": 1.0979,
+      "step": 44680
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.877275011398346e-05,
+      "loss": 0.9291,
+      "step": 44685
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.874896688733246e-05,
+      "loss": 0.953,
+      "step": 44690
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.872518561953178e-05,
+      "loss": 0.9055,
+      "step": 44695
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.72265625,
+      "learning_rate": 6.870140631207207e-05,
+      "loss": 1.0469,
+      "step": 44700
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.86776289664439e-05,
+      "loss": 0.885,
+      "step": 44705
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.865385358413761e-05,
+      "loss": 0.9648,
+      "step": 44710
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.863008016664344e-05,
+      "loss": 0.9232,
+      "step": 44715
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.86063087154516e-05,
+      "loss": 0.974,
+      "step": 44720
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.85825392320521e-05,
+      "loss": 1.0025,
+      "step": 44725
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.855877171793484e-05,
+      "loss": 0.9794,
+      "step": 44730
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.853500617458955e-05,
+      "loss": 1.0223,
+      "step": 44735
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.851124260350588e-05,
+      "loss": 1.0199,
+      "step": 44740
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.848748100617342e-05,
+      "loss": 0.9964,
+      "step": 44745
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.494140625,
+      "learning_rate": 6.846372138408152e-05,
+      "loss": 0.8822,
+      "step": 44750
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.843996373871948e-05,
+      "loss": 0.8908,
+      "step": 44755
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.841620807157647e-05,
+      "loss": 0.9091,
+      "step": 44760
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.839245438414152e-05,
+      "loss": 0.9188,
+      "step": 44765
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.83687026779035e-05,
+      "loss": 1.1365,
+      "step": 44770
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.834495295435123e-05,
+      "loss": 1.0363,
+      "step": 44775
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.832120521497339e-05,
+      "loss": 0.9475,
+      "step": 44780
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.829745946125847e-05,
+      "loss": 0.9322,
+      "step": 44785
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.827371569469489e-05,
+      "loss": 1.1476,
+      "step": 44790
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.824997391677092e-05,
+      "loss": 0.9128,
+      "step": 44795
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.822623412897479e-05,
+      "loss": 0.9565,
+      "step": 44800
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.820249633279448e-05,
+      "loss": 0.9199,
+      "step": 44805
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.817876052971788e-05,
+      "loss": 1.0906,
+      "step": 44810
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.48046875,
+      "learning_rate": 6.815502672123284e-05,
+      "loss": 0.8355,
+      "step": 44815
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.813129490882699e-05,
+      "loss": 0.9748,
+      "step": 44820
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.810756509398786e-05,
+      "loss": 0.9507,
+      "step": 44825
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.808383727820292e-05,
+      "loss": 0.9011,
+      "step": 44830
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.806011146295937e-05,
+      "loss": 1.0287,
+      "step": 44835
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.80363876497444e-05,
+      "loss": 1.0217,
+      "step": 44840
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.801266584004507e-05,
+      "loss": 1.0007,
+      "step": 44845
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.798894603534827e-05,
+      "loss": 0.9481,
+      "step": 44850
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.796522823714079e-05,
+      "loss": 1.0604,
+      "step": 44855
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.79415124469093e-05,
+      "loss": 0.9995,
+      "step": 44860
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.791779866614028e-05,
+      "loss": 0.951,
+      "step": 44865
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.789408689632021e-05,
+      "loss": 0.898,
+      "step": 44870
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.787037713893536e-05,
+      "loss": 1.0025,
+      "step": 44875
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.784666939547182e-05,
+      "loss": 0.974,
+      "step": 44880
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.782296366741574e-05,
+      "loss": 0.9054,
+      "step": 44885
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.779925995625287e-05,
+      "loss": 1.0215,
+      "step": 44890
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.490234375,
+      "learning_rate": 6.777555826346907e-05,
+      "loss": 0.9463,
+      "step": 44895
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.775185859055e-05,
+      "loss": 1.0455,
+      "step": 44900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.772816093898114e-05,
+      "loss": 0.8271,
+      "step": 44905
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.490234375,
+      "learning_rate": 6.77044653102479e-05,
+      "loss": 0.9981,
+      "step": 44910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.768077170583558e-05,
+      "loss": 0.9118,
+      "step": 44915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.765708012722927e-05,
+      "loss": 1.0437,
+      "step": 44920
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.763339057591404e-05,
+      "loss": 0.8474,
+      "step": 44925
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.46875,
+      "learning_rate": 6.760970305337475e-05,
+      "loss": 0.8097,
+      "step": 44930
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.758601756109617e-05,
+      "loss": 0.8699,
+      "step": 44935
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.756233410056292e-05,
+      "loss": 0.9699,
+      "step": 44940
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.753865267325949e-05,
+      "loss": 0.9083,
+      "step": 44945
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.75149732806703e-05,
+      "loss": 0.8572,
+      "step": 44950
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.749129592427958e-05,
+      "loss": 1.0053,
+      "step": 44955
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.453125,
+      "learning_rate": 6.746762060557143e-05,
+      "loss": 0.7956,
+      "step": 44960
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.74439473260299e-05,
+      "loss": 1.0633,
+      "step": 44965
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.742027608713883e-05,
+      "loss": 0.8319,
+      "step": 44970
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.739660689038193e-05,
+      "loss": 0.9019,
+      "step": 44975
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.737293973724287e-05,
+      "loss": 0.9892,
+      "step": 44980
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.73492746292051e-05,
+      "loss": 0.9203,
+      "step": 44985
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.732561156775202e-05,
+      "loss": 1.0347,
+      "step": 44990
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.730195055436677e-05,
+      "loss": 0.9512,
+      "step": 44995
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.727829159053251e-05,
+      "loss": 1.0785,
+      "step": 45000
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.725463467773221e-05,
+      "loss": 0.9478,
+      "step": 45005
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.72309798174487e-05,
+      "loss": 0.943,
+      "step": 45010
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.720732701116468e-05,
+      "loss": 0.9247,
+      "step": 45015
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.718367626036276e-05,
+      "loss": 0.8705,
+      "step": 45020
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.478515625,
+      "learning_rate": 6.71600275665254e-05,
+      "loss": 0.9431,
+      "step": 45025
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.713638093113488e-05,
+      "loss": 0.932,
+      "step": 45030
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.711273635567346e-05,
+      "loss": 1.0349,
+      "step": 45035
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.70890938416232e-05,
+      "loss": 0.988,
+      "step": 45040
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.7065453390466e-05,
+      "loss": 0.9581,
+      "step": 45045
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.704181500368368e-05,
+      "loss": 0.93,
+      "step": 45050
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.701817868275792e-05,
+      "loss": 1.0163,
+      "step": 45055
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.699454442917031e-05,
+      "loss": 0.9731,
+      "step": 45060
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.697091224440221e-05,
+      "loss": 0.8179,
+      "step": 45065
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.6947282129935e-05,
+      "loss": 0.9169,
+      "step": 45070
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.692365408724976e-05,
+      "loss": 0.888,
+      "step": 45075
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.690002811782754e-05,
+      "loss": 0.9491,
+      "step": 45080
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.687640422314927e-05,
+      "loss": 0.8988,
+      "step": 45085
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.685278240469572e-05,
+      "loss": 1.0214,
+      "step": 45090
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.682916266394753e-05,
+      "loss": 1.192,
+      "step": 45095
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.680554500238519e-05,
+      "loss": 0.9118,
+      "step": 45100
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.678192942148907e-05,
+      "loss": 1.0664,
+      "step": 45105
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.675831592273947e-05,
+      "loss": 0.8504,
+      "step": 45110
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.673470450761647e-05,
+      "loss": 0.974,
+      "step": 45115
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.671109517760009e-05,
+      "loss": 0.9236,
+      "step": 45120
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.668748793417017e-05,
+      "loss": 0.9077,
+      "step": 45125
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.666388277880646e-05,
+      "loss": 0.9977,
+      "step": 45130
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.664027971298852e-05,
+      "loss": 1.0262,
+      "step": 45135
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.661667873819586e-05,
+      "loss": 1.0116,
+      "step": 45140
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.659307985590779e-05,
+      "loss": 0.8108,
+      "step": 45145
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.656948306760356e-05,
+      "loss": 0.9178,
+      "step": 45150
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.654588837476216e-05,
+      "loss": 0.8277,
+      "step": 45155
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.50390625,
+      "learning_rate": 6.652229577886258e-05,
+      "loss": 1.0075,
+      "step": 45160
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.649870528138364e-05,
+      "loss": 1.0151,
+      "step": 45165
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.647511688380402e-05,
+      "loss": 0.8129,
+      "step": 45170
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.64515305876022e-05,
+      "loss": 1.0257,
+      "step": 45175
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.642794639425671e-05,
+      "loss": 0.8383,
+      "step": 45180
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.640436430524576e-05,
+      "loss": 0.9116,
+      "step": 45185
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.638078432204749e-05,
+      "loss": 0.8931,
+      "step": 45190
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.635720644613998e-05,
+      "loss": 0.8874,
+      "step": 45195
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.63336306790011e-05,
+      "loss": 0.8541,
+      "step": 45200
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.50390625,
+      "learning_rate": 6.631005702210857e-05,
+      "loss": 0.9455,
+      "step": 45205
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.628648547694006e-05,
+      "loss": 0.8394,
+      "step": 45210
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.626291604497299e-05,
+      "loss": 1.0215,
+      "step": 45215
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.623934872768478e-05,
+      "loss": 0.9224,
+      "step": 45220
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.621578352655267e-05,
+      "loss": 0.9382,
+      "step": 45225
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.619222044305368e-05,
+      "loss": 1.0595,
+      "step": 45230
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.616865947866484e-05,
+      "loss": 0.8566,
+      "step": 45235
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.614510063486296e-05,
+      "loss": 0.8293,
+      "step": 45240
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.61215439131247e-05,
+      "loss": 0.8726,
+      "step": 45245
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.609798931492671e-05,
+      "loss": 1.0158,
+      "step": 45250
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.607443684174533e-05,
+      "loss": 0.9727,
+      "step": 45255
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.605088649505689e-05,
+      "loss": 0.9833,
+      "step": 45260
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.602733827633756e-05,
+      "loss": 0.9667,
+      "step": 45265
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.600379218706331e-05,
+      "loss": 1.0605,
+      "step": 45270
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.625,
+      "learning_rate": 6.598024822871014e-05,
+      "loss": 0.8976,
+      "step": 45275
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.595670640275373e-05,
+      "loss": 0.9378,
+      "step": 45280
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.593316671066972e-05,
+      "loss": 0.9807,
+      "step": 45285
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.470703125,
+      "learning_rate": 6.590962915393364e-05,
+      "loss": 0.9207,
+      "step": 45290
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.625,
+      "learning_rate": 6.588609373402084e-05,
+      "loss": 1.0662,
+      "step": 45295
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.58625604524065e-05,
+      "loss": 0.9258,
+      "step": 45300
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.583902931056582e-05,
+      "loss": 0.9294,
+      "step": 45305
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.581550030997363e-05,
+      "loss": 0.9434,
+      "step": 45310
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.579197345210483e-05,
+      "loss": 1.028,
+      "step": 45315
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.576844873843409e-05,
+      "loss": 0.8776,
+      "step": 45320
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.574492617043596e-05,
+      "loss": 1.091,
+      "step": 45325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.572140574958488e-05,
+      "loss": 0.8153,
+      "step": 45330
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.569788747735515e-05,
+      "loss": 1.0016,
+      "step": 45335
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.567437135522085e-05,
+      "loss": 0.9558,
+      "step": 45340
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.565085738465608e-05,
+      "loss": 1.0025,
+      "step": 45345
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.56273455671347e-05,
+      "loss": 0.8308,
+      "step": 45350
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.560383590413042e-05,
+      "loss": 1.0665,
+      "step": 45355
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.558032839711693e-05,
+      "loss": 0.9764,
+      "step": 45360
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.555682304756761e-05,
+      "loss": 0.8749,
+      "step": 45365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.553331985695586e-05,
+      "loss": 1.0905,
+      "step": 45370
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.550981882675487e-05,
+      "loss": 1.0649,
+      "step": 45375
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.54863199584377e-05,
+      "loss": 0.9374,
+      "step": 45380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.546282325347733e-05,
+      "loss": 1.1363,
+      "step": 45385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.543932871334652e-05,
+      "loss": 0.8631,
+      "step": 45390
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.541583633951795e-05,
+      "loss": 0.7846,
+      "step": 45395
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.539234613346415e-05,
+      "loss": 0.9328,
+      "step": 45400
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.536885809665752e-05,
+      "loss": 1.0687,
+      "step": 45405
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.53453722305703e-05,
+      "loss": 0.7834,
+      "step": 45410
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.532188853667462e-05,
+      "loss": 0.9781,
+      "step": 45415
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4921875,
+      "learning_rate": 6.529840701644245e-05,
+      "loss": 0.9171,
+      "step": 45420
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.527492767134566e-05,
+      "loss": 0.9773,
+      "step": 45425
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.48046875,
+      "learning_rate": 6.525145050285594e-05,
+      "loss": 0.8882,
+      "step": 45430
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.522797551244487e-05,
+      "loss": 0.944,
+      "step": 45435
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.520450270158391e-05,
+      "loss": 0.9338,
+      "step": 45440
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.518103207174436e-05,
+      "loss": 0.9582,
+      "step": 45445
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.515756362439736e-05,
+      "loss": 0.9155,
+      "step": 45450
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.513409736101396e-05,
+      "loss": 0.9894,
+      "step": 45455
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.511063328306504e-05,
+      "loss": 1.0464,
+      "step": 45460
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.508717139202139e-05,
+      "loss": 0.9744,
+      "step": 45465
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.506371168935359e-05,
+      "loss": 0.996,
+      "step": 45470
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.50402541765321e-05,
+      "loss": 0.8536,
+      "step": 45475
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.501679885502731e-05,
+      "loss": 1.0797,
+      "step": 45480
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.499334572630942e-05,
+      "loss": 0.9744,
+      "step": 45485
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.496989479184847e-05,
+      "loss": 0.921,
+      "step": 45490
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.494644605311442e-05,
+      "loss": 1.087,
+      "step": 45495
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.492299951157706e-05,
+      "loss": 1.0464,
+      "step": 45500
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.489955516870601e-05,
+      "loss": 0.8365,
+      "step": 45505
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.487611302597085e-05,
+      "loss": 0.9386,
+      "step": 45510
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.485267308484095e-05,
+      "loss": 0.9561,
+      "step": 45515
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.482923534678552e-05,
+      "loss": 0.8684,
+      "step": 45520
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.480579981327365e-05,
+      "loss": 0.9156,
+      "step": 45525
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.478236648577431e-05,
+      "loss": 0.9276,
+      "step": 45530
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.625,
+      "learning_rate": 6.475893536575639e-05,
+      "loss": 0.8871,
+      "step": 45535
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.473550645468853e-05,
+      "loss": 0.9564,
+      "step": 45540
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.471207975403926e-05,
+      "loss": 0.7958,
+      "step": 45545
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.486328125,
+      "learning_rate": 6.468865526527704e-05,
+      "loss": 0.8408,
+      "step": 45550
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.466523298987013e-05,
+      "loss": 0.8427,
+      "step": 45555
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.464181292928664e-05,
+      "loss": 1.0888,
+      "step": 45560
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.461839508499461e-05,
+      "loss": 0.9834,
+      "step": 45565
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.459497945846189e-05,
+      "loss": 1.12,
+      "step": 45570
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.457156605115615e-05,
+      "loss": 1.0514,
+      "step": 45575
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.454815486454501e-05,
+      "loss": 0.982,
+      "step": 45580
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.452474590009587e-05,
+      "loss": 0.9026,
+      "step": 45585
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.450133915927609e-05,
+      "loss": 0.9451,
+      "step": 45590
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.490234375,
+      "learning_rate": 6.447793464355279e-05,
+      "loss": 0.986,
+      "step": 45595
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.445453235439299e-05,
+      "loss": 0.9617,
+      "step": 45600
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.44311322932636e-05,
+      "loss": 0.7838,
+      "step": 45605
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.440773446163135e-05,
+      "loss": 1.0951,
+      "step": 45610
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.438433886096283e-05,
+      "loss": 0.9337,
+      "step": 45615
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.436094549272457e-05,
+      "loss": 1.0079,
+      "step": 45620
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.433755435838277e-05,
+      "loss": 0.937,
+      "step": 45625
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.43141654594037e-05,
+      "loss": 0.85,
+      "step": 45630
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.429077879725338e-05,
+      "loss": 1.2354,
+      "step": 45635
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.42673943733977e-05,
+      "loss": 1.0328,
+      "step": 45640
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4765625,
+      "learning_rate": 6.424401218930245e-05,
+      "loss": 0.9628,
+      "step": 45645
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.625,
+      "learning_rate": 6.422063224643325e-05,
+      "loss": 1.0466,
+      "step": 45650
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.419725454625554e-05,
+      "loss": 1.0123,
+      "step": 45655
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.417387909023471e-05,
+      "loss": 0.8303,
+      "step": 45660
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4765625,
+      "learning_rate": 6.415050587983593e-05,
+      "loss": 0.9342,
+      "step": 45665
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.412713491652427e-05,
+      "loss": 0.8515,
+      "step": 45670
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.410376620176468e-05,
+      "loss": 0.9153,
+      "step": 45675
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.40803997370219e-05,
+      "loss": 0.895,
+      "step": 45680
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.405703552376057e-05,
+      "loss": 0.9686,
+      "step": 45685
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.486328125,
+      "learning_rate": 6.403367356344517e-05,
+      "loss": 0.9334,
+      "step": 45690
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.401031385754006e-05,
+      "loss": 0.9136,
+      "step": 45695
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.39869564075095e-05,
+      "loss": 1.0172,
+      "step": 45700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.396360121481752e-05,
+      "loss": 1.1214,
+      "step": 45705
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.394024828092804e-05,
+      "loss": 0.891,
+      "step": 45710
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.486328125,
+      "learning_rate": 6.391689760730488e-05,
+      "loss": 0.8239,
+      "step": 45715
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.389354919541169e-05,
+      "loss": 0.9533,
+      "step": 45720
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.387020304671197e-05,
+      "loss": 0.9281,
+      "step": 45725
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.384685916266901e-05,
+      "loss": 0.9008,
+      "step": 45730
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.382351754474614e-05,
+      "loss": 0.8932,
+      "step": 45735
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.380017819440638e-05,
+      "loss": 0.8201,
+      "step": 45740
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.377684111311267e-05,
+      "loss": 0.9648,
+      "step": 45745
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.375350630232782e-05,
+      "loss": 0.916,
+      "step": 45750
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.482421875,
+      "learning_rate": 6.373017376351447e-05,
+      "loss": 0.9437,
+      "step": 45755
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.370684349813515e-05,
+      "loss": 0.894,
+      "step": 45760
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.36835155076522e-05,
+      "loss": 1.1337,
+      "step": 45765
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.366018979352786e-05,
+      "loss": 1.0193,
+      "step": 45770
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.363686635722421e-05,
+      "loss": 1.1689,
+      "step": 45775
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.361354520020324e-05,
+      "loss": 1.0801,
+      "step": 45780
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.359022632392663e-05,
+      "loss": 0.8107,
+      "step": 45785
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.356690972985612e-05,
+      "loss": 0.9673,
+      "step": 45790
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.354359541945323e-05,
+      "loss": 0.963,
+      "step": 45795
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.352028339417926e-05,
+      "loss": 0.7841,
+      "step": 45800
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.349697365549549e-05,
+      "loss": 1.0819,
+      "step": 45805
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.3473666204863e-05,
+      "loss": 1.0213,
+      "step": 45810
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.345036104374267e-05,
+      "loss": 0.9277,
+      "step": 45815
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.342705817359538e-05,
+      "loss": 1.005,
+      "step": 45820
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.340375759588173e-05,
+      "loss": 1.0071,
+      "step": 45825
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.338045931206224e-05,
+      "loss": 0.9861,
+      "step": 45830
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.335716332359725e-05,
+      "loss": 0.9195,
+      "step": 45835
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.333386963194699e-05,
+      "loss": 1.1381,
+      "step": 45840
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.331057823857156e-05,
+      "loss": 0.8934,
+      "step": 45845
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.328728914493085e-05,
+      "loss": 0.8748,
+      "step": 45850
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.326400235248466e-05,
+      "loss": 0.8566,
+      "step": 45855
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.50390625,
+      "learning_rate": 6.324071786269268e-05,
+      "loss": 0.8186,
+      "step": 45860
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.321743567701435e-05,
+      "loss": 1.1196,
+      "step": 45865
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.319415579690902e-05,
+      "loss": 0.9681,
+      "step": 45870
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.317087822383596e-05,
+      "loss": 0.8658,
+      "step": 45875
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.314760295925418e-05,
+      "loss": 1.0102,
+      "step": 45880
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.312433000462266e-05,
+      "loss": 1.0275,
+      "step": 45885
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.310105936140009e-05,
+      "loss": 1.1024,
+      "step": 45890
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.307779103104513e-05,
+      "loss": 0.876,
+      "step": 45895
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.498046875,
+      "learning_rate": 6.30545250150163e-05,
+      "loss": 0.8698,
+      "step": 45900
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.303126131477193e-05,
+      "loss": 0.8891,
+      "step": 45905
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.300799993177017e-05,
+      "loss": 0.8649,
+      "step": 45910
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.298474086746913e-05,
+      "loss": 0.9766,
+      "step": 45915
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.29614841233267e-05,
+      "loss": 0.9994,
+      "step": 45920
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.293822970080059e-05,
+      "loss": 0.9491,
+      "step": 45925
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.291497760134848e-05,
+      "loss": 0.8919,
+      "step": 45930
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49609375,
+      "learning_rate": 6.289172782642782e-05,
+      "loss": 1.018,
+      "step": 45935
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.286848037749593e-05,
+      "loss": 0.9458,
+      "step": 45940
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.284523525600996e-05,
+      "loss": 0.8374,
+      "step": 45945
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.282199246342694e-05,
+      "loss": 0.9878,
+      "step": 45950
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.27987520012038e-05,
+      "loss": 0.874,
+      "step": 45955
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.277551387079725e-05,
+      "loss": 0.9338,
+      "step": 45960
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.275227807366387e-05,
+      "loss": 0.95,
+      "step": 45965
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.625,
+      "learning_rate": 6.272904461126012e-05,
+      "loss": 0.9569,
+      "step": 45970
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.270581348504233e-05,
+      "loss": 0.9139,
+      "step": 45975
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.494140625,
+      "learning_rate": 6.26825846964666e-05,
+      "loss": 1.0085,
+      "step": 45980
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.265935824698897e-05,
+      "loss": 1.03,
+      "step": 45985
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.263613413806532e-05,
+      "loss": 0.9165,
+      "step": 45990
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.261291237115132e-05,
+      "loss": 1.1229,
+      "step": 45995
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.258969294770255e-05,
+      "loss": 0.9052,
+      "step": 46000
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.256647586917441e-05,
+      "loss": 0.9974,
+      "step": 46005
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.50390625,
+      "learning_rate": 6.254326113702222e-05,
+      "loss": 0.8159,
+      "step": 46010
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.252004875270107e-05,
+      "loss": 0.9784,
+      "step": 46015
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.50390625,
+      "learning_rate": 6.249683871766594e-05,
+      "loss": 0.8097,
+      "step": 46020
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.24736310333717e-05,
+      "loss": 1.0112,
+      "step": 46025
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.245042570127299e-05,
+      "loss": 0.9121,
+      "step": 46030
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.242722272282436e-05,
+      "loss": 0.9148,
+      "step": 46035
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.240402209948024e-05,
+      "loss": 1.0444,
+      "step": 46040
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.23808238326948e-05,
+      "loss": 0.8895,
+      "step": 46045
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.235762792392215e-05,
+      "loss": 0.839,
+      "step": 46050
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.233443437461628e-05,
+      "loss": 1.0243,
+      "step": 46055
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.231124318623094e-05,
+      "loss": 0.81,
+      "step": 46060
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.22880543602198e-05,
+      "loss": 1.1031,
+      "step": 46065
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.226486789803638e-05,
+      "loss": 0.9281,
+      "step": 46070
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.2241683801134e-05,
+      "loss": 1.0277,
+      "step": 46075
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.221850207096589e-05,
+      "loss": 0.9444,
+      "step": 46080
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.219532270898511e-05,
+      "loss": 0.9583,
+      "step": 46085
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.439453125,
+      "learning_rate": 6.217214571664453e-05,
+      "loss": 0.7989,
+      "step": 46090
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.214897109539701e-05,
+      "loss": 1.0164,
+      "step": 46095
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.212579884669503e-05,
+      "loss": 0.9877,
+      "step": 46100
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.210262897199113e-05,
+      "loss": 0.8748,
+      "step": 46105
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.20794614727376e-05,
+      "loss": 1.0434,
+      "step": 46110
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.20562963503866e-05,
+      "loss": 1.0184,
+      "step": 46115
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.203313360639019e-05,
+      "loss": 0.9778,
+      "step": 46120
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.200997324220018e-05,
+      "loss": 1.0332,
+      "step": 46125
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.198681525926831e-05,
+      "loss": 0.9329,
+      "step": 46130
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.196365965904617e-05,
+      "loss": 0.8952,
+      "step": 46135
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.625,
+      "learning_rate": 6.194050644298517e-05,
+      "loss": 0.8317,
+      "step": 46140
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.191735561253658e-05,
+      "loss": 1.0471,
+      "step": 46145
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.189420716915149e-05,
+      "loss": 1.0214,
+      "step": 46150
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.187106111428089e-05,
+      "loss": 1.022,
+      "step": 46155
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.184791744937562e-05,
+      "loss": 0.9737,
+      "step": 46160
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.765625,
+      "learning_rate": 6.182477617588634e-05,
+      "loss": 0.9934,
+      "step": 46165
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.180163729526353e-05,
+      "loss": 0.9681,
+      "step": 46170
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.177850080895764e-05,
+      "loss": 0.9406,
+      "step": 46175
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.175536671841882e-05,
+      "loss": 0.8751,
+      "step": 46180
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.498046875,
+      "learning_rate": 6.173223502509717e-05,
+      "loss": 0.8101,
+      "step": 46185
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.170910573044265e-05,
+      "loss": 0.8607,
+      "step": 46190
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.168597883590495e-05,
+      "loss": 0.954,
+      "step": 46195
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.166285434293378e-05,
+      "loss": 0.9483,
+      "step": 46200
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.494140625,
+      "learning_rate": 6.163973225297855e-05,
+      "loss": 1.0113,
+      "step": 46205
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.486328125,
+      "learning_rate": 6.161661256748856e-05,
+      "loss": 0.9647,
+      "step": 46210
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.159349528791302e-05,
+      "loss": 1.0275,
+      "step": 46215
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.157038041570094e-05,
+      "loss": 0.9002,
+      "step": 46220
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.154726795230117e-05,
+      "loss": 1.0381,
+      "step": 46225
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.152415789916246e-05,
+      "loss": 0.9387,
+      "step": 46230
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.150105025773335e-05,
+      "loss": 0.9686,
+      "step": 46235
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.147794502946223e-05,
+      "loss": 1.0323,
+      "step": 46240
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.145484221579743e-05,
+      "loss": 0.8602,
+      "step": 46245
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.143174181818701e-05,
+      "loss": 0.866,
+      "step": 46250
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.140864383807894e-05,
+      "loss": 1.0404,
+      "step": 46255
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.138554827692103e-05,
+      "loss": 1.0325,
+      "step": 46260
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.13624551361609e-05,
+      "loss": 1.0028,
+      "step": 46265
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.13393644172461e-05,
+      "loss": 0.7928,
+      "step": 46270
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.131627612162397e-05,
+      "loss": 1.0011,
+      "step": 46275
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.12931902507417e-05,
+      "loss": 0.9153,
+      "step": 46280
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.127010680604636e-05,
+      "loss": 0.9783,
+      "step": 46285
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.124702578898484e-05,
+      "loss": 1.0021,
+      "step": 46290
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.122394720100386e-05,
+      "loss": 0.9419,
+      "step": 46295
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.120087104355006e-05,
+      "loss": 0.9133,
+      "step": 46300
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.482421875,
+      "learning_rate": 6.117779731806986e-05,
+      "loss": 0.9986,
+      "step": 46305
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.115472602600951e-05,
+      "loss": 0.9771,
+      "step": 46310
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.11316571688152e-05,
+      "loss": 1.0309,
+      "step": 46315
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.110859074793284e-05,
+      "loss": 0.9542,
+      "step": 46320
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.108552676480834e-05,
+      "loss": 0.8965,
+      "step": 46325
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.106246522088735e-05,
+      "loss": 0.9091,
+      "step": 46330
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.103940611761535e-05,
+      "loss": 0.9457,
+      "step": 46335
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.101634945643777e-05,
+      "loss": 1.0486,
+      "step": 46340
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.099329523879981e-05,
+      "loss": 1.0533,
+      "step": 46345
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.09702434661465e-05,
+      "loss": 1.1038,
+      "step": 46350
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.0947194139922824e-05,
+      "loss": 0.8647,
+      "step": 46355
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.092414726157348e-05,
+      "loss": 0.9562,
+      "step": 46360
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.478515625,
+      "learning_rate": 6.090110283254309e-05,
+      "loss": 0.9561,
+      "step": 46365
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.087806085427611e-05,
+      "loss": 0.8962,
+      "step": 46370
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.486328125,
+      "learning_rate": 6.08550213282168e-05,
+      "loss": 0.8846,
+      "step": 46375
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.083198425580936e-05,
+      "loss": 0.8637,
+      "step": 46380
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.080894963849776e-05,
+      "loss": 0.907,
+      "step": 46385
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.0785917477725806e-05,
+      "loss": 1.0066,
+      "step": 46390
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.076288777493723e-05,
+      "loss": 0.925,
+      "step": 46395
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.48046875,
+      "learning_rate": 6.073986053157553e-05,
+      "loss": 0.9988,
+      "step": 46400
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.071683574908407e-05,
+      "loss": 1.0245,
+      "step": 46405
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.0693813428906124e-05,
+      "loss": 1.0269,
+      "step": 46410
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.0670793572484696e-05,
+      "loss": 0.8969,
+      "step": 46415
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.064777618126272e-05,
+      "loss": 0.8421,
+      "step": 46420
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.062476125668293e-05,
+      "loss": 1.0216,
+      "step": 46425
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.060174880018798e-05,
+      "loss": 1.041,
+      "step": 46430
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.478515625,
+      "learning_rate": 6.0578738813220273e-05,
+      "loss": 0.9485,
+      "step": 46435
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.055573129722212e-05,
+      "loss": 1.0267,
+      "step": 46440
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.053272625363562e-05,
+      "loss": 0.98,
+      "step": 46445
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.050972368390282e-05,
+      "loss": 0.994,
+      "step": 46450
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.048672358946552e-05,
+      "loss": 1.0051,
+      "step": 46455
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.0463725971765396e-05,
+      "loss": 1.0058,
+      "step": 46460
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.044073083224393e-05,
+      "loss": 0.9911,
+      "step": 46465
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.041773817234251e-05,
+      "loss": 0.908,
+      "step": 46470
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.0394747993502354e-05,
+      "loss": 1.0566,
+      "step": 46475
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.037176029716448e-05,
+      "loss": 0.9758,
+      "step": 46480
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.034877508476982e-05,
+      "loss": 1.0319,
+      "step": 46485
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.0325792357759106e-05,
+      "loss": 0.855,
+      "step": 46490
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.0302812117572883e-05,
+      "loss": 0.963,
+      "step": 46495
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.027983436565163e-05,
+      "loss": 0.9689,
+      "step": 46500
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.0256859103435595e-05,
+      "loss": 1.0157,
+      "step": 46505
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.0233886332364894e-05,
+      "loss": 1.0156,
+      "step": 46510
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.0210916053879515e-05,
+      "loss": 1.0249,
+      "step": 46515
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.018794826941918e-05,
+      "loss": 0.9484,
+      "step": 46520
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.0164982980423636e-05,
+      "loss": 0.9495,
+      "step": 46525
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.01420201883323e-05,
+      "loss": 1.0027,
+      "step": 46530
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.011905989458453e-05,
+      "loss": 0.967,
+      "step": 46535
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.009610210061951e-05,
+      "loss": 1.1281,
+      "step": 46540
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5,
+      "learning_rate": 6.0073146807876266e-05,
+      "loss": 0.8276,
+      "step": 46545
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.0050194017793635e-05,
+      "loss": 1.025,
+      "step": 46550
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.0027243731810355e-05,
+      "loss": 0.9618,
+      "step": 46555
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.000429595136497e-05,
+      "loss": 1.0105,
+      "step": 46560
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5,
+      "learning_rate": 5.998135067789589e-05,
+      "loss": 0.8233,
+      "step": 46565
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.99584079128413e-05,
+      "loss": 0.9368,
+      "step": 46570
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.9935467657639286e-05,
+      "loss": 0.9314,
+      "step": 46575
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.99125299137278e-05,
+      "loss": 0.921,
+      "step": 46580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.9889594682544604e-05,
+      "loss": 0.9308,
+      "step": 46585
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.986666196552728e-05,
+      "loss": 1.0665,
+      "step": 46590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.984373176411331e-05,
+      "loss": 0.8952,
+      "step": 46595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.982080407973996e-05,
+      "loss": 0.8521,
+      "step": 46600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.9797878913844364e-05,
+      "loss": 1.0417,
+      "step": 46605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.977495626786351e-05,
+      "loss": 1.0472,
+      "step": 46610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.423828125,
+      "learning_rate": 5.975203614323421e-05,
+      "loss": 0.9462,
+      "step": 46615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.9729118541393156e-05,
+      "loss": 1.0091,
+      "step": 46620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.970620346377678e-05,
+      "loss": 1.0154,
+      "step": 46625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.968329091182145e-05,
+      "loss": 0.9673,
+      "step": 46630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.966038088696338e-05,
+      "loss": 1.0245,
+      "step": 46635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.963747339063859e-05,
+      "loss": 0.925,
+      "step": 46640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.96145684242829e-05,
+      "loss": 0.8854,
+      "step": 46645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.959166598933209e-05,
+      "loss": 0.9141,
+      "step": 46650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.956876608722167e-05,
+      "loss": 0.8664,
+      "step": 46655
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.954586871938702e-05,
+      "loss": 0.9962,
+      "step": 46660
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.494140625,
+      "learning_rate": 5.952297388726342e-05,
+      "loss": 0.9225,
+      "step": 46665
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.950008159228593e-05,
+      "loss": 0.8407,
+      "step": 46670
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.9477191835889425e-05,
+      "loss": 1.0356,
+      "step": 46675
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.945430461950869e-05,
+      "loss": 1.0125,
+      "step": 46680
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.9431419944578305e-05,
+      "loss": 1.1157,
+      "step": 46685
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.940853781253274e-05,
+      "loss": 1.0508,
+      "step": 46690
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.938565822480625e-05,
+      "loss": 1.0621,
+      "step": 46695
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.936278118283294e-05,
+      "loss": 1.008,
+      "step": 46700
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.933990668804681e-05,
+      "loss": 0.9433,
+      "step": 46705
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.931703474188164e-05,
+      "loss": 0.9275,
+      "step": 46710
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.51171875,
+      "learning_rate": 5.9294165345771036e-05,
+      "loss": 0.9586,
+      "step": 46715
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.9271298501148545e-05,
+      "loss": 0.8778,
+      "step": 46720
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.9248434209447456e-05,
+      "loss": 0.9347,
+      "step": 46725
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.9225572472100895e-05,
+      "loss": 1.152,
+      "step": 46730
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.92027132905419e-05,
+      "loss": 0.9424,
+      "step": 46735
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.917985666620329e-05,
+      "loss": 1.0837,
+      "step": 46740
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.9157002600517766e-05,
+      "loss": 0.9528,
+      "step": 46745
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.9134151094917834e-05,
+      "loss": 0.9801,
+      "step": 46750
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.9111302150835836e-05,
+      "loss": 0.9547,
+      "step": 46755
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.908845576970401e-05,
+      "loss": 0.96,
+      "step": 46760
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.46875,
+      "learning_rate": 5.906561195295436e-05,
+      "loss": 0.7969,
+      "step": 46765
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.904277070201876e-05,
+      "loss": 0.9844,
+      "step": 46770
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.484375,
+      "learning_rate": 5.901993201832901e-05,
+      "loss": 0.9173,
+      "step": 46775
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.89970959033165e-05,
+      "loss": 0.8882,
+      "step": 46780
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.897426235841277e-05,
+      "loss": 0.9216,
+      "step": 46785
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.895143138504899e-05,
+      "loss": 0.963,
+      "step": 46790
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.478515625,
+      "learning_rate": 5.8928602984656213e-05,
+      "loss": 0.9357,
+      "step": 46795
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.890577715866541e-05,
+      "loss": 0.7911,
+      "step": 46800
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.46484375,
+      "learning_rate": 5.888295390850729e-05,
+      "loss": 0.9172,
+      "step": 46805
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.886013323561244e-05,
+      "loss": 0.8876,
+      "step": 46810
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.8837315141411284e-05,
+      "loss": 0.9875,
+      "step": 46815
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.881449962733412e-05,
+      "loss": 0.8395,
+      "step": 46820
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.8791686694811e-05,
+      "loss": 1.0331,
+      "step": 46825
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.876887634527195e-05,
+      "loss": 1.0062,
+      "step": 46830
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.874606858014662e-05,
+      "loss": 1.0152,
+      "step": 46835
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.872326340086474e-05,
+      "loss": 0.887,
+      "step": 46840
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.87004608088557e-05,
+      "loss": 0.7325,
+      "step": 46845
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.86776608055488e-05,
+      "loss": 1.0844,
+      "step": 46850
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.8654863392373185e-05,
+      "loss": 0.9183,
+      "step": 46855
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.863206857075785e-05,
+      "loss": 0.9098,
+      "step": 46860
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.860927634213154e-05,
+      "loss": 0.9646,
+      "step": 46865
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.8586486707922924e-05,
+      "loss": 0.9475,
+      "step": 46870
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.8563699669560524e-05,
+      "loss": 0.94,
+      "step": 46875
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.8540915228472624e-05,
+      "loss": 0.9383,
+      "step": 46880
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.8518133386087356e-05,
+      "loss": 1.0302,
+      "step": 46885
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.8495354143832716e-05,
+      "loss": 0.9661,
+      "step": 46890
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.847257750313656e-05,
+      "loss": 1.0525,
+      "step": 46895
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.8449803465426545e-05,
+      "loss": 0.8351,
+      "step": 46900
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.842703203213016e-05,
+      "loss": 0.992,
+      "step": 46905
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.8404263204674756e-05,
+      "loss": 1.0304,
+      "step": 46910
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.8381496984487495e-05,
+      "loss": 0.7929,
+      "step": 46915
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.835873337299544e-05,
+      "loss": 1.0232,
+      "step": 46920
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.833597237162538e-05,
+      "loss": 0.952,
+      "step": 46925
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.8313213981804005e-05,
+      "loss": 0.9321,
+      "step": 46930
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.829045820495791e-05,
+      "loss": 0.8802,
+      "step": 46935
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.826770504251339e-05,
+      "loss": 1.1071,
+      "step": 46940
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.82449544958966e-05,
+      "loss": 1.0217,
+      "step": 46945
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.48046875,
+      "learning_rate": 5.822220656653363e-05,
+      "loss": 0.891,
+      "step": 46950
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.8199461255850365e-05,
+      "loss": 0.9092,
+      "step": 46955
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.8176718565272436e-05,
+      "loss": 1.0907,
+      "step": 46960
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.8153978496225415e-05,
+      "loss": 0.9233,
+      "step": 46965
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.813124105013473e-05,
+      "loss": 0.962,
+      "step": 46970
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.81085062284255e-05,
+      "loss": 0.9517,
+      "step": 46975
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.8085774032522814e-05,
+      "loss": 0.919,
+      "step": 46980
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.462890625,
+      "learning_rate": 5.8063044463851623e-05,
+      "loss": 0.8406,
+      "step": 46985
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.8040317523836475e-05,
+      "loss": 0.9927,
+      "step": 46990
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.8017593213902036e-05,
+      "loss": 0.8488,
+      "step": 46995
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.7994871535472684e-05,
+      "loss": 0.9303,
+      "step": 47000
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.79721524899726e-05,
+      "loss": 0.9435,
+      "step": 47005
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.794943607882586e-05,
+      "loss": 0.8904,
+      "step": 47010
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.792672230345639e-05,
+      "loss": 1.0229,
+      "step": 47015
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.790401116528785e-05,
+      "loss": 0.7795,
+      "step": 47020
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.788130266574383e-05,
+      "loss": 0.9443,
+      "step": 47025
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5,
+      "learning_rate": 5.785859680624779e-05,
+      "loss": 0.987,
+      "step": 47030
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.783589358822283e-05,
+      "loss": 1.0296,
+      "step": 47035
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4921875,
+      "learning_rate": 5.781319301309214e-05,
+      "loss": 1.0317,
+      "step": 47040
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.77904950822785e-05,
+      "loss": 0.9373,
+      "step": 47045
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.776779979720475e-05,
+      "loss": 0.9986,
+      "step": 47050
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.7745107159293365e-05,
+      "loss": 1.0654,
+      "step": 47055
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.772241716996679e-05,
+      "loss": 0.9629,
+      "step": 47060
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.76997298306473e-05,
+      "loss": 0.8715,
+      "step": 47065
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.482421875,
+      "learning_rate": 5.7677045142756866e-05,
+      "loss": 0.9028,
+      "step": 47070
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.765436310771746e-05,
+      "loss": 0.9967,
+      "step": 47075
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.763168372695085e-05,
+      "loss": 0.8681,
+      "step": 47080
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.76090070018785e-05,
+      "loss": 1.0474,
+      "step": 47085
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.758633293392191e-05,
+      "loss": 0.927,
+      "step": 47090
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.7563661524502275e-05,
+      "loss": 0.9565,
+      "step": 47095
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.754099277504064e-05,
+      "loss": 0.8303,
+      "step": 47100
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.751832668695792e-05,
+      "loss": 1.017,
+      "step": 47105
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.749566326167492e-05,
+      "loss": 0.99,
+      "step": 47110
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.7473002500612114e-05,
+      "loss": 1.0008,
+      "step": 47115
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.48828125,
+      "learning_rate": 5.745034440518993e-05,
+      "loss": 0.9646,
+      "step": 47120
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.742768897682868e-05,
+      "loss": 0.9078,
+      "step": 47125
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.7405036216948315e-05,
+      "loss": 0.9729,
+      "step": 47130
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.73823861269688e-05,
+      "loss": 0.9689,
+      "step": 47135
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.7359738708309885e-05,
+      "loss": 0.987,
+      "step": 47140
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.733709396239113e-05,
+      "loss": 0.9306,
+      "step": 47145
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.731445189063187e-05,
+      "loss": 0.8408,
+      "step": 47150
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.51171875,
+      "learning_rate": 5.729181249445137e-05,
+      "loss": 0.9015,
+      "step": 47155
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.726917577526876e-05,
+      "loss": 0.9337,
+      "step": 47160
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.724654173450281e-05,
+      "loss": 0.6944,
+      "step": 47165
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.722391037357234e-05,
+      "loss": 1.0884,
+      "step": 47170
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.720128169389593e-05,
+      "loss": 0.8736,
+      "step": 47175
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.717865569689187e-05,
+      "loss": 0.9979,
+      "step": 47180
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.715603238397845e-05,
+      "loss": 0.9299,
+      "step": 47185
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.713341175657376e-05,
+      "loss": 0.9496,
+      "step": 47190
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.7110793816095634e-05,
+      "loss": 0.8789,
+      "step": 47195
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.708817856396176e-05,
+      "loss": 0.9788,
+      "step": 47200
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.7065566001589764e-05,
+      "loss": 0.837,
+      "step": 47205
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.704295613039694e-05,
+      "loss": 0.8921,
+      "step": 47210
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.7020348951800574e-05,
+      "loss": 1.0042,
+      "step": 47215
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.699774446721771e-05,
+      "loss": 0.9389,
+      "step": 47220
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.6975142678065166e-05,
+      "loss": 0.8968,
+      "step": 47225
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.695254358575966e-05,
+      "loss": 0.9572,
+      "step": 47230
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.6929947191717804e-05,
+      "loss": 0.9398,
+      "step": 47235
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5,
+      "learning_rate": 5.6907353497355874e-05,
+      "loss": 0.8282,
+      "step": 47240
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.688476250409011e-05,
+      "loss": 0.8719,
+      "step": 47245
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.686217421333656e-05,
+      "loss": 0.9109,
+      "step": 47250
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.683958862651106e-05,
+      "loss": 0.9396,
+      "step": 47255
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.48046875,
+      "learning_rate": 5.6817005745029284e-05,
+      "loss": 0.9259,
+      "step": 47260
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.679442557030674e-05,
+      "loss": 0.9045,
+      "step": 47265
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.6771848103758875e-05,
+      "loss": 0.9453,
+      "step": 47270
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.674927334680075e-05,
+      "loss": 0.9764,
+      "step": 47275
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.672670130084743e-05,
+      "loss": 0.9651,
+      "step": 47280
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.670413196731379e-05,
+      "loss": 0.8923,
+      "step": 47285
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.668156534761444e-05,
+      "loss": 0.8737,
+      "step": 47290
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.665900144316391e-05,
+      "loss": 0.9184,
+      "step": 47295
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.48828125,
+      "learning_rate": 5.66364402553766e-05,
+      "loss": 1.0944,
+      "step": 47300
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.661388178566653e-05,
+      "loss": 0.9557,
+      "step": 47305
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.659132603544776e-05,
+      "loss": 1.2007,
+      "step": 47310
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.6568773006134144e-05,
+      "loss": 0.7776,
+      "step": 47315
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.654622269913927e-05,
+      "loss": 1.1321,
+      "step": 47320
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.652367511587665e-05,
+      "loss": 0.957,
+      "step": 47325
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.650113025775963e-05,
+      "loss": 0.9662,
+      "step": 47330
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.647858812620126e-05,
+      "loss": 0.8585,
+      "step": 47335
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.645604872261455e-05,
+      "loss": 0.9303,
+      "step": 47340
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.643351204841234e-05,
+      "loss": 0.9571,
+      "step": 47345
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.6410978105007175e-05,
+      "loss": 0.9188,
+      "step": 47350
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.6388446893811585e-05,
+      "loss": 0.9066,
+      "step": 47355
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.6365918416237775e-05,
+      "loss": 0.9174,
+      "step": 47360
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.634339267369794e-05,
+      "loss": 0.9251,
+      "step": 47365
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.6320869667603925e-05,
+      "loss": 0.9401,
+      "step": 47370
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.629834939936755e-05,
+      "loss": 1.0269,
+      "step": 47375
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.6275831870400444e-05,
+      "loss": 0.9956,
+      "step": 47380
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.625331708211394e-05,
+      "loss": 0.9787,
+      "step": 47385
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.6230805035919365e-05,
+      "loss": 0.9678,
+      "step": 47390
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.620829573322781e-05,
+      "loss": 0.8578,
+      "step": 47395
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.618578917545012e-05,
+      "loss": 0.8684,
+      "step": 47400
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.61632853639971e-05,
+      "loss": 1.0012,
+      "step": 47405
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.614078430027924e-05,
+      "loss": 1.0491,
+      "step": 47410
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.6118285985707006e-05,
+      "loss": 0.9195,
+      "step": 47415
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.6095790421690554e-05,
+      "loss": 0.8142,
+      "step": 47420
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.607329760964e-05,
+      "loss": 0.9404,
+      "step": 47425
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.6050807550965125e-05,
+      "loss": 0.8679,
+      "step": 47430
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.60283202470757e-05,
+      "loss": 0.826,
+      "step": 47435
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.600583569938127e-05,
+      "loss": 0.9905,
+      "step": 47440
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.598335390929114e-05,
+      "loss": 0.9148,
+      "step": 47445
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.596087487821451e-05,
+      "loss": 1.0001,
+      "step": 47450
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.593839860756044e-05,
+      "loss": 0.9885,
+      "step": 47455
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.591592509873772e-05,
+      "loss": 0.9997,
+      "step": 47460
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.589345435315498e-05,
+      "loss": 0.9769,
+      "step": 47465
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.587098637222077e-05,
+      "loss": 0.892,
+      "step": 47470
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.584852115734343e-05,
+      "loss": 0.9215,
+      "step": 47475
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.582605870993103e-05,
+      "loss": 0.8581,
+      "step": 47480
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.580359903139156e-05,
+      "loss": 0.8876,
+      "step": 47485
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.5781142123132904e-05,
+      "loss": 1.1096,
+      "step": 47490
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.575868798656256e-05,
+      "loss": 0.9166,
+      "step": 47495
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.573623662308804e-05,
+      "loss": 0.9829,
+      "step": 47500
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.5713788034116673e-05,
+      "loss": 0.9592,
+      "step": 47505
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.48046875,
+      "learning_rate": 5.569134222105551e-05,
+      "loss": 0.9754,
+      "step": 47510
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.566889918531144e-05,
+      "loss": 0.9408,
+      "step": 47515
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.564645892829126e-05,
+      "loss": 0.9398,
+      "step": 47520
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.5624021451401576e-05,
+      "loss": 0.9964,
+      "step": 47525
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.560158675604873e-05,
+      "loss": 0.9931,
+      "step": 47530
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.5579154843639e-05,
+      "loss": 0.9833,
+      "step": 47535
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.555672571557846e-05,
+      "loss": 0.9731,
+      "step": 47540
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.5534299373272944e-05,
+      "loss": 1.032,
+      "step": 47545
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5,
+      "learning_rate": 5.5511875818128176e-05,
+      "loss": 0.9431,
+      "step": 47550
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.5489455051549756e-05,
+      "loss": 0.9872,
+      "step": 47555
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.546703707494293e-05,
+      "loss": 0.9955,
+      "step": 47560
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.544462188971298e-05,
+      "loss": 0.8273,
+      "step": 47565
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.765625,
+      "learning_rate": 5.5422209497264885e-05,
+      "loss": 1.0531,
+      "step": 47570
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.5399799899003414e-05,
+      "loss": 0.9159,
+      "step": 47575
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.53773930963333e-05,
+      "loss": 0.8641,
+      "step": 47580
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.535498909065904e-05,
+      "loss": 0.83,
+      "step": 47585
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.5332587883384854e-05,
+      "loss": 0.9063,
+      "step": 47590
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.5310189475914956e-05,
+      "loss": 1.099,
+      "step": 47595
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.5287793869653305e-05,
+      "loss": 0.8519,
+      "step": 47600
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.5265401066003617e-05,
+      "loss": 1.0752,
+      "step": 47605
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.5243011066369534e-05,
+      "loss": 0.8892,
+      "step": 47610
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.522062387215454e-05,
+      "loss": 0.9662,
+      "step": 47615
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.519823948476184e-05,
+      "loss": 0.9145,
+      "step": 47620
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.517585790559448e-05,
+      "loss": 0.9575,
+      "step": 47625
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.515347913605539e-05,
+      "loss": 0.9744,
+      "step": 47630
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.513110317754734e-05,
+      "loss": 0.8392,
+      "step": 47635
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.510873003147281e-05,
+      "loss": 1.0129,
+      "step": 47640
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.5086359699234205e-05,
+      "loss": 0.9679,
+      "step": 47645
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.506399218223377e-05,
+      "loss": 1.0772,
+      "step": 47650
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.504162748187344e-05,
+      "loss": 0.9394,
+      "step": 47655
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.50192655995551e-05,
+      "loss": 0.9517,
+      "step": 47660
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.499690653668046e-05,
+      "loss": 0.9318,
+      "step": 47665
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.4974550294650985e-05,
+      "loss": 0.9086,
+      "step": 47670
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.495219687486792e-05,
+      "loss": 0.9837,
+      "step": 47675
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.4929846278732525e-05,
+      "loss": 0.9616,
+      "step": 47680
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.4907498507645626e-05,
+      "loss": 0.9245,
+      "step": 47685
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.4885153563008095e-05,
+      "loss": 1.0802,
+      "step": 47690
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.4862811446220564e-05,
+      "loss": 0.9033,
+      "step": 47695
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.484047215868336e-05,
+      "loss": 0.9888,
+      "step": 47700
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.4818135701796814e-05,
+      "loss": 0.9262,
+      "step": 47705
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.4795802076961e-05,
+      "loss": 0.9719,
+      "step": 47710
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.477347128557577e-05,
+      "loss": 0.8993,
+      "step": 47715
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.47511433290409e-05,
+      "loss": 0.8943,
+      "step": 47720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.4728818208755864e-05,
+      "loss": 0.909,
+      "step": 47725
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.498046875,
+      "learning_rate": 5.47064959261201e-05,
+      "loss": 0.9259,
+      "step": 47730
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.625,
+      "learning_rate": 5.4684176482532723e-05,
+      "loss": 0.9023,
+      "step": 47735
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.4661859879392764e-05,
+      "loss": 0.8699,
+      "step": 47740
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.463954611809912e-05,
+      "loss": 0.9654,
+      "step": 47745
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.72265625,
+      "learning_rate": 5.461723520005033e-05,
+      "loss": 0.9852,
+      "step": 47750
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.459492712664493e-05,
+      "loss": 0.8854,
+      "step": 47755
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.457262189928125e-05,
+      "loss": 1.0325,
+      "step": 47760
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.455031951935732e-05,
+      "loss": 1.0155,
+      "step": 47765
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.452801998827114e-05,
+      "loss": 0.9049,
+      "step": 47770
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.450572330742052e-05,
+      "loss": 0.978,
+      "step": 47775
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.4483429478202906e-05,
+      "loss": 0.9369,
+      "step": 47780
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.446113850201578e-05,
+      "loss": 1.041,
+      "step": 47785
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.443885038025638e-05,
+      "loss": 0.9147,
+      "step": 47790
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.441656511432169e-05,
+      "loss": 0.854,
+      "step": 47795
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.466796875,
+      "learning_rate": 5.4394282705608626e-05,
+      "loss": 0.918,
+      "step": 47800
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.4372003155513896e-05,
+      "loss": 0.8515,
+      "step": 47805
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.434972646543393e-05,
+      "loss": 1.1099,
+      "step": 47810
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.432745263676511e-05,
+      "loss": 0.8933,
+      "step": 47815
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5,
+      "learning_rate": 5.430518167090359e-05,
+      "loss": 1.062,
+      "step": 47820
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.428291356924535e-05,
+      "loss": 0.8631,
+      "step": 47825
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.4260648333186114e-05,
+      "loss": 0.998,
+      "step": 47830
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.423838596412152e-05,
+      "loss": 0.9869,
+      "step": 47835
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.421612646344707e-05,
+      "loss": 1.0076,
+      "step": 47840
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.41938698325579e-05,
+      "loss": 0.9925,
+      "step": 47845
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.49609375,
+      "learning_rate": 5.417161607284915e-05,
+      "loss": 0.8627,
+      "step": 47850
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.625,
+      "learning_rate": 5.4149365185715726e-05,
+      "loss": 1.0033,
+      "step": 47855
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.412711717255228e-05,
+      "loss": 1.0183,
+      "step": 47860
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.410487203475338e-05,
+      "loss": 0.9942,
+      "step": 47865
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.40826297737134e-05,
+      "loss": 0.9463,
+      "step": 47870
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.4060390390826444e-05,
+      "loss": 0.9178,
+      "step": 47875
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.403815388748659e-05,
+      "loss": 1.0525,
+      "step": 47880
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.4015920265087574e-05,
+      "loss": 0.8748,
+      "step": 47885
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.49609375,
+      "learning_rate": 5.399368952502302e-05,
+      "loss": 0.91,
+      "step": 47890
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.397146166868641e-05,
+      "loss": 0.9965,
+      "step": 47895
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.3949236697471025e-05,
+      "loss": 0.8512,
+      "step": 47900
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.3927014612769925e-05,
+      "loss": 1.1435,
+      "step": 47905
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.3904795415975996e-05,
+      "loss": 0.92,
+      "step": 47910
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.3882579108482024e-05,
+      "loss": 0.9276,
+      "step": 47915
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.386036569168049e-05,
+      "loss": 1.0497,
+      "step": 47920
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.383815516696378e-05,
+      "loss": 0.9812,
+      "step": 47925
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.3815947535724124e-05,
+      "loss": 0.973,
+      "step": 47930
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.3793742799353464e-05,
+      "loss": 1.0856,
+      "step": 47935
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.377154095924359e-05,
+      "loss": 0.9601,
+      "step": 47940
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.3749342016786186e-05,
+      "loss": 0.9086,
+      "step": 47945
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.372714597337274e-05,
+      "loss": 0.9846,
+      "step": 47950
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.370495283039445e-05,
+      "loss": 1.0759,
+      "step": 47955
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.368276258924243e-05,
+      "loss": 0.9919,
+      "step": 47960
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.3660575251307646e-05,
+      "loss": 0.9385,
+      "step": 47965
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.3638390817980744e-05,
+      "loss": 0.8852,
+      "step": 47970
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.3616209290652296e-05,
+      "loss": 0.8375,
+      "step": 47975
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.359403067071274e-05,
+      "loss": 1.0317,
+      "step": 47980
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.625,
+      "learning_rate": 5.357185495955217e-05,
+      "loss": 0.9038,
+      "step": 47985
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4921875,
+      "learning_rate": 5.3549682158560574e-05,
+      "loss": 0.9486,
+      "step": 47990
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.3527512269127844e-05,
+      "loss": 0.9461,
+      "step": 47995
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.48828125,
+      "learning_rate": 5.350534529264353e-05,
+      "loss": 0.9131,
+      "step": 48000
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.348318123049714e-05,
+      "loss": 0.9683,
+      "step": 48005
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.3461020084077964e-05,
+      "loss": 0.9655,
+      "step": 48010
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.3438861854775e-05,
+      "loss": 0.8866,
+      "step": 48015
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.341670654397721e-05,
+      "loss": 0.924,
+      "step": 48020
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.339455415307335e-05,
+      "loss": 0.9218,
+      "step": 48025
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.337240468345187e-05,
+      "loss": 0.8772,
+      "step": 48030
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.335025813650121e-05,
+      "loss": 0.9987,
+      "step": 48035
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.332811451360946e-05,
+      "loss": 0.9083,
+      "step": 48040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.330597381616469e-05,
+      "loss": 1.1062,
+      "step": 48045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.73828125,
+      "learning_rate": 5.328383604555462e-05,
+      "loss": 0.9678,
+      "step": 48050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.48046875,
+      "learning_rate": 5.326170120316691e-05,
+      "loss": 0.8508,
+      "step": 48055
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.3239569290389043e-05,
+      "loss": 1.0304,
+      "step": 48060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5,
+      "learning_rate": 5.321744030860819e-05,
+      "loss": 0.9192,
+      "step": 48065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.319531425921146e-05,
+      "loss": 1.0754,
+      "step": 48070
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.317319114358576e-05,
+      "loss": 1.098,
+      "step": 48075
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.9140625,
+      "learning_rate": 5.3151070963117754e-05,
+      "loss": 0.9126,
+      "step": 48080
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.3128953719193975e-05,
+      "loss": 1.1703,
+      "step": 48085
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.310683941320082e-05,
+      "loss": 1.0177,
+      "step": 48090
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.3084728046524315e-05,
+      "loss": 0.9678,
+      "step": 48095
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.306261962055048e-05,
+      "loss": 0.9722,
+      "step": 48100
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.3040514136665154e-05,
+      "loss": 1.0295,
+      "step": 48105
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.3018411596253824e-05,
+      "loss": 1.0052,
+      "step": 48110
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.2996312000701986e-05,
+      "loss": 0.9311,
+      "step": 48115
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.2974215351394863e-05,
+      "loss": 1.0246,
+      "step": 48120
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.474609375,
+      "learning_rate": 5.295212164971746e-05,
+      "loss": 0.8808,
+      "step": 48125
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.293003089705464e-05,
+      "loss": 0.9481,
+      "step": 48130
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.290794309479113e-05,
+      "loss": 0.9324,
+      "step": 48135
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.288585824431139e-05,
+      "loss": 0.8866,
+      "step": 48140
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.286377634699967e-05,
+      "loss": 0.9753,
+      "step": 48145
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.284169740424014e-05,
+      "loss": 0.9452,
+      "step": 48150
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.2819621417416755e-05,
+      "loss": 1.1019,
+      "step": 48155
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.27975483879132e-05,
+      "loss": 0.8032,
+      "step": 48160
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.277547831711307e-05,
+      "loss": 0.9602,
+      "step": 48165
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.275341120639977e-05,
+      "loss": 1.0003,
+      "step": 48170
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.273134705715643e-05,
+      "loss": 1.1599,
+      "step": 48175
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.270928587076611e-05,
+      "loss": 0.9654,
+      "step": 48180
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.268722764861164e-05,
+      "loss": 0.8464,
+      "step": 48185
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.266517239207558e-05,
+      "loss": 0.8963,
+      "step": 48190
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.2643120102540466e-05,
+      "loss": 1.0255,
+      "step": 48195
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.80078125,
+      "learning_rate": 5.2621070781388485e-05,
+      "loss": 1.0463,
+      "step": 48200
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.259902443000177e-05,
+      "loss": 0.8804,
+      "step": 48205
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.257698104976217e-05,
+      "loss": 0.8568,
+      "step": 48210
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.25549406420514e-05,
+      "loss": 0.894,
+      "step": 48215
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.253290320825104e-05,
+      "loss": 0.8669,
+      "step": 48220
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.25108687497423e-05,
+      "loss": 0.8684,
+      "step": 48225
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.248883726790646e-05,
+      "loss": 0.9638,
+      "step": 48230
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.734375,
+      "learning_rate": 5.246680876412435e-05,
+      "loss": 0.898,
+      "step": 48235
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.51171875,
+      "learning_rate": 5.244478323977681e-05,
+      "loss": 0.954,
+      "step": 48240
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.242276069624445e-05,
+      "loss": 0.8637,
+      "step": 48245
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.240074113490765e-05,
+      "loss": 0.8344,
+      "step": 48250
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.2378724557146566e-05,
+      "loss": 0.9092,
+      "step": 48255
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.235671096434125e-05,
+      "loss": 0.9158,
+      "step": 48260
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.233470035787159e-05,
+      "loss": 0.8737,
+      "step": 48265
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.2312692739117165e-05,
+      "loss": 0.9542,
+      "step": 48270
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.229068810945748e-05,
+      "loss": 0.8575,
+      "step": 48275
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.498046875,
+      "learning_rate": 5.226868647027183e-05,
+      "loss": 0.931,
+      "step": 48280
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.224668782293923e-05,
+      "loss": 0.954,
+      "step": 48285
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.222469216883863e-05,
+      "loss": 0.8797,
+      "step": 48290
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.220269950934877e-05,
+      "loss": 1.0477,
+      "step": 48295
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.218070984584815e-05,
+      "loss": 0.912,
+      "step": 48300
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.2158723179715064e-05,
+      "loss": 1.0759,
+      "step": 48305
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.490234375,
+      "learning_rate": 5.213673951232768e-05,
+      "loss": 0.871,
+      "step": 48310
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.2114758845064026e-05,
+      "loss": 0.8752,
+      "step": 48315
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.2092781179301776e-05,
+      "loss": 0.9911,
+      "step": 48320
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.2070806516418565e-05,
+      "loss": 1.1045,
+      "step": 48325
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.204883485779184e-05,
+      "loss": 0.8972,
+      "step": 48330
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.2026866204798704e-05,
+      "loss": 0.9271,
+      "step": 48335
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.200490055881624e-05,
+      "loss": 0.9302,
+      "step": 48340
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.198293792122131e-05,
+      "loss": 0.9276,
+      "step": 48345
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.196097829339051e-05,
+      "loss": 0.93,
+      "step": 48350
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.193902167670026e-05,
+      "loss": 1.0635,
+      "step": 48355
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.191706807252692e-05,
+      "loss": 0.888,
+      "step": 48360
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.189511748224647e-05,
+      "loss": 0.9246,
+      "step": 48365
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.1873169907234843e-05,
+      "loss": 1.0039,
+      "step": 48370
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.625,
+      "learning_rate": 5.185122534886777e-05,
+      "loss": 1.0164,
+      "step": 48375
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.18292838085207e-05,
+      "loss": 0.8643,
+      "step": 48380
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.1807345287568984e-05,
+      "loss": 0.9587,
+      "step": 48385
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.178540978738779e-05,
+      "loss": 1.0061,
+      "step": 48390
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.176347730935198e-05,
+      "loss": 1.1532,
+      "step": 48395
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.625,
+      "learning_rate": 5.1741547854836344e-05,
+      "loss": 0.7755,
+      "step": 48400
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.17196214252155e-05,
+      "loss": 0.9659,
+      "step": 48405
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.169769802186377e-05,
+      "loss": 1.0091,
+      "step": 48410
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.16757776461553e-05,
+      "loss": 0.9977,
+      "step": 48415
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.165386029946412e-05,
+      "loss": 1.0268,
+      "step": 48420
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.163194598316408e-05,
+      "loss": 0.8544,
+      "step": 48425
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.1610034698628704e-05,
+      "loss": 0.9178,
+      "step": 48430
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.158812644723148e-05,
+      "loss": 1.0783,
+      "step": 48435
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.1566221230345655e-05,
+      "loss": 0.9574,
+      "step": 48440
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.15443190493442e-05,
+      "loss": 0.9738,
+      "step": 48445
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.152241990560003e-05,
+      "loss": 0.8551,
+      "step": 48450
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.150052380048587e-05,
+      "loss": 1.049,
+      "step": 48455
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.625,
+      "learning_rate": 5.147863073537402e-05,
+      "loss": 0.8248,
+      "step": 48460
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.1456740711636866e-05,
+      "loss": 1.0644,
+      "step": 48465
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.143485373064654e-05,
+      "loss": 0.8591,
+      "step": 48470
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.1412969793774835e-05,
+      "loss": 0.8676,
+      "step": 48475
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.139108890239353e-05,
+      "loss": 0.9285,
+      "step": 48480
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.136921105787417e-05,
+      "loss": 1.0226,
+      "step": 48485
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.134733626158801e-05,
+      "loss": 0.8493,
+      "step": 48490
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.1325464514906216e-05,
+      "loss": 1.1499,
+      "step": 48495
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.130359581919979e-05,
+      "loss": 0.8768,
+      "step": 48500
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.1281730175839394e-05,
+      "loss": 0.9771,
+      "step": 48505
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.125986758619567e-05,
+      "loss": 1.0365,
+      "step": 48510
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.123800805163893e-05,
+      "loss": 0.9222,
+      "step": 48515
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.1216151573539404e-05,
+      "loss": 1.0843,
+      "step": 48520
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.49609375,
+      "learning_rate": 5.119429815326704e-05,
+      "loss": 1.0634,
+      "step": 48525
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.1172447792191637e-05,
+      "loss": 0.9098,
+      "step": 48530
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.115060049168286e-05,
+      "loss": 0.9464,
+      "step": 48535
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.1128756253110044e-05,
+      "loss": 0.853,
+      "step": 48540
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.110691507784244e-05,
+      "loss": 0.95,
+      "step": 48545
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.108507696724911e-05,
+      "loss": 0.8101,
+      "step": 48550
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4609375,
+      "learning_rate": 5.106324192269884e-05,
+      "loss": 0.9499,
+      "step": 48555
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.104140994556035e-05,
+      "loss": 1.067,
+      "step": 48560
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.1019581037202036e-05,
+      "loss": 0.9056,
+      "step": 48565
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.099775519899213e-05,
+      "loss": 0.972,
+      "step": 48570
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.097593243229872e-05,
+      "loss": 0.8852,
+      "step": 48575
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.0954112738489754e-05,
+      "loss": 0.9366,
+      "step": 48580
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.093229611893281e-05,
+      "loss": 1.0244,
+      "step": 48585
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.0910482574995445e-05,
+      "loss": 0.9055,
+      "step": 48590
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.088867210804496e-05,
+      "loss": 0.828,
+      "step": 48595
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.0866864719448414e-05,
+      "loss": 1.0362,
+      "step": 48600
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.084506041057274e-05,
+      "loss": 1.0137,
+      "step": 48605
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.08232591827847e-05,
+      "loss": 0.9519,
+      "step": 48610
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.498046875,
+      "learning_rate": 5.0801461037450794e-05,
+      "loss": 0.8714,
+      "step": 48615
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.077966597593728e-05,
+      "loss": 0.8594,
+      "step": 48620
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.0757873999610385e-05,
+      "loss": 0.9234,
+      "step": 48625
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.073608510983606e-05,
+      "loss": 0.9935,
+      "step": 48630
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.0714299307979994e-05,
+      "loss": 1.1072,
+      "step": 48635
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.51171875,
+      "learning_rate": 5.069251659540777e-05,
+      "loss": 1.0978,
+      "step": 48640
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.06707369734848e-05,
+      "loss": 0.877,
+      "step": 48645
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.0648960443576185e-05,
+      "loss": 0.9152,
+      "step": 48650
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.062718700704693e-05,
+      "loss": 0.9402,
+      "step": 48655
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.0605416665261864e-05,
+      "loss": 0.9516,
+      "step": 48660
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.058364941958553e-05,
+      "loss": 0.85,
+      "step": 48665
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.0561885271382316e-05,
+      "loss": 0.9248,
+      "step": 48670
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.79296875,
+      "learning_rate": 5.054012422201646e-05,
+      "loss": 1.043,
+      "step": 48675
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.051836627285191e-05,
+      "loss": 0.8783,
+      "step": 48680
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.0496611425252515e-05,
+      "loss": 1.0747,
+      "step": 48685
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.047485968058194e-05,
+      "loss": 0.9795,
+      "step": 48690
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.045311104020353e-05,
+      "loss": 1.0219,
+      "step": 48695
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.0431365505480536e-05,
+      "loss": 0.924,
+      "step": 48700
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.040962307777606e-05,
+      "loss": 0.8713,
+      "step": 48705
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.0387883758452836e-05,
+      "loss": 1.0171,
+      "step": 48710
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.036614754887356e-05,
+      "loss": 0.9331,
+      "step": 48715
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.0344414450400734e-05,
+      "loss": 0.9752,
+      "step": 48720
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.032268446439656e-05,
+      "loss": 0.909,
+      "step": 48725
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.030095759222306e-05,
+      "loss": 0.9741,
+      "step": 48730
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.0279233835242135e-05,
+      "loss": 0.9015,
+      "step": 48735
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.025751319481551e-05,
+      "loss": 0.8833,
+      "step": 48740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.023579567230456e-05,
+      "loss": 0.8863,
+      "step": 48745
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.021408126907063e-05,
+      "loss": 1.0758,
+      "step": 48750
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.019236998647482e-05,
+      "loss": 0.9996,
+      "step": 48755
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.017066182587794e-05,
+      "loss": 0.905,
+      "step": 48760
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.0148956788640736e-05,
+      "loss": 1.0319,
+      "step": 48765
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.012725487612376e-05,
+      "loss": 0.8233,
+      "step": 48770
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.0105556089687186e-05,
+      "loss": 0.9083,
+      "step": 48775
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.0083860430691175e-05,
+      "loss": 1.1075,
+      "step": 48780
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.0062167900495674e-05,
+      "loss": 1.0288,
+      "step": 48785
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.49609375,
+      "learning_rate": 5.004047850046034e-05,
+      "loss": 0.9933,
+      "step": 48790
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.625,
+      "learning_rate": 5.00187922319447e-05,
+      "loss": 0.9871,
+      "step": 48795
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.999710909630813e-05,
+      "loss": 0.9935,
+      "step": 48800
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.997542909490968e-05,
+      "loss": 0.8804,
+      "step": 48805
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.9953752229108295e-05,
+      "loss": 0.9151,
+      "step": 48810
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.993207850026276e-05,
+      "loss": 0.93,
+      "step": 48815
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.991040790973153e-05,
+      "loss": 0.9725,
+      "step": 48820
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.9888740458873016e-05,
+      "loss": 0.8681,
+      "step": 48825
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.986707614904529e-05,
+      "loss": 0.8718,
+      "step": 48830
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.984541498160635e-05,
+      "loss": 0.957,
+      "step": 48835
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.982375695791389e-05,
+      "loss": 1.0334,
+      "step": 48840
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.98021020793255e-05,
+      "loss": 1.1653,
+      "step": 48845
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.9780450347198546e-05,
+      "loss": 0.9699,
+      "step": 48850
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.975880176289014e-05,
+      "loss": 0.9317,
+      "step": 48855
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.973715632775724e-05,
+      "loss": 1.0843,
+      "step": 48860
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.9715514043156675e-05,
+      "loss": 0.8991,
+      "step": 48865
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.9693874910444914e-05,
+      "loss": 0.9711,
+      "step": 48870
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.967223893097841e-05,
+      "loss": 0.8734,
+      "step": 48875
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.965060610611325e-05,
+      "loss": 0.8456,
+      "step": 48880
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.9628976437205485e-05,
+      "loss": 0.882,
+      "step": 48885
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.96073499256108e-05,
+      "loss": 0.9383,
+      "step": 48890
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.958572657268482e-05,
+      "loss": 1.0069,
+      "step": 48895
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.625,
+      "learning_rate": 4.956410637978295e-05,
+      "loss": 1.146,
+      "step": 48900
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.954248934826029e-05,
+      "loss": 0.8967,
+      "step": 48905
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.9520875479471904e-05,
+      "loss": 1.0125,
+      "step": 48910
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.94992647747725e-05,
+      "loss": 0.954,
+      "step": 48915
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.9477657235516696e-05,
+      "loss": 0.9786,
+      "step": 48920
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.945605286305891e-05,
+      "loss": 0.865,
+      "step": 48925
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.94344516587533e-05,
+      "loss": 0.8614,
+      "step": 48930
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.75,
+      "learning_rate": 4.941285362395382e-05,
+      "loss": 0.9378,
+      "step": 48935
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.939125876001427e-05,
+      "loss": 0.9931,
+      "step": 48940
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.9369667068288303e-05,
+      "loss": 1.0125,
+      "step": 48945
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.934807855012924e-05,
+      "loss": 0.9566,
+      "step": 48950
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.932649320689029e-05,
+      "loss": 1.149,
+      "step": 48955
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.93049110399245e-05,
+      "loss": 1.0048,
+      "step": 48960
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.928333205058459e-05,
+      "loss": 1.0096,
+      "step": 48965
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.9261756240223175e-05,
+      "loss": 0.9787,
+      "step": 48970
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.924018361019271e-05,
+      "loss": 0.9944,
+      "step": 48975
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.9218614161845345e-05,
+      "loss": 0.8065,
+      "step": 48980
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.919704789653304e-05,
+      "loss": 1.0229,
+      "step": 48985
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.917548481560764e-05,
+      "loss": 0.8176,
+      "step": 48990
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.915392492042078e-05,
+      "loss": 0.9873,
+      "step": 48995
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.913236821232377e-05,
+      "loss": 0.9549,
+      "step": 49000
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.9110814692667836e-05,
+      "loss": 1.1025,
+      "step": 49005
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5,
+      "learning_rate": 4.908926436280406e-05,
+      "loss": 0.9839,
+      "step": 49010
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.906771722408312e-05,
+      "loss": 0.9247,
+      "step": 49015
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.904617327785568e-05,
+      "loss": 0.9522,
+      "step": 49020
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.902463252547217e-05,
+      "loss": 0.9195,
+      "step": 49025
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.90030949682827e-05,
+      "loss": 0.9347,
+      "step": 49030
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.898156060763737e-05,
+      "loss": 0.9197,
+      "step": 49035
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.896002944488593e-05,
+      "loss": 0.8434,
+      "step": 49040
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.893850148137792e-05,
+      "loss": 0.8596,
+      "step": 49045
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.891697671846279e-05,
+      "loss": 0.9378,
+      "step": 49050
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.889545515748979e-05,
+      "loss": 0.9958,
+      "step": 49055
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.8873936799807816e-05,
+      "loss": 0.8579,
+      "step": 49060
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.885242164676571e-05,
+      "loss": 0.8905,
+      "step": 49065
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.883090969971211e-05,
+      "loss": 0.9437,
+      "step": 49070
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.8809400959995335e-05,
+      "loss": 0.949,
+      "step": 49075
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.87878954289636e-05,
+      "loss": 0.9098,
+      "step": 49080
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.876639310796495e-05,
+      "loss": 0.94,
+      "step": 49085
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.874489399834712e-05,
+      "loss": 0.9608,
+      "step": 49090
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.872339810145767e-05,
+      "loss": 1.0145,
+      "step": 49095
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.8701905418644044e-05,
+      "loss": 1.0344,
+      "step": 49100
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.8680415951253436e-05,
+      "loss": 0.9625,
+      "step": 49105
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.865892970063276e-05,
+      "loss": 0.8345,
+      "step": 49110
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.863744666812885e-05,
+      "loss": 1.032,
+      "step": 49115
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.861596685508832e-05,
+      "loss": 1.0246,
+      "step": 49120
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.8594490262857474e-05,
+      "loss": 1.0238,
+      "step": 49125
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.857301689278252e-05,
+      "loss": 1.1131,
+      "step": 49130
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.8551546746209474e-05,
+      "loss": 1.04,
+      "step": 49135
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.853007982448408e-05,
+      "loss": 0.9312,
+      "step": 49140
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.8508616128951844e-05,
+      "loss": 1.0589,
+      "step": 49145
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.848715566095825e-05,
+      "loss": 0.9474,
+      "step": 49150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.846569842184836e-05,
+      "loss": 1.0166,
+      "step": 49155
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.8444244412967186e-05,
+      "loss": 1.1753,
+      "step": 49160
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.8422793635659535e-05,
+      "loss": 0.9829,
+      "step": 49165
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.8401346091269887e-05,
+      "loss": 0.89,
+      "step": 49170
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.837990178114261e-05,
+      "loss": 0.9116,
+      "step": 49175
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.8358460706621934e-05,
+      "loss": 0.8536,
+      "step": 49180
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.833702286905172e-05,
+      "loss": 0.9304,
+      "step": 49185
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.831558826977578e-05,
+      "loss": 1.0011,
+      "step": 49190
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.82941569101376e-05,
+      "loss": 0.807,
+      "step": 49195
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.827272879148058e-05,
+      "loss": 0.8885,
+      "step": 49200
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.8251303915147805e-05,
+      "loss": 0.8502,
+      "step": 49205
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.822988228248224e-05,
+      "loss": 1.0452,
+      "step": 49210
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.8208463894826635e-05,
+      "loss": 0.9482,
+      "step": 49215
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.818704875352348e-05,
+      "loss": 0.8321,
+      "step": 49220
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.816563685991511e-05,
+      "loss": 1.0368,
+      "step": 49225
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.81442282153437e-05,
+      "loss": 0.9997,
+      "step": 49230
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.812282282115108e-05,
+      "loss": 0.9241,
+      "step": 49235
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.810142067867901e-05,
+      "loss": 1.0477,
+      "step": 49240
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.808002178926908e-05,
+      "loss": 0.9772,
+      "step": 49245
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.805862615426243e-05,
+      "loss": 0.9679,
+      "step": 49250
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.8037233775000254e-05,
+      "loss": 0.8722,
+      "step": 49255
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.8015844652823494e-05,
+      "loss": 1.0573,
+      "step": 49260
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.799445878907275e-05,
+      "loss": 0.9677,
+      "step": 49265
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.797307618508855e-05,
+      "loss": 0.939,
+      "step": 49270
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.7951696842211235e-05,
+      "loss": 0.9858,
+      "step": 49275
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.79303207617808e-05,
+      "loss": 1.0011,
+      "step": 49280
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.7908947945137164e-05,
+      "loss": 0.8575,
+      "step": 49285
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.788757839362004e-05,
+      "loss": 0.9424,
+      "step": 49290
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.786621210856884e-05,
+      "loss": 0.9988,
+      "step": 49295
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.784484909132281e-05,
+      "loss": 0.8938,
+      "step": 49300
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.782348934322104e-05,
+      "loss": 0.9112,
+      "step": 49305
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.7802132865602425e-05,
+      "loss": 0.9679,
+      "step": 49310
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.778077965980552e-05,
+      "loss": 0.8787,
+      "step": 49315
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.775942972716883e-05,
+      "loss": 1.1053,
+      "step": 49320
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.773808306903061e-05,
+      "loss": 0.9912,
+      "step": 49325
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.771673968672883e-05,
+      "loss": 1.0677,
+      "step": 49330
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.769539958160135e-05,
+      "loss": 0.8735,
+      "step": 49335
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.767406275498584e-05,
+      "loss": 1.1303,
+      "step": 49340
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.765272920821963e-05,
+      "loss": 0.9533,
+      "step": 49345
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.763139894264e-05,
+      "loss": 0.97,
+      "step": 49350
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.7610071959583935e-05,
+      "loss": 0.9927,
+      "step": 49355
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.758874826038819e-05,
+      "loss": 0.9571,
+      "step": 49360
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.75674278463894e-05,
+      "loss": 0.8435,
+      "step": 49365
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.7546110718923974e-05,
+      "loss": 1.0449,
+      "step": 49370
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.7524796879328035e-05,
+      "loss": 0.9614,
+      "step": 49375
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.750348632893759e-05,
+      "loss": 0.9131,
+      "step": 49380
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.748217906908846e-05,
+      "loss": 0.9521,
+      "step": 49385
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.7460875101116106e-05,
+      "loss": 0.9026,
+      "step": 49390
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.7439574426355946e-05,
+      "loss": 0.9808,
+      "step": 49395
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.741827704614317e-05,
+      "loss": 0.8594,
+      "step": 49400
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.739698296181265e-05,
+      "loss": 0.9514,
+      "step": 49405
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.737569217469913e-05,
+      "loss": 0.8326,
+      "step": 49410
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.625,
+      "learning_rate": 4.7354404686137155e-05,
+      "loss": 1.1199,
+      "step": 49415
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.733312049746108e-05,
+      "loss": 0.9097,
+      "step": 49420
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.731183961000496e-05,
+      "loss": 0.8891,
+      "step": 49425
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.729056202510275e-05,
+      "loss": 0.934,
+      "step": 49430
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.726928774408818e-05,
+      "loss": 0.8811,
+      "step": 49435
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.724801676829466e-05,
+      "loss": 1.2237,
+      "step": 49440
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.7226749099055537e-05,
+      "loss": 0.9113,
+      "step": 49445
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.720548473770392e-05,
+      "loss": 0.9965,
+      "step": 49450
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.718422368557265e-05,
+      "loss": 0.9313,
+      "step": 49455
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.716296594399434e-05,
+      "loss": 1.0277,
+      "step": 49460
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.714171151430156e-05,
+      "loss": 1.0236,
+      "step": 49465
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.712046039782645e-05,
+      "loss": 0.8902,
+      "step": 49470
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.709921259590111e-05,
+      "loss": 0.8605,
+      "step": 49475
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.707796810985742e-05,
+      "loss": 0.9146,
+      "step": 49480
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.705672694102691e-05,
+      "loss": 1.152,
+      "step": 49485
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.703548909074106e-05,
+      "loss": 0.9783,
+      "step": 49490
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.7014254560331115e-05,
+      "loss": 0.9032,
+      "step": 49495
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.699302335112799e-05,
+      "loss": 0.9469,
+      "step": 49500
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.6971795464462584e-05,
+      "loss": 0.9139,
+      "step": 49505
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.695057090166539e-05,
+      "loss": 0.852,
+      "step": 49510
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.692934966406687e-05,
+      "loss": 0.9828,
+      "step": 49515
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.690813175299712e-05,
+      "loss": 0.9554,
+      "step": 49520
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.625,
+      "learning_rate": 4.688691716978613e-05,
+      "loss": 0.9984,
+      "step": 49525
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.68657059157637e-05,
+      "loss": 0.9265,
+      "step": 49530
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.684449799225931e-05,
+      "loss": 0.9029,
+      "step": 49535
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.682329340060232e-05,
+      "loss": 0.9228,
+      "step": 49540
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.68020921421219e-05,
+      "loss": 0.9272,
+      "step": 49545
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.6780894218146886e-05,
+      "loss": 0.8161,
+      "step": 49550
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.6759699630006046e-05,
+      "loss": 0.9438,
+      "step": 49555
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.67385083790279e-05,
+      "loss": 0.8106,
+      "step": 49560
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.6717320466540715e-05,
+      "loss": 0.9336,
+      "step": 49565
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.6696135893872525e-05,
+      "loss": 0.9585,
+      "step": 49570
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.667495466235125e-05,
+      "loss": 0.9121,
+      "step": 49575
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.665377677330458e-05,
+      "loss": 1.0132,
+      "step": 49580
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.66326022280599e-05,
+      "loss": 1.0023,
+      "step": 49585
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.661143102794453e-05,
+      "loss": 0.9591,
+      "step": 49590
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.659026317428543e-05,
+      "loss": 0.8945,
+      "step": 49595
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.656909866840947e-05,
+      "loss": 0.9979,
+      "step": 49600
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.654793751164329e-05,
+      "loss": 0.7919,
+      "step": 49605
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.652677970531323e-05,
+      "loss": 1.1003,
+      "step": 49610
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.6505625250745557e-05,
+      "loss": 1.0737,
+      "step": 49615
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.6484474149266166e-05,
+      "loss": 0.842,
+      "step": 49620
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.6463326402200934e-05,
+      "loss": 1.0013,
+      "step": 49625
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.6442182010875345e-05,
+      "loss": 0.8792,
+      "step": 49630
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.642104097661477e-05,
+      "loss": 0.7644,
+      "step": 49635
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.639990330074442e-05,
+      "loss": 0.9765,
+      "step": 49640
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.637876898458913e-05,
+      "loss": 0.8962,
+      "step": 49645
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.635763802947367e-05,
+      "loss": 0.9404,
+      "step": 49650
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.633651043672259e-05,
+      "loss": 0.9556,
+      "step": 49655
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.631538620766012e-05,
+      "loss": 0.8018,
+      "step": 49660
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.629426534361042e-05,
+      "loss": 0.9562,
+      "step": 49665
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.627314784589729e-05,
+      "loss": 0.8734,
+      "step": 49670
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.625203371584449e-05,
+      "loss": 0.9724,
+      "step": 49675
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.6230922954775394e-05,
+      "loss": 0.9479,
+      "step": 49680
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.620981556401327e-05,
+      "loss": 0.995,
+      "step": 49685
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.6188711544881215e-05,
+      "loss": 0.8704,
+      "step": 49690
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.616761089870197e-05,
+      "loss": 0.8862,
+      "step": 49695
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.6146513626798186e-05,
+      "loss": 0.9441,
+      "step": 49700
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.612541973049229e-05,
+      "loss": 0.8817,
+      "step": 49705
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.6104329211106426e-05,
+      "loss": 0.9116,
+      "step": 49710
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.6083242069962564e-05,
+      "loss": 0.9826,
+      "step": 49715
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.6062158308382584e-05,
+      "loss": 0.9834,
+      "step": 49720
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.604107792768787e-05,
+      "loss": 0.8969,
+      "step": 49725
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.6020000929199856e-05,
+      "loss": 0.9264,
+      "step": 49730
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.5998927314239694e-05,
+      "loss": 0.9001,
+      "step": 49735
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.597785708412823e-05,
+      "loss": 1.0587,
+      "step": 49740
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.595679024018621e-05,
+      "loss": 0.914,
+      "step": 49745
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.593572678373417e-05,
+      "loss": 0.9823,
+      "step": 49750
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.59146667160923e-05,
+      "loss": 0.9742,
+      "step": 49755
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.589361003858072e-05,
+      "loss": 1.2027,
+      "step": 49760
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.5872556752519335e-05,
+      "loss": 1.0227,
+      "step": 49765
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.585150685922773e-05,
+      "loss": 0.9103,
+      "step": 49770
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.5830460360025315e-05,
+      "loss": 0.9096,
+      "step": 49775
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.580941725623134e-05,
+      "loss": 1.0385,
+      "step": 49780
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.578837754916483e-05,
+      "loss": 0.8463,
+      "step": 49785
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.576734124014454e-05,
+      "loss": 0.9174,
+      "step": 49790
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.574630833048905e-05,
+      "loss": 0.9285,
+      "step": 49795
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.572527882151678e-05,
+      "loss": 0.95,
+      "step": 49800
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.5704252714545824e-05,
+      "loss": 0.9859,
+      "step": 49805
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.5683230010894143e-05,
+      "loss": 1.0285,
+      "step": 49810
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.5662210711879495e-05,
+      "loss": 0.9972,
+      "step": 49815
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.564119481881933e-05,
+      "loss": 0.925,
+      "step": 49820
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.5620182333031025e-05,
+      "loss": 0.8579,
+      "step": 49825
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.559917325583162e-05,
+      "loss": 0.8751,
+      "step": 49830
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.557816758853796e-05,
+      "loss": 0.9685,
+      "step": 49835
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.5557165332466736e-05,
+      "loss": 0.9727,
+      "step": 49840
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.553616648893443e-05,
+      "loss": 1.0464,
+      "step": 49845
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.55151710592572e-05,
+      "loss": 0.9426,
+      "step": 49850
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.549417904475111e-05,
+      "loss": 0.8065,
+      "step": 49855
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.547319044673197e-05,
+      "loss": 0.9354,
+      "step": 49860
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.545220526651534e-05,
+      "loss": 0.825,
+      "step": 49865
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.54312235054166e-05,
+      "loss": 0.8915,
+      "step": 49870
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.5410245164750955e-05,
+      "loss": 0.8697,
+      "step": 49875
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.5389270245833316e-05,
+      "loss": 0.884,
+      "step": 49880
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.536829874997838e-05,
+      "loss": 0.9207,
+      "step": 49885
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.5347330678500686e-05,
+      "loss": 0.9551,
+      "step": 49890
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.5326366032714595e-05,
+      "loss": 0.9075,
+      "step": 49895
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.530540481393412e-05,
+      "loss": 0.8319,
+      "step": 49900
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.528444702347315e-05,
+      "loss": 0.8604,
+      "step": 49905
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.5263492662645404e-05,
+      "loss": 0.8437,
+      "step": 49910
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.524254173276423e-05,
+      "loss": 1.0757,
+      "step": 49915
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.522159423514292e-05,
+      "loss": 0.9885,
+      "step": 49920
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.5200650171094495e-05,
+      "loss": 1.059,
+      "step": 49925
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.517970954193172e-05,
+      "loss": 0.9149,
+      "step": 49930
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.515877234896717e-05,
+      "loss": 0.8068,
+      "step": 49935
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.5137838593513246e-05,
+      "loss": 0.9659,
+      "step": 49940
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.5116908276882064e-05,
+      "loss": 0.9508,
+      "step": 49945
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.509598140038557e-05,
+      "loss": 0.9591,
+      "step": 49950
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.507505796533552e-05,
+      "loss": 0.9213,
+      "step": 49955
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.505413797304336e-05,
+      "loss": 0.8213,
+      "step": 49960
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.5033221424820415e-05,
+      "loss": 1.0068,
+      "step": 49965
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.501230832197778e-05,
+      "loss": 0.8619,
+      "step": 49970
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.499139866582626e-05,
+      "loss": 1.0136,
+      "step": 49975
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.497049245767655e-05,
+      "loss": 1.099,
+      "step": 49980
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.4949589698839015e-05,
+      "loss": 0.9329,
+      "step": 49985
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.4928690390623916e-05,
+      "loss": 1.0729,
+      "step": 49990
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5,
+      "learning_rate": 4.490779453434121e-05,
+      "loss": 0.8685,
+      "step": 49995
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.488690213130067e-05,
+      "loss": 1.0066,
+      "step": 50000
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.486601318281192e-05,
+      "loss": 0.9442,
+      "step": 50005
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.484512769018422e-05,
+      "loss": 1.0699,
+      "step": 50010
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.482424565472673e-05,
+      "loss": 0.9667,
+      "step": 50015
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.480336707774839e-05,
+      "loss": 1.105,
+      "step": 50020
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.4782491960557835e-05,
+      "loss": 1.1043,
+      "step": 50025
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.476162030446357e-05,
+      "loss": 0.9254,
+      "step": 50030
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.474075211077393e-05,
+      "loss": 1.0297,
+      "step": 50035
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.4719887380796824e-05,
+      "loss": 0.9864,
+      "step": 50040
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.4699026115840115e-05,
+      "loss": 0.8713,
+      "step": 50045
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.4678168317211475e-05,
+      "loss": 1.0196,
+      "step": 50050
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.465731398621822e-05,
+      "loss": 0.8765,
+      "step": 50055
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.463646312416754e-05,
+      "loss": 0.9513,
+      "step": 50060
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.461561573236645e-05,
+      "loss": 0.9159,
+      "step": 50065
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.459477181212161e-05,
+      "loss": 0.9396,
+      "step": 50070
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.457393136473957e-05,
+      "loss": 0.9387,
+      "step": 50075
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.4553094391526684e-05,
+      "loss": 0.9666,
+      "step": 50080
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.453226089378898e-05,
+      "loss": 1.2035,
+      "step": 50085
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.4511430872832294e-05,
+      "loss": 0.9999,
+      "step": 50090
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.449060432996233e-05,
+      "loss": 1.0427,
+      "step": 50095
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.4469781266484534e-05,
+      "loss": 0.9715,
+      "step": 50100
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.444896168370406e-05,
+      "loss": 1.0078,
+      "step": 50105
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.442814558292593e-05,
+      "loss": 1.0199,
+      "step": 50110
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.4407332965454975e-05,
+      "loss": 0.9279,
+      "step": 50115
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.438652383259565e-05,
+      "loss": 0.8802,
+      "step": 50120
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.436571818565236e-05,
+      "loss": 1.0009,
+      "step": 50125
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.434491602592925e-05,
+      "loss": 0.9599,
+      "step": 50130
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.432411735473015e-05,
+      "loss": 0.9421,
+      "step": 50135
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.430332217335883e-05,
+      "loss": 0.9521,
+      "step": 50140
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.42825304831187e-05,
+      "loss": 0.9073,
+      "step": 50145
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.426174228531299e-05,
+      "loss": 1.0736,
+      "step": 50150
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.424095758124474e-05,
+      "loss": 0.9449,
+      "step": 50155
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.4220176372216826e-05,
+      "loss": 0.9873,
+      "step": 50160
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.419939865953173e-05,
+      "loss": 0.9748,
+      "step": 50165
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.417862444449189e-05,
+      "loss": 0.883,
+      "step": 50170
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.4157853728399475e-05,
+      "loss": 0.8066,
+      "step": 50175
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.4137086512556356e-05,
+      "loss": 0.8805,
+      "step": 50180
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.4116322798264276e-05,
+      "loss": 0.8867,
+      "step": 50185
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.4095562586824766e-05,
+      "loss": 1.1039,
+      "step": 50190
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.4074805879539064e-05,
+      "loss": 0.9521,
+      "step": 50195
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.405405267770818e-05,
+      "loss": 0.9552,
+      "step": 50200
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.4033302982633e-05,
+      "loss": 0.9926,
+      "step": 50205
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.401255679561416e-05,
+      "loss": 1.058,
+      "step": 50210
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.3991814117951994e-05,
+      "loss": 0.8447,
+      "step": 50215
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.39710749509467e-05,
+      "loss": 1.0724,
+      "step": 50220
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.395033929589828e-05,
+      "loss": 1.06,
+      "step": 50225
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.392960715410639e-05,
+      "loss": 1.0209,
+      "step": 50230
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.3908878526870575e-05,
+      "loss": 0.9987,
+      "step": 50235
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5,
+      "learning_rate": 4.388815341549019e-05,
+      "loss": 0.966,
+      "step": 50240
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5,
+      "learning_rate": 4.386743182126424e-05,
+      "loss": 0.9384,
+      "step": 50245
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.384671374549156e-05,
+      "loss": 0.9489,
+      "step": 50250
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.382599918947082e-05,
+      "loss": 0.9616,
+      "step": 50255
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.3805288154500456e-05,
+      "loss": 1.1693,
+      "step": 50260
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.3784580641878606e-05,
+      "loss": 1.0767,
+      "step": 50265
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.376387665290329e-05,
+      "loss": 0.9063,
+      "step": 50270
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.3743176188872205e-05,
+      "loss": 1.0128,
+      "step": 50275
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.37224792510829e-05,
+      "loss": 0.9175,
+      "step": 50280
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.370178584083274e-05,
+      "loss": 0.9061,
+      "step": 50285
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.368109595941872e-05,
+      "loss": 0.8362,
+      "step": 50290
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.366040960813778e-05,
+      "loss": 0.949,
+      "step": 50295
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5,
+      "learning_rate": 4.3639726788286495e-05,
+      "loss": 0.9306,
+      "step": 50300
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.361904750116136e-05,
+      "loss": 1.1314,
+      "step": 50305
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.359837174805851e-05,
+      "loss": 1.0325,
+      "step": 50310
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.357769953027394e-05,
+      "loss": 0.9517,
+      "step": 50315
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.3557030849103474e-05,
+      "loss": 0.8991,
+      "step": 50320
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.353636570584255e-05,
+      "loss": 0.9992,
+      "step": 50325
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.351570410178654e-05,
+      "loss": 0.8851,
+      "step": 50330
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.3495046038230556e-05,
+      "loss": 0.794,
+      "step": 50335
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.3474391516469416e-05,
+      "loss": 0.9003,
+      "step": 50340
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.345374053779778e-05,
+      "loss": 1.0992,
+      "step": 50345
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.343309310351013e-05,
+      "loss": 0.9804,
+      "step": 50350
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.341244921490063e-05,
+      "loss": 1.01,
+      "step": 50355
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.3391808873263215e-05,
+      "loss": 0.7897,
+      "step": 50360
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.33711720798917e-05,
+      "loss": 0.9257,
+      "step": 50365
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.335053883607965e-05,
+      "loss": 1.0943,
+      "step": 50370
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.332990914312031e-05,
+      "loss": 1.0418,
+      "step": 50375
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.33092830023068e-05,
+      "loss": 0.913,
+      "step": 50380
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.3288660414932034e-05,
+      "loss": 0.9002,
+      "step": 50385
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.3268041382288594e-05,
+      "loss": 0.7632,
+      "step": 50390
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.324742590566893e-05,
+      "loss": 1.0351,
+      "step": 50395
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.322681398636531e-05,
+      "loss": 0.8847,
+      "step": 50400
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.3206205625669574e-05,
+      "loss": 1.0687,
+      "step": 50405
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.3185600824873565e-05,
+      "loss": 1.0468,
+      "step": 50410
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.316499958526883e-05,
+      "loss": 1.1039,
+      "step": 50415
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.3144401908146614e-05,
+      "loss": 0.9662,
+      "step": 50420
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.3123807794798046e-05,
+      "loss": 0.8069,
+      "step": 50425
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.3103217246514015e-05,
+      "loss": 0.9225,
+      "step": 50430
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.308263026458509e-05,
+      "loss": 0.9417,
+      "step": 50435
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.3062046850301716e-05,
+      "loss": 0.9334,
+      "step": 50440
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.304146700495414e-05,
+      "loss": 0.8333,
+      "step": 50445
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.3020890729832244e-05,
+      "loss": 0.9604,
+      "step": 50450
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.300031802622585e-05,
+      "loss": 0.8044,
+      "step": 50455
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.2979748895424396e-05,
+      "loss": 0.9168,
+      "step": 50460
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.2959183338717255e-05,
+      "loss": 0.9734,
+      "step": 50465
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.2938621357393436e-05,
+      "loss": 0.9601,
+      "step": 50470
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.291806295274182e-05,
+      "loss": 1.0021,
+      "step": 50475
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.289750812605107e-05,
+      "loss": 0.9122,
+      "step": 50480
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.28769568786095e-05,
+      "loss": 0.8227,
+      "step": 50485
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.285640921170533e-05,
+      "loss": 0.9285,
+      "step": 50490
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.283586512662655e-05,
+      "loss": 0.8889,
+      "step": 50495
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.2815324624660814e-05,
+      "loss": 1.0344,
+      "step": 50500
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.279478770709569e-05,
+      "loss": 0.9646,
+      "step": 50505
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.277425437521843e-05,
+      "loss": 0.842,
+      "step": 50510
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.275372463031604e-05,
+      "loss": 0.8737,
+      "step": 50515
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.273319847367539e-05,
+      "loss": 0.8028,
+      "step": 50520
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.271267590658311e-05,
+      "loss": 0.9467,
+      "step": 50525
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.269215693032552e-05,
+      "loss": 1.0773,
+      "step": 50530
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.26716415461888e-05,
+      "loss": 1.038,
+      "step": 50535
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.2651129755458916e-05,
+      "loss": 1.0356,
+      "step": 50540
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.26306215594215e-05,
+      "loss": 1.1298,
+      "step": 50545
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.2610116959362057e-05,
+      "loss": 0.9046,
+      "step": 50550
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.2589615956565885e-05,
+      "loss": 0.8925,
+      "step": 50555
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.256911855231798e-05,
+      "loss": 0.97,
+      "step": 50560
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.2548624747903076e-05,
+      "loss": 1.0284,
+      "step": 50565
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.2528134544605813e-05,
+      "loss": 1.0616,
+      "step": 50570
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.250764794371057e-05,
+      "loss": 0.9394,
+      "step": 50575
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5,
+      "learning_rate": 4.24871649465014e-05,
+      "loss": 0.9177,
+      "step": 50580
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.246668555426223e-05,
+      "loss": 0.965,
+      "step": 50585
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.244620976827677e-05,
+      "loss": 0.9909,
+      "step": 50590
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.242573758982839e-05,
+      "loss": 1.0053,
+      "step": 50595
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.240526902020035e-05,
+      "loss": 0.9533,
+      "step": 50600
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.238480406067567e-05,
+      "loss": 0.9128,
+      "step": 50605
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.236434271253711e-05,
+      "loss": 1.0219,
+      "step": 50610
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.234388497706715e-05,
+      "loss": 0.8261,
+      "step": 50615
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.2323430855548174e-05,
+      "loss": 0.8961,
+      "step": 50620
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.23029803492622e-05,
+      "loss": 0.8444,
+      "step": 50625
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.228253345949114e-05,
+      "loss": 1.001,
+      "step": 50630
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.2262090187516644e-05,
+      "loss": 0.955,
+      "step": 50635
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.224165053462006e-05,
+      "loss": 0.8988,
+      "step": 50640
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.2221214502082594e-05,
+      "loss": 0.8975,
+      "step": 50645
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.220078209118525e-05,
+      "loss": 0.9368,
+      "step": 50650
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.218035330320869e-05,
+      "loss": 0.8767,
+      "step": 50655
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.2159928139433427e-05,
+      "loss": 1.0704,
+      "step": 50660
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.2139506601139766e-05,
+      "loss": 0.8408,
+      "step": 50665
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.211908868960774e-05,
+      "loss": 0.9176,
+      "step": 50670
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.209867440611712e-05,
+      "loss": 0.9723,
+      "step": 50675
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.2078263751947535e-05,
+      "loss": 0.9628,
+      "step": 50680
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.205785672837837e-05,
+      "loss": 1.221,
+      "step": 50685
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.2037453336688715e-05,
+      "loss": 0.9129,
+      "step": 50690
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.201705357815748e-05,
+      "loss": 1.0118,
+      "step": 50695
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.1996657454063415e-05,
+      "loss": 1.0532,
+      "step": 50700
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.197626496568488e-05,
+      "loss": 0.9336,
+      "step": 50705
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.195587611430014e-05,
+      "loss": 1.0174,
+      "step": 50710
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.193549090118727e-05,
+      "loss": 1.0614,
+      "step": 50715
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.191510932762388e-05,
+      "loss": 1.0202,
+      "step": 50720
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.189473139488759e-05,
+      "loss": 0.9998,
+      "step": 50725
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.187435710425576e-05,
+      "loss": 0.8776,
+      "step": 50730
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.1853986457005376e-05,
+      "loss": 0.9999,
+      "step": 50735
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.183361945441335e-05,
+      "loss": 0.8458,
+      "step": 50740
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.181325609775634e-05,
+      "loss": 0.7763,
+      "step": 50745
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.179289638831067e-05,
+      "loss": 1.0876,
+      "step": 50750
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.625,
+      "learning_rate": 4.177254032735254e-05,
+      "loss": 1.1139,
+      "step": 50755
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.1752187916157945e-05,
+      "loss": 1.0421,
+      "step": 50760
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.173183915600251e-05,
+      "loss": 0.9571,
+      "step": 50765
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.171149404816179e-05,
+      "loss": 0.9549,
+      "step": 50770
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.1691152593910975e-05,
+      "loss": 0.9044,
+      "step": 50775
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.167081479452516e-05,
+      "loss": 1.0115,
+      "step": 50780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.165048065127907e-05,
+      "loss": 1.009,
+      "step": 50785
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.16301501654473e-05,
+      "loss": 1.0415,
+      "step": 50790
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.1609823338304246e-05,
+      "loss": 0.9369,
+      "step": 50795
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.158950017112392e-05,
+      "loss": 0.8816,
+      "step": 50800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.156918066518025e-05,
+      "loss": 0.9797,
+      "step": 50805
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.154886482174691e-05,
+      "loss": 0.8669,
+      "step": 50810
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.152855264209727e-05,
+      "loss": 0.9375,
+      "step": 50815
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.150824412750458e-05,
+      "loss": 1.0531,
+      "step": 50820
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5,
+      "learning_rate": 4.148793927924176e-05,
+      "loss": 0.998,
+      "step": 50825
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.146763809858151e-05,
+      "loss": 0.856,
+      "step": 50830
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.144734058679636e-05,
+      "loss": 0.9849,
+      "step": 50835
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.142704674515863e-05,
+      "loss": 0.9253,
+      "step": 50840
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.140675657494029e-05,
+      "loss": 0.9308,
+      "step": 50845
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.1386470077413166e-05,
+      "loss": 0.9967,
+      "step": 50850
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.13661872538489e-05,
+      "loss": 0.9413,
+      "step": 50855
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.134590810551875e-05,
+      "loss": 0.8395,
+      "step": 50860
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.1325632633693886e-05,
+      "loss": 1.0183,
+      "step": 50865
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.130536083964524e-05,
+      "loss": 1.0262,
+      "step": 50870
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.12850927246434e-05,
+      "loss": 1.027,
+      "step": 50875
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.12648282899588e-05,
+      "loss": 1.0127,
+      "step": 50880
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.124456753686166e-05,
+      "loss": 0.9388,
+      "step": 50885
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.1224310466621965e-05,
+      "loss": 0.9285,
+      "step": 50890
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.120405708050941e-05,
+      "loss": 1.0692,
+      "step": 50895
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.11838073797935e-05,
+      "loss": 0.9283,
+      "step": 50900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.116356136574359e-05,
+      "loss": 0.8697,
+      "step": 50905
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.11433190396286e-05,
+      "loss": 1.0036,
+      "step": 50910
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.1123080402717415e-05,
+      "loss": 0.9792,
+      "step": 50915
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.110284545627865e-05,
+      "loss": 1.0078,
+      "step": 50920
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.1082614201580604e-05,
+      "loss": 1.0572,
+      "step": 50925
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.37890625,
+      "learning_rate": 4.106238663989137e-05,
+      "loss": 0.7276,
+      "step": 50930
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.10421627724789e-05,
+      "loss": 0.9774,
+      "step": 50935
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.102194260061078e-05,
+      "loss": 1.0234,
+      "step": 50940
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.100172612555446e-05,
+      "loss": 0.873,
+      "step": 50945
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.098151334857718e-05,
+      "loss": 1.0226,
+      "step": 50950
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.0961304270945824e-05,
+      "loss": 1.0075,
+      "step": 50955
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.094109889392715e-05,
+      "loss": 0.8681,
+      "step": 50960
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.0920897218787704e-05,
+      "loss": 1.0173,
+      "step": 50965
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.090069924679367e-05,
+      "loss": 1.1367,
+      "step": 50970
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.088050497921111e-05,
+      "loss": 0.8676,
+      "step": 50975
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.086031441730587e-05,
+      "loss": 1.0029,
+      "step": 50980
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.0840127562343476e-05,
+      "loss": 0.8638,
+      "step": 50985
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.081994441558923e-05,
+      "loss": 1.0133,
+      "step": 50990
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.0799764978308265e-05,
+      "loss": 0.9686,
+      "step": 50995
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.0779589251765495e-05,
+      "loss": 1.0507,
+      "step": 51000
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.075941723722547e-05,
+      "loss": 0.8381,
+      "step": 51005
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.0739248935952646e-05,
+      "loss": 0.9353,
+      "step": 51010
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.071908434921123e-05,
+      "loss": 0.8137,
+      "step": 51015
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.069892347826509e-05,
+      "loss": 0.9861,
+      "step": 51020
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.0678766324377957e-05,
+      "loss": 0.8659,
+      "step": 51025
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.065861288881335e-05,
+      "loss": 0.9603,
+      "step": 51030
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.0638463172834484e-05,
+      "loss": 0.9187,
+      "step": 51035
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.0618317177704304e-05,
+      "loss": 0.9147,
+      "step": 51040
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.059817490468564e-05,
+      "loss": 1.0506,
+      "step": 51045
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.057803635504105e-05,
+      "loss": 0.9171,
+      "step": 51050
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.0557901530032794e-05,
+      "loss": 1.1271,
+      "step": 51055
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.053777043092296e-05,
+      "loss": 0.9112,
+      "step": 51060
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.051764305897344e-05,
+      "loss": 0.9855,
+      "step": 51065
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.049751941544576e-05,
+      "loss": 0.8994,
+      "step": 51070
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.047739950160137e-05,
+      "loss": 0.9347,
+      "step": 51075
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.045728331870132e-05,
+      "loss": 0.9463,
+      "step": 51080
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.043717086800661e-05,
+      "loss": 0.8492,
+      "step": 51085
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.041706215077784e-05,
+      "loss": 1.1038,
+      "step": 51090
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.03969571682755e-05,
+      "loss": 0.8422,
+      "step": 51095
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.0376855921759725e-05,
+      "loss": 1.0364,
+      "step": 51100
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.0356758412490535e-05,
+      "loss": 0.9729,
+      "step": 51105
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.0336664641727697e-05,
+      "loss": 1.0945,
+      "step": 51110
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.031657461073063e-05,
+      "loss": 0.9299,
+      "step": 51115
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.029648832075863e-05,
+      "loss": 0.9332,
+      "step": 51120
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.027640577307078e-05,
+      "loss": 1.052,
+      "step": 51125
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.025632696892581e-05,
+      "loss": 1.0119,
+      "step": 51130
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.0236251909582325e-05,
+      "loss": 0.9903,
+      "step": 51135
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.021618059629861e-05,
+      "loss": 1.0225,
+      "step": 51140
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.0196113030332814e-05,
+      "loss": 1.0148,
+      "step": 51145
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.017604921294273e-05,
+      "loss": 0.9159,
+      "step": 51150
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.015598914538603e-05,
+      "loss": 0.9721,
+      "step": 51155
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.013593282892011e-05,
+      "loss": 0.9497,
+      "step": 51160
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.011588026480206e-05,
+      "loss": 0.9915,
+      "step": 51165
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.009583145428884e-05,
+      "loss": 0.7746,
+      "step": 51170
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.007578639863717e-05,
+      "loss": 0.8888,
+      "step": 51175
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.005574509910342e-05,
+      "loss": 0.9815,
+      "step": 51180
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.0035707556943834e-05,
+      "loss": 0.9475,
+      "step": 51185
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.0015673773414464e-05,
+      "loss": 0.8932,
+      "step": 51190
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.99956437497709e-05,
+      "loss": 0.8803,
+      "step": 51195
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.490234375,
+      "learning_rate": 3.9975617487268744e-05,
+      "loss": 0.7229,
+      "step": 51200
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.995559498716327e-05,
+      "loss": 0.9048,
+      "step": 51205
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.993557625070945e-05,
+      "loss": 0.8673,
+      "step": 51210
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.9915561279162125e-05,
+      "loss": 0.9174,
+      "step": 51215
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.989555007377588e-05,
+      "loss": 0.902,
+      "step": 51220
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4765625,
+      "learning_rate": 3.9875542635804976e-05,
+      "loss": 0.9872,
+      "step": 51225
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.985553896650354e-05,
+      "loss": 0.9109,
+      "step": 51230
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.983553906712544e-05,
+      "loss": 1.0102,
+      "step": 51235
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.494140625,
+      "learning_rate": 3.9815542938924286e-05,
+      "loss": 0.8146,
+      "step": 51240
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.9795550583153404e-05,
+      "loss": 0.9904,
+      "step": 51245
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.977556200106598e-05,
+      "loss": 0.9373,
+      "step": 51250
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.975557719391496e-05,
+      "loss": 0.9666,
+      "step": 51255
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.973559616295294e-05,
+      "loss": 1.1397,
+      "step": 51260
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.971561890943237e-05,
+      "loss": 1.0834,
+      "step": 51265
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.969564543460552e-05,
+      "loss": 0.9221,
+      "step": 51270
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.967567573972425e-05,
+      "loss": 1.0058,
+      "step": 51275
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.965570982604033e-05,
+      "loss": 0.9367,
+      "step": 51280
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.963574769480528e-05,
+      "loss": 1.0296,
+      "step": 51285
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.9615789347270285e-05,
+      "loss": 0.8648,
+      "step": 51290
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.9595834784686414e-05,
+      "loss": 1.0255,
+      "step": 51295
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.957588400830441e-05,
+      "loss": 0.9507,
+      "step": 51300
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.955593701937479e-05,
+      "loss": 1.0923,
+      "step": 51305
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.953599381914787e-05,
+      "loss": 1.0396,
+      "step": 51310
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.09375,
+      "learning_rate": 3.951605440887375e-05,
+      "loss": 0.9572,
+      "step": 51315
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.9496118789802196e-05,
+      "loss": 0.9524,
+      "step": 51320
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.947618696318282e-05,
+      "loss": 0.8093,
+      "step": 51325
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.945625893026502e-05,
+      "loss": 0.862,
+      "step": 51330
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.943633469229783e-05,
+      "loss": 0.8987,
+      "step": 51335
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.941641425053014e-05,
+      "loss": 1.0828,
+      "step": 51340
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.939649760621066e-05,
+      "loss": 0.9208,
+      "step": 51345
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.937658476058772e-05,
+      "loss": 0.8918,
+      "step": 51350
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.9356675714909455e-05,
+      "loss": 1.0362,
+      "step": 51355
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.933677047042382e-05,
+      "loss": 0.9233,
+      "step": 51360
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.931686902837854e-05,
+      "loss": 1.0053,
+      "step": 51365
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.9296971390021e-05,
+      "loss": 0.9761,
+      "step": 51370
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.9277077556598415e-05,
+      "loss": 0.8039,
+      "step": 51375
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.9257187529357806e-05,
+      "loss": 0.9285,
+      "step": 51380
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.9237301309545826e-05,
+      "loss": 0.8729,
+      "step": 51385
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.9217418898408996e-05,
+      "loss": 0.8877,
+      "step": 51390
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.458984375,
+      "learning_rate": 3.919754029719363e-05,
+      "loss": 0.7871,
+      "step": 51395
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.917766550714567e-05,
+      "loss": 0.8088,
+      "step": 51400
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.915779452951087e-05,
+      "loss": 0.9516,
+      "step": 51405
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.498046875,
+      "learning_rate": 3.913792736553484e-05,
+      "loss": 0.8929,
+      "step": 51410
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.9118064016462806e-05,
+      "loss": 0.8689,
+      "step": 51415
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.909820448353986e-05,
+      "loss": 0.8161,
+      "step": 51420
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.907834876801085e-05,
+      "loss": 0.9621,
+      "step": 51425
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.9058496871120295e-05,
+      "loss": 0.9807,
+      "step": 51430
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.903864879411255e-05,
+      "loss": 0.9334,
+      "step": 51435
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.9018804538231776e-05,
+      "loss": 0.9602,
+      "step": 51440
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.8998964104721745e-05,
+      "loss": 0.8113,
+      "step": 51445
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.897912749482615e-05,
+      "loss": 1.0266,
+      "step": 51450
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.895929470978831e-05,
+      "loss": 1.0061,
+      "step": 51455
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.8939465750851434e-05,
+      "loss": 0.9953,
+      "step": 51460
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.891964061925835e-05,
+      "loss": 1.0519,
+      "step": 51465
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.8899819316251753e-05,
+      "loss": 0.7409,
+      "step": 51470
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.888000184307411e-05,
+      "loss": 0.9194,
+      "step": 51475
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.8860188200967516e-05,
+      "loss": 1.035,
+      "step": 51480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.884037839117396e-05,
+      "loss": 1.0426,
+      "step": 51485
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.8820572414935185e-05,
+      "loss": 1.03,
+      "step": 51490
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.880077027349257e-05,
+      "loss": 0.9797,
+      "step": 51495
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.878097196808737e-05,
+      "loss": 0.8733,
+      "step": 51500
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.876117749996064e-05,
+      "loss": 0.911,
+      "step": 51505
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.8741386870352994e-05,
+      "loss": 1.0591,
+      "step": 51510
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.872160008050497e-05,
+      "loss": 0.8526,
+      "step": 51515
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.870181713165688e-05,
+      "loss": 0.9296,
+      "step": 51520
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.868203802504867e-05,
+      "loss": 0.8969,
+      "step": 51525
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.866226276192016e-05,
+      "loss": 0.9046,
+      "step": 51530
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.864249134351091e-05,
+      "loss": 0.8912,
+      "step": 51535
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.8622723771060145e-05,
+      "loss": 1.1907,
+      "step": 51540
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.860296004580696e-05,
+      "loss": 0.9601,
+      "step": 51545
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.8583200168990195e-05,
+      "loss": 0.9729,
+      "step": 51550
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.856344414184839e-05,
+      "loss": 0.9276,
+      "step": 51555
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.854369196561984e-05,
+      "loss": 1.0324,
+      "step": 51560
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.852394364154268e-05,
+      "loss": 0.8486,
+      "step": 51565
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.850419917085478e-05,
+      "loss": 1.0582,
+      "step": 51570
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.848445855479368e-05,
+      "loss": 0.9645,
+      "step": 51575
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.846472179459678e-05,
+      "loss": 0.8136,
+      "step": 51580
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.8444988891501224e-05,
+      "loss": 1.1109,
+      "step": 51585
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.8425259846743845e-05,
+      "loss": 0.9099,
+      "step": 51590
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.8405534661561304e-05,
+      "loss": 1.062,
+      "step": 51595
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.838581333719004e-05,
+      "loss": 0.9472,
+      "step": 51600
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.836609587486612e-05,
+      "loss": 0.9514,
+      "step": 51605
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.490234375,
+      "learning_rate": 3.834638227582555e-05,
+      "loss": 1.0129,
+      "step": 51610
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.832667254130396e-05,
+      "loss": 0.9279,
+      "step": 51615
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.830696667253674e-05,
+      "loss": 0.9951,
+      "step": 51620
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.8287264670759106e-05,
+      "loss": 0.9776,
+      "step": 51625
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.826756653720605e-05,
+      "loss": 0.9665,
+      "step": 51630
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.824787227311218e-05,
+      "loss": 0.9334,
+      "step": 51635
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.822818187971201e-05,
+      "loss": 1.0076,
+      "step": 51640
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.8208495358239796e-05,
+      "loss": 0.8727,
+      "step": 51645
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.818881270992943e-05,
+      "loss": 0.9704,
+      "step": 51650
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.816913393601468e-05,
+      "loss": 0.9842,
+      "step": 51655
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.8149459037729076e-05,
+      "loss": 0.9255,
+      "step": 51660
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.8129788016305814e-05,
+      "loss": 1.0225,
+      "step": 51665
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.811012087297786e-05,
+      "loss": 0.9501,
+      "step": 51670
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.809045760897802e-05,
+      "loss": 1.0628,
+      "step": 51675
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.807079822553885e-05,
+      "loss": 1.0562,
+      "step": 51680
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.805114272389254e-05,
+      "loss": 0.9083,
+      "step": 51685
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.803149110527116e-05,
+      "loss": 0.9235,
+      "step": 51690
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.801184337090653e-05,
+      "loss": 1.0131,
+      "step": 51695
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.7992199522030115e-05,
+      "loss": 1.0319,
+      "step": 51700
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.797255955987326e-05,
+      "loss": 1.0243,
+      "step": 51705
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.7952923485667045e-05,
+      "loss": 0.885,
+      "step": 51710
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.793329130064225e-05,
+      "loss": 0.9075,
+      "step": 51715
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.791366300602941e-05,
+      "loss": 1.0565,
+      "step": 51720
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.789403860305889e-05,
+      "loss": 0.8994,
+      "step": 51725
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.7874418092960796e-05,
+      "loss": 0.9463,
+      "step": 51730
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.7854801476964895e-05,
+      "loss": 0.9532,
+      "step": 51735
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.47265625,
+      "learning_rate": 3.783518875630081e-05,
+      "loss": 0.8684,
+      "step": 51740
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.781557993219794e-05,
+      "loss": 0.904,
+      "step": 51745
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.77959750058853e-05,
+      "loss": 0.8483,
+      "step": 51750
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.777637397859183e-05,
+      "loss": 1.0518,
+      "step": 51755
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.775677685154606e-05,
+      "loss": 0.8509,
+      "step": 51760
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.7737183625976446e-05,
+      "loss": 0.9962,
+      "step": 51765
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.466796875,
+      "learning_rate": 3.771759430311105e-05,
+      "loss": 0.7977,
+      "step": 51770
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.7698008884177794e-05,
+      "loss": 0.9294,
+      "step": 51775
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.767842737040427e-05,
+      "loss": 1.0377,
+      "step": 51780
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.76588497630179e-05,
+      "loss": 1.0127,
+      "step": 51785
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.7639276063245855e-05,
+      "loss": 1.0065,
+      "step": 51790
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.761970627231498e-05,
+      "loss": 0.8589,
+      "step": 51795
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.760014039145197e-05,
+      "loss": 0.8028,
+      "step": 51800
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.758057842188325e-05,
+      "loss": 0.8983,
+      "step": 51805
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.756102036483493e-05,
+      "loss": 0.9926,
+      "step": 51810
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.754146622153296e-05,
+      "loss": 0.9784,
+      "step": 51815
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.752191599320307e-05,
+      "loss": 0.9877,
+      "step": 51820
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.7502369681070635e-05,
+      "loss": 1.0543,
+      "step": 51825
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.748282728636081e-05,
+      "loss": 0.9123,
+      "step": 51830
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.746328881029858e-05,
+      "loss": 0.8175,
+      "step": 51835
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.744375425410867e-05,
+      "loss": 0.92,
+      "step": 51840
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.742422361901544e-05,
+      "loss": 0.9807,
+      "step": 51845
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.740469690624314e-05,
+      "loss": 0.9612,
+      "step": 51850
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.738517411701578e-05,
+      "loss": 0.9129,
+      "step": 51855
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.7365655252556965e-05,
+      "loss": 0.8131,
+      "step": 51860
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.7346140314090226e-05,
+      "loss": 0.9964,
+      "step": 51865
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.732662930283883e-05,
+      "loss": 1.0277,
+      "step": 51870
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.7307122220025625e-05,
+      "loss": 0.9241,
+      "step": 51875
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.728761906687339e-05,
+      "loss": 0.9222,
+      "step": 51880
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.726811984460467e-05,
+      "loss": 1.0535,
+      "step": 51885
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.724862455444159e-05,
+      "loss": 1.073,
+      "step": 51890
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.72291331976062e-05,
+      "loss": 0.9135,
+      "step": 51895
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.7209645775320265e-05,
+      "loss": 0.959,
+      "step": 51900
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.7190162288805205e-05,
+      "loss": 0.9446,
+      "step": 51905
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.717068273928232e-05,
+      "loss": 1.0108,
+      "step": 51910
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.498046875,
+      "learning_rate": 3.715120712797262e-05,
+      "loss": 0.8893,
+      "step": 51915
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.713173545609681e-05,
+      "loss": 1.1068,
+      "step": 51920
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.711226772487545e-05,
+      "loss": 0.7963,
+      "step": 51925
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.7092803935528734e-05,
+      "loss": 0.9644,
+      "step": 51930
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.7073344089276754e-05,
+      "loss": 0.9825,
+      "step": 51935
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.705388818733919e-05,
+      "loss": 0.9845,
+      "step": 51940
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.703443623093562e-05,
+      "loss": 0.9744,
+      "step": 51945
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.7014988221285315e-05,
+      "loss": 0.8523,
+      "step": 51950
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.699554415960727e-05,
+      "loss": 0.8706,
+      "step": 51955
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.6976104047120264e-05,
+      "loss": 1.1154,
+      "step": 51960
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.695666788504286e-05,
+      "loss": 0.9287,
+      "step": 51965
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.693723567459329e-05,
+      "loss": 0.8874,
+      "step": 51970
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.48046875,
+      "learning_rate": 3.691780741698964e-05,
+      "loss": 0.9014,
+      "step": 51975
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.689838311344966e-05,
+      "loss": 1.0094,
+      "step": 51980
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.687896276519086e-05,
+      "loss": 1.0438,
+      "step": 51985
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.6859546373430576e-05,
+      "loss": 0.8317,
+      "step": 51990
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.6840133939385854e-05,
+      "loss": 0.9784,
+      "step": 51995
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.682072546427344e-05,
+      "loss": 1.0079,
+      "step": 52000
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.680132094930992e-05,
+      "loss": 0.9256,
+      "step": 52005
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.494140625,
+      "learning_rate": 3.678192039571161e-05,
+      "loss": 0.8137,
+      "step": 52010
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.676252380469448e-05,
+      "loss": 0.9156,
+      "step": 52015
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.67431311774744e-05,
+      "loss": 0.9449,
+      "step": 52020
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.6723742515266924e-05,
+      "loss": 1.0522,
+      "step": 52025
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.6704357819287336e-05,
+      "loss": 0.887,
+      "step": 52030
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.668497709075065e-05,
+      "loss": 0.879,
+      "step": 52035
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.666560033087172e-05,
+      "loss": 1.0117,
+      "step": 52040
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.6646227540865105e-05,
+      "loss": 0.9591,
+      "step": 52045
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.662685872194509e-05,
+      "loss": 0.9551,
+      "step": 52050
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.660749387532574e-05,
+      "loss": 0.8861,
+      "step": 52055
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.658813300222091e-05,
+      "loss": 1.0416,
+      "step": 52060
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.656877610384407e-05,
+      "loss": 0.9706,
+      "step": 52065
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.65494231814086e-05,
+      "loss": 1.0754,
+      "step": 52070
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.6530074236127585e-05,
+      "loss": 0.869,
+      "step": 52075
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.6510729269213805e-05,
+      "loss": 0.9369,
+      "step": 52080
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.649138828187978e-05,
+      "loss": 1.013,
+      "step": 52085
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.6472051275337904e-05,
+      "loss": 1.021,
+      "step": 52090
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.6452718250800174e-05,
+      "loss": 0.9296,
+      "step": 52095
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.68359375,
+      "learning_rate": 3.643338920947844e-05,
+      "loss": 1.0464,
+      "step": 52100
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.64140641525843e-05,
+      "loss": 0.9818,
+      "step": 52105
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.639474308132901e-05,
+      "loss": 1.1438,
+      "step": 52110
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.637542599692365e-05,
+      "loss": 0.9877,
+      "step": 52115
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.63561129005791e-05,
+      "loss": 0.9625,
+      "step": 52120
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5,
+      "learning_rate": 3.633680379350583e-05,
+      "loss": 0.8335,
+      "step": 52125
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.631749867691421e-05,
+      "loss": 1.027,
+      "step": 52130
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.6298197552014336e-05,
+      "loss": 0.8022,
+      "step": 52135
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.627890042001598e-05,
+      "loss": 1.0426,
+      "step": 52140
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.625960728212869e-05,
+      "loss": 1.2428,
+      "step": 52145
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.6240318139561826e-05,
+      "loss": 0.882,
+      "step": 52150
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.622103299352445e-05,
+      "loss": 0.9101,
+      "step": 52155
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.620175184522534e-05,
+      "loss": 0.9604,
+      "step": 52160
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.6182474695873084e-05,
+      "loss": 1.1251,
+      "step": 52165
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.616320154667603e-05,
+      "loss": 1.1223,
+      "step": 52170
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.614393239884216e-05,
+      "loss": 0.8701,
+      "step": 52175
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.612466725357935e-05,
+      "loss": 0.8804,
+      "step": 52180
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.6105406112095207e-05,
+      "loss": 1.0503,
+      "step": 52185
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.6086148975596914e-05,
+      "loss": 0.9704,
+      "step": 52190
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.6066895845291595e-05,
+      "loss": 0.9705,
+      "step": 52195
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.604764672238609e-05,
+      "loss": 0.9165,
+      "step": 52200
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.602840160808688e-05,
+      "loss": 0.9666,
+      "step": 52205
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.6009160503600326e-05,
+      "loss": 0.9914,
+      "step": 52210
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5,
+      "learning_rate": 3.5989923410132495e-05,
+      "loss": 0.9454,
+      "step": 52215
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.458984375,
+      "learning_rate": 3.597069032888915e-05,
+      "loss": 0.9694,
+      "step": 52220
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.5951461261075845e-05,
+      "loss": 0.9661,
+      "step": 52225
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.593223620789793e-05,
+      "loss": 1.0651,
+      "step": 52230
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.5913015170560385e-05,
+      "loss": 0.933,
+      "step": 52235
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.589379815026806e-05,
+      "loss": 0.9405,
+      "step": 52240
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.5874585148225456e-05,
+      "loss": 0.9426,
+      "step": 52245
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.5855376165636924e-05,
+      "loss": 0.9903,
+      "step": 52250
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.5836171203706425e-05,
+      "loss": 1.2026,
+      "step": 52255
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.49609375,
+      "learning_rate": 3.5816970263637796e-05,
+      "loss": 0.965,
+      "step": 52260
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.579777334663461e-05,
+      "loss": 0.9093,
+      "step": 52265
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.577858045390007e-05,
+      "loss": 1.1153,
+      "step": 52270
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.575939158663725e-05,
+      "loss": 0.9002,
+      "step": 52275
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.5740206746048965e-05,
+      "loss": 0.9292,
+      "step": 52280
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.572102593333767e-05,
+      "loss": 0.8759,
+      "step": 52285
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.570184914970571e-05,
+      "loss": 0.946,
+      "step": 52290
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.568267639635507e-05,
+      "loss": 0.9866,
+      "step": 52295
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.5663507674487505e-05,
+      "loss": 0.892,
+      "step": 52300
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.48046875,
+      "learning_rate": 3.5644342985304545e-05,
+      "loss": 0.8986,
+      "step": 52305
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.562518233000749e-05,
+      "loss": 0.9364,
+      "step": 52310
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4609375,
+      "learning_rate": 3.56060257097973e-05,
+      "loss": 0.8437,
+      "step": 52315
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.558687312587474e-05,
+      "loss": 0.9295,
+      "step": 52320
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.5567724579440386e-05,
+      "loss": 1.0193,
+      "step": 52325
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.554858007169439e-05,
+      "loss": 1.019,
+      "step": 52330
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.552943960383681e-05,
+      "loss": 0.9998,
+      "step": 52335
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.551030317706742e-05,
+      "loss": 1.1862,
+      "step": 52340
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.549117079258566e-05,
+      "loss": 1.1737,
+      "step": 52345
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.547204245159075e-05,
+      "loss": 0.9703,
+      "step": 52350
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.545291815528171e-05,
+      "loss": 0.9504,
+      "step": 52355
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.543379790485732e-05,
+      "loss": 0.8241,
+      "step": 52360
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.541468170151597e-05,
+      "loss": 0.9014,
+      "step": 52365
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.539556954645593e-05,
+      "loss": 0.9119,
+      "step": 52370
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.53764614408752e-05,
+      "loss": 0.9338,
+      "step": 52375
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.535735738597144e-05,
+      "loss": 0.9221,
+      "step": 52380
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.49609375,
+      "learning_rate": 3.533825738294213e-05,
+      "loss": 0.888,
+      "step": 52385
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.5319161432984525e-05,
+      "loss": 1.0516,
+      "step": 52390
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.5300069537295556e-05,
+      "loss": 0.8644,
+      "step": 52395
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.528098169707187e-05,
+      "loss": 0.8862,
+      "step": 52400
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.5261897913509964e-05,
+      "loss": 1.0723,
+      "step": 52405
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.524281818780607e-05,
+      "loss": 0.9493,
+      "step": 52410
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.522374252115604e-05,
+      "loss": 0.9562,
+      "step": 52415
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.520467091475561e-05,
+      "loss": 0.9403,
+      "step": 52420
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.518560336980024e-05,
+      "loss": 0.9999,
+      "step": 52425
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.516653988748503e-05,
+      "loss": 0.866,
+      "step": 52430
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.514748046900497e-05,
+      "loss": 0.9026,
+      "step": 52435
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.5128425115554656e-05,
+      "loss": 1.0292,
+      "step": 52440
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5,
+      "learning_rate": 3.510937382832854e-05,
+      "loss": 0.9959,
+      "step": 52445
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.509032660852082e-05,
+      "loss": 1.0212,
+      "step": 52450
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.5071283457325344e-05,
+      "loss": 0.8349,
+      "step": 52455
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.5052244375935736e-05,
+      "loss": 0.9994,
+      "step": 52460
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.503320936554543e-05,
+      "loss": 0.8113,
+      "step": 52465
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.501417842734758e-05,
+      "loss": 0.97,
+      "step": 52470
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.625,
+      "learning_rate": 3.4995151562535e-05,
+      "loss": 0.9237,
+      "step": 52475
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.4976128772300364e-05,
+      "loss": 1.1457,
+      "step": 52480
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.4957110057836065e-05,
+      "loss": 1.0911,
+      "step": 52485
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.494140625,
+      "learning_rate": 3.493809542033414e-05,
+      "loss": 0.8484,
+      "step": 52490
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.4919084860986506e-05,
+      "loss": 1.061,
+      "step": 52495
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.490007838098478e-05,
+      "loss": 0.9506,
+      "step": 52500
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.4881075981520284e-05,
+      "loss": 0.8539,
+      "step": 52505
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.625,
+      "learning_rate": 3.4862077663784074e-05,
+      "loss": 0.962,
+      "step": 52510
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.494140625,
+      "learning_rate": 3.484308342896703e-05,
+      "loss": 0.8708,
+      "step": 52515
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.482421875,
+      "learning_rate": 3.482409327825975e-05,
+      "loss": 0.8459,
+      "step": 52520
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.4805107212852504e-05,
+      "loss": 0.9101,
+      "step": 52525
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.4786125233935386e-05,
+      "loss": 0.9337,
+      "step": 52530
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.4767147342698244e-05,
+      "loss": 0.8437,
+      "step": 52535
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.474817354033058e-05,
+      "loss": 1.005,
+      "step": 52540
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.4729203828021694e-05,
+      "loss": 0.9563,
+      "step": 52545
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.47102382069607e-05,
+      "loss": 0.9094,
+      "step": 52550
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.469127667833631e-05,
+      "loss": 0.8738,
+      "step": 52555
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.467231924333707e-05,
+      "loss": 0.8692,
+      "step": 52560
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.465336590315128e-05,
+      "loss": 0.9131,
+      "step": 52565
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.463441665896692e-05,
+      "loss": 0.9032,
+      "step": 52570
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.461547151197175e-05,
+      "loss": 1.0797,
+      "step": 52575
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.4596530463353336e-05,
+      "loss": 0.9467,
+      "step": 52580
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.457759351429884e-05,
+      "loss": 0.8499,
+      "step": 52585
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.453125,
+      "learning_rate": 3.455866066599531e-05,
+      "loss": 0.8842,
+      "step": 52590
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.453973191962948e-05,
+      "loss": 0.7765,
+      "step": 52595
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.452080727638778e-05,
+      "loss": 1.0101,
+      "step": 52600
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.450188673745648e-05,
+      "loss": 0.9075,
+      "step": 52605
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.448297030402149e-05,
+      "loss": 0.9609,
+      "step": 52610
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.446405797726857e-05,
+      "loss": 1.009,
+      "step": 52615
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.4445149758383096e-05,
+      "loss": 1.0084,
+      "step": 52620
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.44262456485503e-05,
+      "loss": 0.8443,
+      "step": 52625
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.440734564895515e-05,
+      "loss": 0.7951,
+      "step": 52630
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.438844976078224e-05,
+      "loss": 0.9905,
+      "step": 52635
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.436955798521602e-05,
+      "loss": 0.995,
+      "step": 52640
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.4350670323440684e-05,
+      "loss": 1.0952,
+      "step": 52645
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.4331786776640075e-05,
+      "loss": 0.8872,
+      "step": 52650
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.431290734599785e-05,
+      "loss": 0.9607,
+      "step": 52655
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.429403203269748e-05,
+      "loss": 1.0511,
+      "step": 52660
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.427516083792194e-05,
+      "loss": 0.9045,
+      "step": 52665
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.425629376285418e-05,
+      "loss": 1.0049,
+      "step": 52670
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.423743080867684e-05,
+      "loss": 1.0459,
+      "step": 52675
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.421857197657219e-05,
+      "loss": 0.9,
+      "step": 52680
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.419971726772238e-05,
+      "loss": 1.0364,
+      "step": 52685
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.4180866683309255e-05,
+      "loss": 1.015,
+      "step": 52690
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.416202022451433e-05,
+      "loss": 0.9177,
+      "step": 52695
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.4143177892518975e-05,
+      "loss": 1.0352,
+      "step": 52700
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.412433968850426e-05,
+      "loss": 1.0523,
+      "step": 52705
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.4105505613650956e-05,
+      "loss": 0.886,
+      "step": 52710
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.408667566913958e-05,
+      "loss": 0.9176,
+      "step": 52715
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.406784985615044e-05,
+      "loss": 0.9577,
+      "step": 52720
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.40490281758636e-05,
+      "loss": 1.008,
+      "step": 52725
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.403021062945875e-05,
+      "loss": 0.8865,
+      "step": 52730
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.4011397218115425e-05,
+      "loss": 0.9519,
+      "step": 52735
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.399258794301291e-05,
+      "loss": 1.0096,
+      "step": 52740
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.3973782805330135e-05,
+      "loss": 1.1155,
+      "step": 52745
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.395498180624584e-05,
+      "loss": 0.9467,
+      "step": 52750
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.3936184946938544e-05,
+      "loss": 0.8723,
+      "step": 52755
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.391739222858639e-05,
+      "loss": 0.928,
+      "step": 52760
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.3898603652367364e-05,
+      "loss": 0.8712,
+      "step": 52765
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.387981921945916e-05,
+      "loss": 0.9256,
+      "step": 52770
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.386103893103916e-05,
+      "loss": 1.0196,
+      "step": 52775
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.384226278828456e-05,
+      "loss": 0.8886,
+      "step": 52780
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.382349079237232e-05,
+      "loss": 1.0258,
+      "step": 52785
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.3804722944479004e-05,
+      "loss": 0.8668,
+      "step": 52790
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.378595924578104e-05,
+      "loss": 0.8021,
+      "step": 52795
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.37671996974546e-05,
+      "loss": 0.9738,
+      "step": 52800
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.3748444300675484e-05,
+      "loss": 0.8922,
+      "step": 52805
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.372969305661934e-05,
+      "loss": 0.9705,
+      "step": 52810
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.371094596646153e-05,
+      "loss": 0.9663,
+      "step": 52815
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.369220303137712e-05,
+      "loss": 1.0698,
+      "step": 52820
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.367346425254093e-05,
+      "loss": 0.8487,
+      "step": 52825
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.365472963112752e-05,
+      "loss": 0.9631,
+      "step": 52830
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.363599916831126e-05,
+      "loss": 0.9238,
+      "step": 52835
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5,
+      "learning_rate": 3.361727286526612e-05,
+      "loss": 0.9257,
+      "step": 52840
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.359855072316592e-05,
+      "loss": 0.9111,
+      "step": 52845
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.357983274318422e-05,
+      "loss": 0.9004,
+      "step": 52850
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.356111892649423e-05,
+      "loss": 1.0007,
+      "step": 52855
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.354240927426895e-05,
+      "loss": 1.0987,
+      "step": 52860
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.352370378768119e-05,
+      "loss": 1.0265,
+      "step": 52865
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.350500246790339e-05,
+      "loss": 1.024,
+      "step": 52870
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.348630531610773e-05,
+      "loss": 0.8475,
+      "step": 52875
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.346761233346624e-05,
+      "loss": 0.8669,
+      "step": 52880
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.344892352115055e-05,
+      "loss": 0.8269,
+      "step": 52885
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4765625,
+      "learning_rate": 3.3430238880332124e-05,
+      "loss": 0.868,
+      "step": 52890
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.3411558412182165e-05,
+      "loss": 0.8541,
+      "step": 52895
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.339288211787155e-05,
+      "loss": 0.8378,
+      "step": 52900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.3374209998570924e-05,
+      "loss": 0.9926,
+      "step": 52905
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.335554205545075e-05,
+      "loss": 0.9391,
+      "step": 52910
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.333687828968105e-05,
+      "loss": 0.9534,
+      "step": 52915
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.331821870243179e-05,
+      "loss": 0.9321,
+      "step": 52920
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.3299563294872485e-05,
+      "loss": 0.9496,
+      "step": 52925
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.3280912068172554e-05,
+      "loss": 1.0303,
+      "step": 52930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.3262265023501e-05,
+      "loss": 1.015,
+      "step": 52935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.324362216202669e-05,
+      "loss": 1.0326,
+      "step": 52940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.3224983484918205e-05,
+      "loss": 0.8786,
+      "step": 52945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.320634899334377e-05,
+      "loss": 0.9051,
+      "step": 52950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.3187718688471446e-05,
+      "loss": 0.9922,
+      "step": 52955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.316909257146905e-05,
+      "loss": 0.9637,
+      "step": 52960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.315047064350402e-05,
+      "loss": 0.9554,
+      "step": 52965
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.313185290574361e-05,
+      "loss": 0.7879,
+      "step": 52970
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.311323935935489e-05,
+      "loss": 1.035,
+      "step": 52975
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.3094630005504435e-05,
+      "loss": 0.9484,
+      "step": 52980
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.307602484535877e-05,
+      "loss": 1.1787,
+      "step": 52985
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.3057423880084114e-05,
+      "loss": 0.8673,
+      "step": 52990
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.303882711084635e-05,
+      "loss": 0.9424,
+      "step": 52995
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.3020234538811155e-05,
+      "loss": 0.9029,
+      "step": 53000
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.3001646165143986e-05,
+      "loss": 1.0614,
+      "step": 53005
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.29830619910099e-05,
+      "loss": 1.0087,
+      "step": 53010
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.2964482017573816e-05,
+      "loss": 0.9137,
+      "step": 53015
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.294590624600038e-05,
+      "loss": 0.9672,
+      "step": 53020
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.29273346774539e-05,
+      "loss": 0.932,
+      "step": 53025
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.2908767313098453e-05,
+      "loss": 0.9715,
+      "step": 53030
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.289020415409787e-05,
+      "loss": 0.9634,
+      "step": 53035
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.2871645201615765e-05,
+      "loss": 1.0967,
+      "step": 53040
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.285309045681535e-05,
+      "loss": 1.0092,
+      "step": 53045
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.498046875,
+      "learning_rate": 3.283453992085971e-05,
+      "loss": 0.902,
+      "step": 53050
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.281599359491162e-05,
+      "loss": 0.9241,
+      "step": 53055
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.279745148013354e-05,
+      "loss": 0.9882,
+      "step": 53060
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.2778913577687754e-05,
+      "loss": 0.9183,
+      "step": 53065
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.2760379888736236e-05,
+      "loss": 1.0216,
+      "step": 53070
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.274185041444067e-05,
+      "loss": 1.0704,
+      "step": 53075
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.272332515596254e-05,
+      "loss": 1.0757,
+      "step": 53080
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.270480411446298e-05,
+      "loss": 0.9288,
+      "step": 53085
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.268628729110298e-05,
+      "loss": 1.0104,
+      "step": 53090
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.49609375,
+      "learning_rate": 3.2667774687043116e-05,
+      "loss": 0.8726,
+      "step": 53095
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.264926630344386e-05,
+      "loss": 0.9184,
+      "step": 53100
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.2630762141465265e-05,
+      "loss": 0.9145,
+      "step": 53105
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.261226220226722e-05,
+      "loss": 1.1971,
+      "step": 53110
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.259376648700937e-05,
+      "loss": 0.8967,
+      "step": 53115
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.257527499685096e-05,
+      "loss": 0.9529,
+      "step": 53120
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.255678773295111e-05,
+      "loss": 1.1128,
+      "step": 53125
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.253830469646866e-05,
+      "loss": 0.8973,
+      "step": 53130
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.474609375,
+      "learning_rate": 3.251982588856208e-05,
+      "loss": 1.0136,
+      "step": 53135
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.250135131038966e-05,
+      "loss": 0.9316,
+      "step": 53140
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.24828809631094e-05,
+      "loss": 0.8264,
+      "step": 53145
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.2464414847879084e-05,
+      "loss": 0.972,
+      "step": 53150
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.244595296585614e-05,
+      "loss": 0.847,
+      "step": 53155
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.24274953181978e-05,
+      "loss": 0.9734,
+      "step": 53160
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.240904190606105e-05,
+      "loss": 1.0219,
+      "step": 53165
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.239059273060249e-05,
+      "loss": 0.9728,
+      "step": 53170
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.237214779297858e-05,
+      "loss": 1.1157,
+      "step": 53175
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.15625,
+      "learning_rate": 3.235370709434551e-05,
+      "loss": 0.9806,
+      "step": 53180
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.23352706358591e-05,
+      "loss": 0.942,
+      "step": 53185
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.494140625,
+      "learning_rate": 3.2316838418674975e-05,
+      "loss": 1.0261,
+      "step": 53190
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.2298410443948504e-05,
+      "loss": 0.9171,
+      "step": 53195
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.22799867128348e-05,
+      "loss": 0.9859,
+      "step": 53200
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.2261567226488634e-05,
+      "loss": 0.7889,
+      "step": 53205
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.2243151986064565e-05,
+      "loss": 1.0395,
+      "step": 53210
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.222474099271694e-05,
+      "loss": 0.9108,
+      "step": 53215
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.22063342475997e-05,
+      "loss": 1.0814,
+      "step": 53220
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.218793175186664e-05,
+      "loss": 0.9087,
+      "step": 53225
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.216953350667129e-05,
+      "loss": 0.9244,
+      "step": 53230
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.482421875,
+      "learning_rate": 3.2151139513166825e-05,
+      "loss": 0.8413,
+      "step": 53235
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.2132749772506176e-05,
+      "loss": 0.9695,
+      "step": 53240
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.2114364285842104e-05,
+      "loss": 0.8746,
+      "step": 53245
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4921875,
+      "learning_rate": 3.2095983054326964e-05,
+      "loss": 1.1472,
+      "step": 53250
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.2077606079112934e-05,
+      "loss": 1.0815,
+      "step": 53255
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.205923336135195e-05,
+      "loss": 1.0107,
+      "step": 53260
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.2040864902195556e-05,
+      "loss": 1.0251,
+      "step": 53265
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.2022500702795156e-05,
+      "loss": 0.8614,
+      "step": 53270
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.200414076430186e-05,
+      "loss": 0.9612,
+      "step": 53275
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.198578508786643e-05,
+      "loss": 0.9162,
+      "step": 53280
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.196743367463946e-05,
+      "loss": 1.081,
+      "step": 53285
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.194908652577126e-05,
+      "loss": 1.0349,
+      "step": 53290
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.193074364241181e-05,
+      "loss": 0.8826,
+      "step": 53295
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.191240502571085e-05,
+      "loss": 1.0065,
+      "step": 53300
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.1894070676817876e-05,
+      "loss": 0.9465,
+      "step": 53305
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.187574059688216e-05,
+      "loss": 0.8663,
+      "step": 53310
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.185741478705259e-05,
+      "loss": 1.1527,
+      "step": 53315
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.1839093248477846e-05,
+      "loss": 1.0791,
+      "step": 53320
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.1820775982306417e-05,
+      "loss": 0.922,
+      "step": 53325
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.1802462989686355e-05,
+      "loss": 0.9416,
+      "step": 53330
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.178415427176559e-05,
+      "loss": 0.9373,
+      "step": 53335
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.1765849829691785e-05,
+      "loss": 1.1576,
+      "step": 53340
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.1747549664612165e-05,
+      "loss": 0.9918,
+      "step": 53345
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.1729253777673864e-05,
+      "loss": 1.1734,
+      "step": 53350
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.1710962170023715e-05,
+      "loss": 0.8821,
+      "step": 53355
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.169267484280821e-05,
+      "loss": 0.8931,
+      "step": 53360
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.167439179717364e-05,
+      "loss": 0.9137,
+      "step": 53365
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.1656113034266046e-05,
+      "loss": 1.0828,
+      "step": 53370
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.163783855523108e-05,
+      "loss": 0.9159,
+      "step": 53375
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.1619568361214256e-05,
+      "loss": 1.0501,
+      "step": 53380
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5,
+      "learning_rate": 3.1601302453360796e-05,
+      "loss": 1.0275,
+      "step": 53385
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.158304083281557e-05,
+      "loss": 0.9749,
+      "step": 53390
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.1564783500723296e-05,
+      "loss": 0.9175,
+      "step": 53395
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.494140625,
+      "learning_rate": 3.154653045822829e-05,
+      "loss": 1.0527,
+      "step": 53400
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.466796875,
+      "learning_rate": 3.152828170647477e-05,
+      "loss": 0.933,
+      "step": 53405
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.1510037246606496e-05,
+      "loss": 0.9635,
+      "step": 53410
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.1491797079767086e-05,
+      "loss": 1.0825,
+      "step": 53415
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.147356120709989e-05,
+      "loss": 0.8809,
+      "step": 53420
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.1455329629747884e-05,
+      "loss": 1.0626,
+      "step": 53425
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.14371023488539e-05,
+      "loss": 0.8831,
+      "step": 53430
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.141887936556045e-05,
+      "loss": 0.9965,
+      "step": 53435
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.140066068100972e-05,
+      "loss": 0.9626,
+      "step": 53440
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.138244629634374e-05,
+      "loss": 1.0613,
+      "step": 53445
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.136423621270417e-05,
+      "loss": 0.9647,
+      "step": 53450
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.134603043123241e-05,
+      "loss": 0.8799,
+      "step": 53455
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.132782895306966e-05,
+      "loss": 0.9257,
+      "step": 53460
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.130963177935683e-05,
+      "loss": 0.9144,
+      "step": 53465
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.1291438911234474e-05,
+      "loss": 1.0308,
+      "step": 53470
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.1273250349842985e-05,
+      "loss": 1.0441,
+      "step": 53475
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.49609375,
+      "learning_rate": 3.125506609632247e-05,
+      "loss": 0.8567,
+      "step": 53480
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.123688615181267e-05,
+      "loss": 1.1832,
+      "step": 53485
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.121871051745317e-05,
+      "loss": 0.9365,
+      "step": 53490
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.120053919438326e-05,
+      "loss": 0.8913,
+      "step": 53495
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.11823721837419e-05,
+      "loss": 1.0006,
+      "step": 53500
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.116420948666781e-05,
+      "loss": 0.9045,
+      "step": 53505
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.114605110429946e-05,
+      "loss": 0.926,
+      "step": 53510
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.1127897037775077e-05,
+      "loss": 0.9321,
+      "step": 53515
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.1109747288232505e-05,
+      "loss": 0.9293,
+      "step": 53520
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.109160185680945e-05,
+      "loss": 0.9626,
+      "step": 53525
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.107346074464329e-05,
+      "loss": 0.8654,
+      "step": 53530
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.1055323952871086e-05,
+      "loss": 0.8872,
+      "step": 53535
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.1037191482629694e-05,
+      "loss": 1.0243,
+      "step": 53540
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.101906333505571e-05,
+      "loss": 0.9513,
+      "step": 53545
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.100093951128541e-05,
+      "loss": 0.9745,
+      "step": 53550
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.098282001245476e-05,
+      "loss": 1.0569,
+      "step": 53555
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.0964704839699596e-05,
+      "loss": 1.0132,
+      "step": 53560
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.094659399415533e-05,
+      "loss": 0.9108,
+      "step": 53565
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5,
+      "learning_rate": 3.092848747695718e-05,
+      "loss": 0.8874,
+      "step": 53570
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.091038528924015e-05,
+      "loss": 1.0389,
+      "step": 53575
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.089228743213881e-05,
+      "loss": 0.8209,
+      "step": 53580
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.0874193906787616e-05,
+      "loss": 1.1522,
+      "step": 53585
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.625,
+      "learning_rate": 3.08561047143207e-05,
+      "loss": 0.919,
+      "step": 53590
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.0838019855871855e-05,
+      "loss": 1.0189,
+      "step": 53595
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.458984375,
+      "learning_rate": 3.08199393325747e-05,
+      "loss": 1.0154,
+      "step": 53600
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.080186314556256e-05,
+      "loss": 0.9181,
+      "step": 53605
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.0783791295968454e-05,
+      "loss": 0.8568,
+      "step": 53610
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.0765723784925116e-05,
+      "loss": 1.0142,
+      "step": 53615
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.0747660613565054e-05,
+      "loss": 1.0121,
+      "step": 53620
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.072960178302052e-05,
+      "loss": 0.8851,
+      "step": 53625
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.071154729442343e-05,
+      "loss": 0.9259,
+      "step": 53630
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.069349714890545e-05,
+      "loss": 1.1588,
+      "step": 53635
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.0675451347598036e-05,
+      "loss": 1.0215,
+      "step": 53640
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.065740989163225e-05,
+      "loss": 0.9885,
+      "step": 53645
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.0639372782138995e-05,
+      "loss": 0.9443,
+      "step": 53650
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.062134002024891e-05,
+      "loss": 0.8408,
+      "step": 53655
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.060331160709219e-05,
+      "loss": 0.9833,
+      "step": 53660
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.0585287543798924e-05,
+      "loss": 0.9061,
+      "step": 53665
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4921875,
+      "learning_rate": 3.0567267831498937e-05,
+      "loss": 0.989,
+      "step": 53670
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.0549252471321635e-05,
+      "loss": 0.9358,
+      "step": 53675
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.0531241464396285e-05,
+      "loss": 0.8769,
+      "step": 53680
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.051323481185189e-05,
+      "loss": 0.8853,
+      "step": 53685
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.0495232514817016e-05,
+      "loss": 0.966,
+      "step": 53690
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.0477234574420143e-05,
+      "loss": 1.0596,
+      "step": 53695
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.045924099178942e-05,
+      "loss": 0.9349,
+      "step": 53700
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.0441251768052636e-05,
+      "loss": 0.882,
+      "step": 53705
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.0423266904337444e-05,
+      "loss": 0.9613,
+      "step": 53710
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.0405286401771095e-05,
+      "loss": 0.9987,
+      "step": 53715
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.0387310261480684e-05,
+      "loss": 0.9974,
+      "step": 53720
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.0369338484592914e-05,
+      "loss": 0.8701,
+      "step": 53725
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.0351371072234315e-05,
+      "loss": 0.9919,
+      "step": 53730
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.033340802553113e-05,
+      "loss": 0.8744,
+      "step": 53735
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.0315449345609248e-05,
+      "loss": 0.9822,
+      "step": 53740
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.029749503359436e-05,
+      "loss": 1.0412,
+      "step": 53745
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.0279545090611895e-05,
+      "loss": 0.9188,
+      "step": 53750
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.026159951778692e-05,
+      "loss": 0.99,
+      "step": 53755
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.024365831624435e-05,
+      "loss": 0.9298,
+      "step": 53760
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.0225721487108683e-05,
+      "loss": 1.0029,
+      "step": 53765
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.0207789031504295e-05,
+      "loss": 1.0409,
+      "step": 53770
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.018986095055515e-05,
+      "loss": 0.9551,
+      "step": 53775
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.017193724538505e-05,
+      "loss": 1.0299,
+      "step": 53780
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.015401791711743e-05,
+      "loss": 0.8979,
+      "step": 53785
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.0136102966875502e-05,
+      "loss": 1.1295,
+      "step": 53790
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.011819239578225e-05,
+      "loss": 0.917,
+      "step": 53795
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.0100286204960247e-05,
+      "loss": 0.9848,
+      "step": 53800
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.0082384395531926e-05,
+      "loss": 0.8899,
+      "step": 53805
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.00644869686194e-05,
+      "loss": 1.1058,
+      "step": 53810
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.004659392534449e-05,
+      "loss": 1.0225,
+      "step": 53815
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.0028705266828704e-05,
+      "loss": 0.9437,
+      "step": 53820
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.0010820994193367e-05,
+      "loss": 0.9064,
+      "step": 53825
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.9992941108559514e-05,
+      "loss": 0.8512,
+      "step": 53830
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.9975065611047804e-05,
+      "loss": 0.9165,
+      "step": 53835
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.995719450277874e-05,
+      "loss": 0.8977,
+      "step": 53840
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.993932778487253e-05,
+      "loss": 0.9242,
+      "step": 53845
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.9921465458449006e-05,
+      "loss": 0.8031,
+      "step": 53850
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.9903607524627852e-05,
+      "loss": 0.7719,
+      "step": 53855
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.988575398452843e-05,
+      "loss": 1.0093,
+      "step": 53860
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5,
+      "learning_rate": 2.9867904839269812e-05,
+      "loss": 1.0583,
+      "step": 53865
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.9850060089970753e-05,
+      "loss": 0.9877,
+      "step": 53870
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.9832219737749835e-05,
+      "loss": 0.8176,
+      "step": 53875
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.981438378372532e-05,
+      "loss": 0.8607,
+      "step": 53880
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.9796552229015128e-05,
+      "loss": 0.9302,
+      "step": 53885
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.9778725074737002e-05,
+      "loss": 1.0279,
+      "step": 53890
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.9760902322008398e-05,
+      "loss": 0.9389,
+      "step": 53895
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.9743083971946407e-05,
+      "loss": 0.9887,
+      "step": 53900
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.9725270025667917e-05,
+      "loss": 0.9757,
+      "step": 53905
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.9707460484289573e-05,
+      "loss": 0.9174,
+      "step": 53910
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.9689655348927625e-05,
+      "loss": 1.0818,
+      "step": 53915
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.9671854620698192e-05,
+      "loss": 0.884,
+      "step": 53920
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.9654058300717015e-05,
+      "loss": 1.0192,
+      "step": 53925
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.9636266390099543e-05,
+      "loss": 0.7995,
+      "step": 53930
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.9618478889961032e-05,
+      "loss": 0.9835,
+      "step": 53935
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.9600695801416454e-05,
+      "loss": 0.912,
+      "step": 53940
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.958291712558041e-05,
+      "loss": 0.9939,
+      "step": 53945
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.956514286356732e-05,
+      "loss": 0.8653,
+      "step": 53950
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.954737301649132e-05,
+      "loss": 1.0124,
+      "step": 53955
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4921875,
+      "learning_rate": 2.95296075854662e-05,
+      "loss": 0.9432,
+      "step": 53960
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.9511846571605517e-05,
+      "loss": 0.8652,
+      "step": 53965
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.949408997602262e-05,
+      "loss": 0.9531,
+      "step": 53970
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.9476337799830456e-05,
+      "loss": 0.9282,
+      "step": 53975
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.498046875,
+      "learning_rate": 2.945859004414172e-05,
+      "loss": 1.1041,
+      "step": 53980
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.9440846710068902e-05,
+      "loss": 0.948,
+      "step": 53985
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.9423107798724214e-05,
+      "loss": 0.9718,
+      "step": 53990
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.9405373311219465e-05,
+      "loss": 0.9747,
+      "step": 53995
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.938764324866632e-05,
+      "loss": 0.8916,
+      "step": 54000
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.9369917612176157e-05,
+      "loss": 0.951,
+      "step": 54005
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.9352196402859955e-05,
+      "loss": 1.0411,
+      "step": 54010
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.9334479621828558e-05,
+      "loss": 0.8444,
+      "step": 54015
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.931676727019248e-05,
+      "loss": 1.1292,
+      "step": 54020
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.9299059349061942e-05,
+      "loss": 0.9727,
+      "step": 54025
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.9281355859546843e-05,
+      "loss": 0.8993,
+      "step": 54030
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.9263656802756935e-05,
+      "loss": 1.0242,
+      "step": 54035
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.9245962179801544e-05,
+      "loss": 0.9757,
+      "step": 54040
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.9228271991789834e-05,
+      "loss": 1.0083,
+      "step": 54045
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.9210586239830673e-05,
+      "loss": 0.9915,
+      "step": 54050
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.919290492503255e-05,
+      "loss": 1.0139,
+      "step": 54055
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.9175228048503788e-05,
+      "loss": 1.1561,
+      "step": 54060
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.9157555611352427e-05,
+      "loss": 1.0815,
+      "step": 54065
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.9139887614686134e-05,
+      "loss": 0.9104,
+      "step": 54070
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.912222405961239e-05,
+      "loss": 1.1101,
+      "step": 54075
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.910456494723839e-05,
+      "loss": 1.0586,
+      "step": 54080
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.908691027867101e-05,
+      "loss": 0.7989,
+      "step": 54085
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.9069260055016813e-05,
+      "loss": 0.9496,
+      "step": 54090
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.905161427738219e-05,
+      "loss": 0.9066,
+      "step": 54095
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.9033972946873212e-05,
+      "loss": 0.9965,
+      "step": 54100
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.9016336064595607e-05,
+      "loss": 0.8419,
+      "step": 54105
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.89987036316549e-05,
+      "loss": 0.7921,
+      "step": 54110
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.898107564915634e-05,
+      "loss": 0.8745,
+      "step": 54115
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.8963452118204827e-05,
+      "loss": 0.7942,
+      "step": 54120
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.8945833039905023e-05,
+      "loss": 0.8932,
+      "step": 54125
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.8928218415361397e-05,
+      "loss": 0.8994,
+      "step": 54130
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.8910608245677918e-05,
+      "loss": 1.0687,
+      "step": 54135
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.8893002531958478e-05,
+      "loss": 0.8616,
+      "step": 54140
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.8875401275306657e-05,
+      "loss": 0.9944,
+      "step": 54145
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.8857804476825655e-05,
+      "loss": 0.7736,
+      "step": 54150
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.8840212137618495e-05,
+      "loss": 0.9387,
+      "step": 54155
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.882262425878791e-05,
+      "loss": 0.8779,
+      "step": 54160
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.880504084143627e-05,
+      "loss": 1.0307,
+      "step": 54165
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.8787461886665756e-05,
+      "loss": 0.8971,
+      "step": 54170
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.8769887395578276e-05,
+      "loss": 0.8709,
+      "step": 54175
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.875231736927534e-05,
+      "loss": 0.997,
+      "step": 54180
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.8734751808858318e-05,
+      "loss": 0.9288,
+      "step": 54185
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.87171907154282e-05,
+      "loss": 1.0063,
+      "step": 54190
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.869963409008577e-05,
+      "loss": 1.0611,
+      "step": 54195
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.868208193393146e-05,
+      "loss": 0.9429,
+      "step": 54200
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.8664534248065467e-05,
+      "loss": 0.8201,
+      "step": 54205
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.8646991033587754e-05,
+      "loss": 0.8655,
+      "step": 54210
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.8629452291597868e-05,
+      "loss": 1.0149,
+      "step": 54215
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.8611918023195193e-05,
+      "loss": 0.9014,
+      "step": 54220
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.8594388229478843e-05,
+      "loss": 0.933,
+      "step": 54225
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.857686291154753e-05,
+      "loss": 1.0044,
+      "step": 54230
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.8559342070499817e-05,
+      "loss": 0.9215,
+      "step": 54235
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.854182570743391e-05,
+      "loss": 0.8283,
+      "step": 54240
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.8524313823447724e-05,
+      "loss": 0.8395,
+      "step": 54245
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.8506806419638953e-05,
+      "loss": 1.0228,
+      "step": 54250
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.8489303497105012e-05,
+      "loss": 1.0182,
+      "step": 54255
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.8471805056942936e-05,
+      "loss": 1.0948,
+      "step": 54260
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.8454311100249588e-05,
+      "loss": 0.9233,
+      "step": 54265
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.8436821628121523e-05,
+      "loss": 1.0684,
+      "step": 54270
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.8419336641654958e-05,
+      "loss": 0.9943,
+      "step": 54275
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.8401856141945893e-05,
+      "loss": 0.8644,
+      "step": 54280
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.8384380130090072e-05,
+      "loss": 0.99,
+      "step": 54285
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.8366908607182852e-05,
+      "loss": 1.0081,
+      "step": 54290
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.834944157431936e-05,
+      "loss": 1.066,
+      "step": 54295
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.833197903259447e-05,
+      "loss": 0.9283,
+      "step": 54300
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.83145209831028e-05,
+      "loss": 0.8595,
+      "step": 54305
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.829706742693855e-05,
+      "loss": 1.1134,
+      "step": 54310
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.827961836519578e-05,
+      "loss": 0.8888,
+      "step": 54315
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.826217379896825e-05,
+      "loss": 1.1328,
+      "step": 54320
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5,
+      "learning_rate": 2.8244733729349338e-05,
+      "loss": 0.8952,
+      "step": 54325
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.8227298157432235e-05,
+      "loss": 0.9766,
+      "step": 54330
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.820986708430986e-05,
+      "loss": 0.9542,
+      "step": 54335
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.8192440511074782e-05,
+      "loss": 0.964,
+      "step": 54340
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.817501843881928e-05,
+      "loss": 1.0103,
+      "step": 54345
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.8157600868635448e-05,
+      "loss": 0.9626,
+      "step": 54350
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.8140187801614993e-05,
+      "loss": 0.8749,
+      "step": 54355
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.8122779238849407e-05,
+      "loss": 1.015,
+      "step": 54360
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.8105375181429906e-05,
+      "loss": 0.9016,
+      "step": 54365
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.808797563044734e-05,
+      "loss": 0.9122,
+      "step": 54370
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.8070580586992356e-05,
+      "loss": 0.8866,
+      "step": 54375
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.8053190052155343e-05,
+      "loss": 1.0144,
+      "step": 54380
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.803580402702628e-05,
+      "loss": 1.0065,
+      "step": 54385
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.8018422512694985e-05,
+      "loss": 0.976,
+      "step": 54390
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.800104551025098e-05,
+      "loss": 1.0068,
+      "step": 54395
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.798367302078344e-05,
+      "loss": 0.9593,
+      "step": 54400
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.7966305045381268e-05,
+      "loss": 1.1082,
+      "step": 54405
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.7948941585133126e-05,
+      "loss": 0.9987,
+      "step": 54410
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.793158264112743e-05,
+      "loss": 0.8221,
+      "step": 54415
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5,
+      "learning_rate": 2.7914228214452165e-05,
+      "loss": 0.7934,
+      "step": 54420
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.7896878306195184e-05,
+      "loss": 0.9969,
+      "step": 54425
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.486328125,
+      "learning_rate": 2.787953291744402e-05,
+      "loss": 0.9567,
+      "step": 54430
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.786219204928584e-05,
+      "loss": 0.923,
+      "step": 54435
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.7844855702807616e-05,
+      "loss": 1.0718,
+      "step": 54440
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.7827523879096052e-05,
+      "loss": 0.9112,
+      "step": 54445
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.781019657923748e-05,
+      "loss": 0.9836,
+      "step": 54450
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.7792873804317976e-05,
+      "loss": 0.8287,
+      "step": 54455
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.7775555555423416e-05,
+      "loss": 1.008,
+      "step": 54460
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.7758241833639252e-05,
+      "loss": 0.9603,
+      "step": 54465
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.7740932640050775e-05,
+      "loss": 0.88,
+      "step": 54470
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.7723627975742962e-05,
+      "loss": 1.0548,
+      "step": 54475
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.7706327841800428e-05,
+      "loss": 1.0476,
+      "step": 54480
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.768903223930761e-05,
+      "loss": 0.9371,
+      "step": 54485
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.7671741169348632e-05,
+      "loss": 0.9731,
+      "step": 54490
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.7654454633007266e-05,
+      "loss": 1.1529,
+      "step": 54495
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.7637172631367104e-05,
+      "loss": 0.9571,
+      "step": 54500
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.7619895165511345e-05,
+      "loss": 0.9497,
+      "step": 54505
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.7602622236523023e-05,
+      "loss": 0.8806,
+      "step": 54510
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.7585353845484764e-05,
+      "loss": 0.9026,
+      "step": 54515
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.7568089993478995e-05,
+      "loss": 0.8992,
+      "step": 54520
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.755083068158787e-05,
+      "loss": 1.0083,
+      "step": 54525
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.7533575910893162e-05,
+      "loss": 1.0233,
+      "step": 54530
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.7516325682476453e-05,
+      "loss": 0.9418,
+      "step": 54535
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.749907999741902e-05,
+      "loss": 1.0514,
+      "step": 54540
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.74818388568018e-05,
+      "loss": 0.9402,
+      "step": 54545
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.746460226170554e-05,
+      "loss": 1.0042,
+      "step": 54550
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.74473702132106e-05,
+      "loss": 0.9899,
+      "step": 54555
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.7430142712397157e-05,
+      "loss": 0.8437,
+      "step": 54560
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.7412919760344978e-05,
+      "loss": 0.9307,
+      "step": 54565
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.7395701358133664e-05,
+      "loss": 0.8752,
+      "step": 54570
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.73784875068425e-05,
+      "loss": 0.8699,
+      "step": 54575
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.7361278207550423e-05,
+      "loss": 0.9109,
+      "step": 54580
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.734407346133616e-05,
+      "loss": 1.0958,
+      "step": 54585
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.7326873269278154e-05,
+      "loss": 0.8842,
+      "step": 54590
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.7309677632454454e-05,
+      "loss": 0.9275,
+      "step": 54595
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.7292486551942987e-05,
+      "loss": 0.8883,
+      "step": 54600
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.727530002882127e-05,
+      "loss": 0.96,
+      "step": 54605
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.725811806416655e-05,
+      "loss": 0.9566,
+      "step": 54610
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.498046875,
+      "learning_rate": 2.7240940659055826e-05,
+      "loss": 0.8742,
+      "step": 54615
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.7223767814565837e-05,
+      "loss": 0.8872,
+      "step": 54620
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.7206599531772948e-05,
+      "loss": 0.8923,
+      "step": 54625
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.7189435811753293e-05,
+      "loss": 0.899,
+      "step": 54630
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.717227665558276e-05,
+      "loss": 0.8567,
+      "step": 54635
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.7155122064336835e-05,
+      "loss": 0.9793,
+      "step": 54640
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.713797203909083e-05,
+      "loss": 0.938,
+      "step": 54645
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.7120826580919746e-05,
+      "loss": 0.9868,
+      "step": 54650
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.466796875,
+      "learning_rate": 2.710368569089826e-05,
+      "loss": 0.8248,
+      "step": 54655
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.7086549370100746e-05,
+      "loss": 0.8802,
+      "step": 54660
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.7069417619601343e-05,
+      "loss": 0.855,
+      "step": 54665
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.705229044047396e-05,
+      "loss": 1.0715,
+      "step": 54670
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.703516783379204e-05,
+      "loss": 0.8297,
+      "step": 54675
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.701804980062891e-05,
+      "loss": 0.9352,
+      "step": 54680
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.700093634205757e-05,
+      "loss": 0.9531,
+      "step": 54685
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.6983827459150646e-05,
+      "loss": 0.9047,
+      "step": 54690
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.486328125,
+      "learning_rate": 2.6966723152980578e-05,
+      "loss": 0.8579,
+      "step": 54695
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.6949623424619507e-05,
+      "loss": 1.0137,
+      "step": 54700
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.6932528275139212e-05,
+      "loss": 1.0013,
+      "step": 54705
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.691543770561129e-05,
+      "loss": 0.9706,
+      "step": 54710
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.6898351717106962e-05,
+      "loss": 1.0458,
+      "step": 54715
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.688127031069718e-05,
+      "loss": 1.0411,
+      "step": 54720
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.686419348745265e-05,
+      "loss": 0.9726,
+      "step": 54725
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.6847121248443796e-05,
+      "loss": 1.1117,
+      "step": 54730
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.6830053594740668e-05,
+      "loss": 0.9083,
+      "step": 54735
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.6812990527413116e-05,
+      "loss": 0.9351,
+      "step": 54740
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.6795932047530705e-05,
+      "loss": 1.0094,
+      "step": 54745
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.625,
+      "learning_rate": 2.67788781561626e-05,
+      "loss": 0.935,
+      "step": 54750
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.6761828854377812e-05,
+      "loss": 0.8739,
+      "step": 54755
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.674478414324504e-05,
+      "loss": 0.9497,
+      "step": 54760
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.6727744023832622e-05,
+      "loss": 1.0655,
+      "step": 54765
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.6710708497208625e-05,
+      "loss": 1.0258,
+      "step": 54770
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.66936775644409e-05,
+      "loss": 1.1597,
+      "step": 54775
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.667665122659697e-05,
+      "loss": 0.8485,
+      "step": 54780
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.665962948474403e-05,
+      "loss": 0.8748,
+      "step": 54785
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.6642612339949037e-05,
+      "loss": 1.0183,
+      "step": 54790
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.6625599793278676e-05,
+      "loss": 0.9696,
+      "step": 54795
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.6608591845799268e-05,
+      "loss": 0.9695,
+      "step": 54800
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.6591588498576903e-05,
+      "loss": 0.9984,
+      "step": 54805
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.6574589752677405e-05,
+      "loss": 0.927,
+      "step": 54810
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.6557595609166242e-05,
+      "loss": 0.9817,
+      "step": 54815
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.65406060691086e-05,
+      "loss": 0.7142,
+      "step": 54820
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.6523621133569464e-05,
+      "loss": 0.8814,
+      "step": 54825
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.6506640803613412e-05,
+      "loss": 1.0073,
+      "step": 54830
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.6489665080304814e-05,
+      "loss": 0.9157,
+      "step": 54835
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.647269396470776e-05,
+      "loss": 0.9138,
+      "step": 54840
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.645572745788597e-05,
+      "loss": 0.8308,
+      "step": 54845
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.6438765560902933e-05,
+      "loss": 0.9832,
+      "step": 54850
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.642180827482188e-05,
+      "loss": 0.8653,
+      "step": 54855
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5,
+      "learning_rate": 2.6404855600705648e-05,
+      "loss": 0.8943,
+      "step": 54860
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.638790753961692e-05,
+      "loss": 0.835,
+      "step": 54865
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.6370964092617957e-05,
+      "loss": 0.9696,
+      "step": 54870
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.6354025260770843e-05,
+      "loss": 0.9426,
+      "step": 54875
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.6337091045137275e-05,
+      "loss": 0.9313,
+      "step": 54880
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.6320161446778735e-05,
+      "loss": 1.0948,
+      "step": 54885
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.630323646675642e-05,
+      "loss": 1.0268,
+      "step": 54890
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.6286316106131148e-05,
+      "loss": 0.7687,
+      "step": 54895
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.6269400365963536e-05,
+      "loss": 0.9385,
+      "step": 54900
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.6252489247313915e-05,
+      "loss": 0.915,
+      "step": 54905
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.6235582751242226e-05,
+      "loss": 0.8307,
+      "step": 54910
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.621868087880822e-05,
+      "loss": 0.8782,
+      "step": 54915
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.6201783631071385e-05,
+      "loss": 1.0213,
+      "step": 54920
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.618489100909074e-05,
+      "loss": 0.9641,
+      "step": 54925
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.61680030139252e-05,
+      "loss": 0.8366,
+      "step": 54930
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.6151119646633347e-05,
+      "loss": 1.034,
+      "step": 54935
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.613424090827339e-05,
+      "loss": 0.9292,
+      "step": 54940
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.611736679990334e-05,
+      "loss": 0.8975,
+      "step": 54945
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.6100497322580908e-05,
+      "loss": 0.9267,
+      "step": 54950
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.6083632477363452e-05,
+      "loss": 1.0463,
+      "step": 54955
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.6066772265308083e-05,
+      "loss": 0.9473,
+      "step": 54960
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.6049916687471666e-05,
+      "loss": 0.8829,
+      "step": 54965
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.625,
+      "learning_rate": 2.603306574491069e-05,
+      "loss": 1.0004,
+      "step": 54970
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.6016219438681367e-05,
+      "loss": 0.9185,
+      "step": 54975
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.599937776983967e-05,
+      "loss": 1.0157,
+      "step": 54980
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.5982540739441284e-05,
+      "loss": 1.0383,
+      "step": 54985
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.5965708348541518e-05,
+      "loss": 1.072,
+      "step": 54990
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.5948880598195467e-05,
+      "loss": 1.0174,
+      "step": 54995
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.593205748945795e-05,
+      "loss": 1.1266,
+      "step": 55000
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.5915239023383387e-05,
+      "loss": 0.9785,
+      "step": 55005
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.5898425201026022e-05,
+      "loss": 0.8967,
+      "step": 55010
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.588161602343979e-05,
+      "loss": 0.9996,
+      "step": 55015
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.5864811491678255e-05,
+      "loss": 0.9827,
+      "step": 55020
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.5848011606794797e-05,
+      "loss": 0.9682,
+      "step": 55025
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.5831216369842416e-05,
+      "loss": 0.803,
+      "step": 55030
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.581442578187384e-05,
+      "loss": 1.03,
+      "step": 55035
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.5797639843941547e-05,
+      "loss": 1.0129,
+      "step": 55040
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.578085855709773e-05,
+      "loss": 0.8681,
+      "step": 55045
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.576408192239419e-05,
+      "loss": 0.8666,
+      "step": 55050
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.5747309940882546e-05,
+      "loss": 0.995,
+      "step": 55055
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5,
+      "learning_rate": 2.5730542613614118e-05,
+      "loss": 0.904,
+      "step": 55060
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.5713779941639826e-05,
+      "loss": 0.9581,
+      "step": 55065
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.625,
+      "learning_rate": 2.5697021926010413e-05,
+      "loss": 0.8722,
+      "step": 55070
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.5680268567776323e-05,
+      "loss": 1.021,
+      "step": 55075
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.566351986798764e-05,
+      "loss": 0.8893,
+      "step": 55080
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.564677582769416e-05,
+      "loss": 1.1019,
+      "step": 55085
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.563003644794546e-05,
+      "loss": 1.16,
+      "step": 55090
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.5613301729790794e-05,
+      "loss": 1.0489,
+      "step": 55095
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.559657167427908e-05,
+      "loss": 0.954,
+      "step": 55100
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.5579846282458987e-05,
+      "loss": 0.9067,
+      "step": 55105
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.5563125555378908e-05,
+      "loss": 0.8201,
+      "step": 55110
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.5546409494086876e-05,
+      "loss": 0.9355,
+      "step": 55115
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.5529698099630694e-05,
+      "loss": 0.951,
+      "step": 55120
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.551299137305787e-05,
+      "loss": 0.9268,
+      "step": 55125
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.549628931541559e-05,
+      "loss": 0.9385,
+      "step": 55130
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.5479591927750713e-05,
+      "loss": 0.9118,
+      "step": 55135
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.5462899211109915e-05,
+      "loss": 0.9354,
+      "step": 55140
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.5446211166539448e-05,
+      "loss": 0.993,
+      "step": 55145
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.5429527795085384e-05,
+      "loss": 1.1194,
+      "step": 55150
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.541284909779348e-05,
+      "loss": 0.9855,
+      "step": 55155
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.5396175075709107e-05,
+      "loss": 1.1419,
+      "step": 55160
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.537950572987744e-05,
+      "loss": 1.0015,
+      "step": 55165
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.536284106134338e-05,
+      "loss": 0.985,
+      "step": 55170
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.534618107115141e-05,
+      "loss": 0.9552,
+      "step": 55175
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.5329525760345875e-05,
+      "loss": 0.9298,
+      "step": 55180
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.531287512997067e-05,
+      "loss": 0.8846,
+      "step": 55185
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.529622918106954e-05,
+      "loss": 0.9645,
+      "step": 55190
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.527958791468582e-05,
+      "loss": 0.9765,
+      "step": 55195
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.5262951331862628e-05,
+      "loss": 1.0036,
+      "step": 55200
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.5246319433642794e-05,
+      "loss": 1.0886,
+      "step": 55205
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.5229692221068767e-05,
+      "loss": 0.9325,
+      "step": 55210
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.521306969518279e-05,
+      "loss": 1.0222,
+      "step": 55215
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.5196451857026805e-05,
+      "loss": 1.0714,
+      "step": 55220
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.517983870764238e-05,
+      "loss": 0.9616,
+      "step": 55225
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.516323024807088e-05,
+      "loss": 0.9596,
+      "step": 55230
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.5146626479353375e-05,
+      "loss": 0.8699,
+      "step": 55235
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.5130027402530564e-05,
+      "loss": 0.8693,
+      "step": 55240
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.5113433018642883e-05,
+      "loss": 0.921,
+      "step": 55245
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.5096843328730503e-05,
+      "loss": 1.1023,
+      "step": 55250
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.508025833383333e-05,
+      "loss": 0.9598,
+      "step": 55255
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.5063678034990855e-05,
+      "loss": 0.9467,
+      "step": 55260
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.5047102433242385e-05,
+      "loss": 0.8873,
+      "step": 55265
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.5030531529626932e-05,
+      "loss": 0.9983,
+      "step": 55270
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.501396532518311e-05,
+      "loss": 0.9313,
+      "step": 55275
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.499740382094937e-05,
+      "loss": 1.0263,
+      "step": 55280
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.4980847017963792e-05,
+      "loss": 1.1104,
+      "step": 55285
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.496429491726413e-05,
+      "loss": 1.0691,
+      "step": 55290
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.4947747519887922e-05,
+      "loss": 1.0045,
+      "step": 55295
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.49312048268724e-05,
+      "loss": 0.7354,
+      "step": 55300
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.491466683925443e-05,
+      "loss": 0.9239,
+      "step": 55305
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.4898133558070647e-05,
+      "loss": 1.1663,
+      "step": 55310
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.4881604984357432e-05,
+      "loss": 0.8713,
+      "step": 55315
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.4865081119150734e-05,
+      "loss": 0.8766,
+      "step": 55320
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.4848561963486318e-05,
+      "loss": 0.9103,
+      "step": 55325
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.483204751839966e-05,
+      "loss": 0.9963,
+      "step": 55330
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.4815537784925846e-05,
+      "loss": 0.9386,
+      "step": 55335
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.4799032764099784e-05,
+      "loss": 0.9169,
+      "step": 55340
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.4782532456955955e-05,
+      "loss": 0.9657,
+      "step": 55345
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.4766036864528696e-05,
+      "loss": 0.8921,
+      "step": 55350
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.4749545987851897e-05,
+      "loss": 0.8551,
+      "step": 55355
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.473305982795926e-05,
+      "loss": 0.9152,
+      "step": 55360
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.4716578385884182e-05,
+      "loss": 1.1017,
+      "step": 55365
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.4700101662659692e-05,
+      "loss": 1.047,
+      "step": 55370
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.4683629659318575e-05,
+      "loss": 0.8259,
+      "step": 55375
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.466716237689337e-05,
+      "loss": 0.8812,
+      "step": 55380
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.465069981641619e-05,
+      "loss": 0.9734,
+      "step": 55385
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.463424197891898e-05,
+      "loss": 0.9448,
+      "step": 55390
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.4617788865433332e-05,
+      "loss": 1.0034,
+      "step": 55395
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.4601340476990498e-05,
+      "loss": 0.9685,
+      "step": 55400
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.4584896814621518e-05,
+      "loss": 0.979,
+      "step": 55405
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.456845787935712e-05,
+      "loss": 0.9138,
+      "step": 55410
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.4552023672227666e-05,
+      "loss": 1.0635,
+      "step": 55415
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.453559419426329e-05,
+      "loss": 0.9942,
+      "step": 55420
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.48828125,
+      "learning_rate": 2.451916944649385e-05,
+      "loss": 0.8659,
+      "step": 55425
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.45027494299488e-05,
+      "loss": 0.8897,
+      "step": 55430
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.4486334145657408e-05,
+      "loss": 1.0097,
+      "step": 55435
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.4469923594648626e-05,
+      "loss": 0.8444,
+      "step": 55440
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.445351777795104e-05,
+      "loss": 0.9892,
+      "step": 55445
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.4437116696592988e-05,
+      "loss": 0.8611,
+      "step": 55450
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.4420720351602512e-05,
+      "loss": 0.912,
+      "step": 55455
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.4404328744007387e-05,
+      "loss": 0.8703,
+      "step": 55460
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.4387941874835008e-05,
+      "loss": 0.9272,
+      "step": 55465
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.437155974511255e-05,
+      "loss": 0.8864,
+      "step": 55470
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.435518235586688e-05,
+      "loss": 0.9701,
+      "step": 55475
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.433880970812451e-05,
+      "loss": 0.9989,
+      "step": 55480
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.432244180291172e-05,
+      "loss": 0.9397,
+      "step": 55485
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.4306078641254493e-05,
+      "loss": 0.9598,
+      "step": 55490
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.428972022417847e-05,
+      "loss": 1.0931,
+      "step": 55495
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.4273366552708975e-05,
+      "loss": 1.1147,
+      "step": 55500
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.425701762787115e-05,
+      "loss": 0.9353,
+      "step": 55505
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.4240673450689676e-05,
+      "loss": 0.9107,
+      "step": 55510
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.422433402218909e-05,
+      "loss": 0.9921,
+      "step": 55515
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.4207999343393574e-05,
+      "loss": 0.9116,
+      "step": 55520
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.4191669415326945e-05,
+      "loss": 0.9435,
+      "step": 55525
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.4175344239012822e-05,
+      "loss": 0.9898,
+      "step": 55530
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.4159023815474502e-05,
+      "loss": 0.932,
+      "step": 55535
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.4142708145734916e-05,
+      "loss": 0.8182,
+      "step": 55540
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.4126397230816778e-05,
+      "loss": 0.8407,
+      "step": 55545
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.41100910717425e-05,
+      "loss": 1.0069,
+      "step": 55550
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.409378966953415e-05,
+      "loss": 0.9041,
+      "step": 55555
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.4077493025213483e-05,
+      "loss": 0.893,
+      "step": 55560
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.4061201139802024e-05,
+      "loss": 1.0359,
+      "step": 55565
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.4044914014320995e-05,
+      "loss": 0.8965,
+      "step": 55570
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.4028631649791233e-05,
+      "loss": 0.8205,
+      "step": 55575
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.4012354047233354e-05,
+      "loss": 0.8566,
+      "step": 55580
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.3996081207667698e-05,
+      "loss": 0.8672,
+      "step": 55585
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.3979813132114215e-05,
+      "loss": 0.9229,
+      "step": 55590
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.3963549821592613e-05,
+      "loss": 1.1278,
+      "step": 55595
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.3947291277122362e-05,
+      "loss": 1.1218,
+      "step": 55600
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.3931037499722466e-05,
+      "loss": 1.045,
+      "step": 55605
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4609375,
+      "learning_rate": 2.3914788490411765e-05,
+      "loss": 0.9761,
+      "step": 55610
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.625,
+      "learning_rate": 2.3898544250208808e-05,
+      "loss": 0.9598,
+      "step": 55615
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.3882304780131738e-05,
+      "loss": 0.9488,
+      "step": 55620
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.38660700811985e-05,
+      "loss": 0.888,
+      "step": 55625
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.625,
+      "learning_rate": 2.3849840154426716e-05,
+      "loss": 1.1086,
+      "step": 55630
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.3833615000833666e-05,
+      "loss": 0.9319,
+      "step": 55635
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.381739462143636e-05,
+      "loss": 0.9932,
+      "step": 55640
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.380117901725156e-05,
+      "loss": 0.7757,
+      "step": 55645
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.378496818929561e-05,
+      "loss": 1.0451,
+      "step": 55650
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.3768762138584688e-05,
+      "loss": 0.9445,
+      "step": 55655
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.3752560866134544e-05,
+      "loss": 0.8609,
+      "step": 55660
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.3736364372960763e-05,
+      "loss": 0.9798,
+      "step": 55665
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.3720172660078478e-05,
+      "loss": 1.0855,
+      "step": 55670
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.370398572850264e-05,
+      "loss": 1.0424,
+      "step": 55675
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.3687803579247915e-05,
+      "loss": 0.9406,
+      "step": 55680
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.3671626213328534e-05,
+      "loss": 1.0556,
+      "step": 55685
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.365545363175856e-05,
+      "loss": 1.0199,
+      "step": 55690
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.363928583555173e-05,
+      "loss": 0.9645,
+      "step": 55695
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.36231228257214e-05,
+      "loss": 0.9919,
+      "step": 55700
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.3606964603280746e-05,
+      "loss": 0.9985,
+      "step": 55705
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.3590811169242554e-05,
+      "loss": 0.9128,
+      "step": 55710
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.3574662524619318e-05,
+      "loss": 1.0251,
+      "step": 55715
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.3558518670423268e-05,
+      "loss": 0.9368,
+      "step": 55720
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.3542379607666365e-05,
+      "loss": 1.0604,
+      "step": 55725
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4921875,
+      "learning_rate": 2.3526245337360153e-05,
+      "loss": 1.0838,
+      "step": 55730
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.351011586051598e-05,
+      "loss": 0.9904,
+      "step": 55735
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.349399117814488e-05,
+      "loss": 1.0388,
+      "step": 55740
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.3477871291257525e-05,
+      "loss": 1.089,
+      "step": 55745
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.346175620086435e-05,
+      "loss": 0.7987,
+      "step": 55750
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.3445645907975488e-05,
+      "loss": 1.0289,
+      "step": 55755
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.625,
+      "learning_rate": 2.3429540413600736e-05,
+      "loss": 0.9907,
+      "step": 55760
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.3413439718749562e-05,
+      "loss": 0.9355,
+      "step": 55765
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.3397343824431216e-05,
+      "loss": 0.9169,
+      "step": 55770
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.3381252731654633e-05,
+      "loss": 0.97,
+      "step": 55775
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.3365166441428366e-05,
+      "loss": 0.9656,
+      "step": 55780
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.3349084954760735e-05,
+      "loss": 1.0667,
+      "step": 55785
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.3333008272659784e-05,
+      "loss": 0.8738,
+      "step": 55790
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.331693639613317e-05,
+      "loss": 1.0558,
+      "step": 55795
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.3300869326188313e-05,
+      "loss": 0.9356,
+      "step": 55800
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.328480706383236e-05,
+      "loss": 0.8565,
+      "step": 55805
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.3268749610072062e-05,
+      "loss": 0.884,
+      "step": 55810
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.325269696591389e-05,
+      "loss": 0.8756,
+      "step": 55815
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.323664913236412e-05,
+      "loss": 0.7658,
+      "step": 55820
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.322060611042858e-05,
+      "loss": 0.882,
+      "step": 55825
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.3204567901112895e-05,
+      "loss": 0.9241,
+      "step": 55830
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.3188534505422377e-05,
+      "loss": 1.0328,
+      "step": 55835
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.3172505924361976e-05,
+      "loss": 0.8019,
+      "step": 55840
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4921875,
+      "learning_rate": 2.31564821589364e-05,
+      "loss": 0.9576,
+      "step": 55845
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.3140463210150065e-05,
+      "loss": 1.0144,
+      "step": 55850
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.3124449079007004e-05,
+      "loss": 0.8785,
+      "step": 55855
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.3108439766511038e-05,
+      "loss": 1.0206,
+      "step": 55860
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.3092435273665657e-05,
+      "loss": 0.9757,
+      "step": 55865
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.3076435601474024e-05,
+      "loss": 0.9286,
+      "step": 55870
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.3060440750938994e-05,
+      "loss": 0.9818,
+      "step": 55875
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.304445072306316e-05,
+      "loss": 0.8368,
+      "step": 55880
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.3028465518848828e-05,
+      "loss": 0.9012,
+      "step": 55885
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.301248513929791e-05,
+      "loss": 1.0057,
+      "step": 55890
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.2996509585412117e-05,
+      "loss": 0.8797,
+      "step": 55895
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.2980538858192825e-05,
+      "loss": 0.9028,
+      "step": 55900
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.296457295864104e-05,
+      "loss": 0.9417,
+      "step": 55905
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.2948611887757566e-05,
+      "loss": 0.9962,
+      "step": 55910
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.2932655646542876e-05,
+      "loss": 0.9032,
+      "step": 55915
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.29167042359971e-05,
+      "loss": 1.0349,
+      "step": 55920
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5,
+      "learning_rate": 2.2900757657120075e-05,
+      "loss": 0.9714,
+      "step": 55925
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.288481591091136e-05,
+      "loss": 1.1025,
+      "step": 55930
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.2868878998370247e-05,
+      "loss": 0.916,
+      "step": 55935
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.2852946920495612e-05,
+      "loss": 0.92,
+      "step": 55940
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.2837019678286143e-05,
+      "loss": 0.9217,
+      "step": 55945
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.2821097272740143e-05,
+      "loss": 0.9212,
+      "step": 55950
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.2805179704855674e-05,
+      "loss": 0.8215,
+      "step": 55955
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.2789266975630474e-05,
+      "loss": 0.967,
+      "step": 55960
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.2773359086061928e-05,
+      "loss": 0.8872,
+      "step": 55965
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.275745603714723e-05,
+      "loss": 0.9856,
+      "step": 55970
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.274155782988312e-05,
+      "loss": 0.8701,
+      "step": 55975
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.2725664465266182e-05,
+      "loss": 0.9997,
+      "step": 55980
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.2709775944292576e-05,
+      "loss": 1.0167,
+      "step": 55985
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.269389226795825e-05,
+      "loss": 0.924,
+      "step": 55990
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.2678013437258815e-05,
+      "loss": 1.0758,
+      "step": 55995
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.2662139453189547e-05,
+      "loss": 0.9348,
+      "step": 56000
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.2646270316745445e-05,
+      "loss": 0.8797,
+      "step": 56005
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.2630406028921258e-05,
+      "loss": 0.9258,
+      "step": 56010
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.2614546590711295e-05,
+      "loss": 1.0569,
+      "step": 56015
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.259869200310972e-05,
+      "loss": 1.0872,
+      "step": 56020
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.258284226711026e-05,
+      "loss": 0.9533,
+      "step": 56025
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.2566997383706445e-05,
+      "loss": 0.9205,
+      "step": 56030
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.2551157353891393e-05,
+      "loss": 1.0329,
+      "step": 56035
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.2535322178658003e-05,
+      "loss": 0.7919,
+      "step": 56040
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.2519491858998875e-05,
+      "loss": 0.9221,
+      "step": 56045
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.2503666395906208e-05,
+      "loss": 0.9291,
+      "step": 56050
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.2487845790371998e-05,
+      "loss": 0.9941,
+      "step": 56055
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.2472030043387914e-05,
+      "loss": 0.9139,
+      "step": 56060
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.2456219155945256e-05,
+      "loss": 0.9235,
+      "step": 56065
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.244041312903511e-05,
+      "loss": 0.9313,
+      "step": 56070
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.2424611963648244e-05,
+      "loss": 0.9421,
+      "step": 56075
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.2408815660774995e-05,
+      "loss": 0.9214,
+      "step": 56080
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.2393024221405555e-05,
+      "loss": 0.999,
+      "step": 56085
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.237723764652977e-05,
+      "loss": 0.8612,
+      "step": 56090
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.236145593713711e-05,
+      "loss": 0.9955,
+      "step": 56095
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.455078125,
+      "learning_rate": 2.234567909421681e-05,
+      "loss": 0.8391,
+      "step": 56100
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.2329907118757807e-05,
+      "loss": 0.9794,
+      "step": 56105
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.2314140011748662e-05,
+      "loss": 1.0065,
+      "step": 56110
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.2298377774177702e-05,
+      "loss": 1.0236,
+      "step": 56115
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.486328125,
+      "learning_rate": 2.228262040703294e-05,
+      "loss": 1.0006,
+      "step": 56120
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.2266867911302048e-05,
+      "loss": 0.7434,
+      "step": 56125
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.2251120287972384e-05,
+      "loss": 0.8261,
+      "step": 56130
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.2235377538031033e-05,
+      "loss": 0.8794,
+      "step": 56135
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.221963966246482e-05,
+      "loss": 0.9186,
+      "step": 56140
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.2203906662260156e-05,
+      "loss": 0.6777,
+      "step": 56145
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.2188178538403213e-05,
+      "loss": 0.9449,
+      "step": 56150
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.21724552918799e-05,
+      "loss": 0.9514,
+      "step": 56155
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.2156736923675693e-05,
+      "loss": 0.8412,
+      "step": 56160
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.48828125,
+      "learning_rate": 2.2141023434775866e-05,
+      "loss": 0.9219,
+      "step": 56165
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.21253148261654e-05,
+      "loss": 0.8796,
+      "step": 56170
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.498046875,
+      "learning_rate": 2.2109611098828863e-05,
+      "loss": 0.9103,
+      "step": 56175
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.48828125,
+      "learning_rate": 2.209391225375064e-05,
+      "loss": 0.8296,
+      "step": 56180
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.207821829191472e-05,
+      "loss": 1.0703,
+      "step": 56185
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.2062529214304804e-05,
+      "loss": 1.0274,
+      "step": 56190
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.2046845021904316e-05,
+      "loss": 1.0084,
+      "step": 56195
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.203116571569639e-05,
+      "loss": 0.9722,
+      "step": 56200
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.201549129666377e-05,
+      "loss": 0.8741,
+      "step": 56205
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.1999821765788965e-05,
+      "loss": 1.1016,
+      "step": 56210
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.1984157124054207e-05,
+      "loss": 0.8361,
+      "step": 56215
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.1968497372441288e-05,
+      "loss": 0.8822,
+      "step": 56220
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.195284251193184e-05,
+      "loss": 0.9323,
+      "step": 56225
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.1937192543507136e-05,
+      "loss": 0.9843,
+      "step": 56230
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.1921547468148105e-05,
+      "loss": 0.8911,
+      "step": 56235
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.190590728683537e-05,
+      "loss": 1.0064,
+      "step": 56240
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.1890272000549317e-05,
+      "loss": 1.0378,
+      "step": 56245
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.1874641610270008e-05,
+      "loss": 0.9608,
+      "step": 56250
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.1859016116977106e-05,
+      "loss": 0.8337,
+      "step": 56255
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.1843395521650056e-05,
+      "loss": 0.9049,
+      "step": 56260
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.1827779825268036e-05,
+      "loss": 0.9015,
+      "step": 56265
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.181216902880977e-05,
+      "loss": 0.9828,
+      "step": 56270
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.179656313325379e-05,
+      "loss": 0.9996,
+      "step": 56275
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.178096213957834e-05,
+      "loss": 1.0021,
+      "step": 56280
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.176536604876126e-05,
+      "loss": 1.0434,
+      "step": 56285
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.1749774861780115e-05,
+      "loss": 1.0075,
+      "step": 56290
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.1734188579612225e-05,
+      "loss": 0.8832,
+      "step": 56295
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.171860720323451e-05,
+      "loss": 0.857,
+      "step": 56300
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.1703030733623642e-05,
+      "loss": 0.8859,
+      "step": 56305
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.1687459171756008e-05,
+      "loss": 0.8828,
+      "step": 56310
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.1671892518607607e-05,
+      "loss": 0.9719,
+      "step": 56315
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.1656330775154175e-05,
+      "loss": 0.9589,
+      "step": 56320
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.625,
+      "learning_rate": 2.1640773942371195e-05,
+      "loss": 1.0611,
+      "step": 56325
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.1625222021233714e-05,
+      "loss": 0.9626,
+      "step": 56330
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.1609675012716613e-05,
+      "loss": 0.9591,
+      "step": 56335
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.159413291779433e-05,
+      "loss": 0.918,
+      "step": 56340
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.157859573744112e-05,
+      "loss": 1.0512,
+      "step": 56345
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.1563063472630818e-05,
+      "loss": 1.0652,
+      "step": 56350
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5,
+      "learning_rate": 2.1547536124337032e-05,
+      "loss": 0.9323,
+      "step": 56355
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.153201369353306e-05,
+      "loss": 0.9307,
+      "step": 56360
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.15164961811918e-05,
+      "loss": 0.9425,
+      "step": 56365
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.150098358828595e-05,
+      "loss": 0.9132,
+      "step": 56370
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.148547591578788e-05,
+      "loss": 0.8788,
+      "step": 56375
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.1469973164669567e-05,
+      "loss": 0.8095,
+      "step": 56380
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.1454475335902778e-05,
+      "loss": 1.05,
+      "step": 56385
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.1438982430458986e-05,
+      "loss": 0.9308,
+      "step": 56390
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.14234944493092e-05,
+      "loss": 1.1513,
+      "step": 56395
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.1408011393424265e-05,
+      "loss": 1.1842,
+      "step": 56400
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.1392533263774716e-05,
+      "loss": 1.0473,
+      "step": 56405
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.1377060061330677e-05,
+      "loss": 1.0914,
+      "step": 56410
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.1361591787062064e-05,
+      "loss": 0.9997,
+      "step": 56415
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5,
+      "learning_rate": 2.1346128441938463e-05,
+      "loss": 0.8518,
+      "step": 56420
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.133067002692908e-05,
+      "loss": 0.9078,
+      "step": 56425
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.13152165430029e-05,
+      "loss": 1.0808,
+      "step": 56430
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.129976799112858e-05,
+      "loss": 0.9562,
+      "step": 56435
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.1284324372274454e-05,
+      "loss": 1.0333,
+      "step": 56440
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.1268885687408478e-05,
+      "loss": 0.9312,
+      "step": 56445
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.1253451937498426e-05,
+      "loss": 0.8903,
+      "step": 56450
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.123802312351172e-05,
+      "loss": 1.0886,
+      "step": 56455
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.12225992464154e-05,
+      "loss": 0.9818,
+      "step": 56460
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.1207180307176266e-05,
+      "loss": 0.9473,
+      "step": 56465
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.1191766306760852e-05,
+      "loss": 0.9522,
+      "step": 56470
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.1176357246135247e-05,
+      "loss": 0.819,
+      "step": 56475
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.1160953126265336e-05,
+      "loss": 1.0077,
+      "step": 56480
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5,
+      "learning_rate": 2.11455539481167e-05,
+      "loss": 0.9595,
+      "step": 56485
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.113015971265454e-05,
+      "loss": 0.9976,
+      "step": 56490
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.111477042084381e-05,
+      "loss": 1.0494,
+      "step": 56495
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.48828125,
+      "learning_rate": 2.1099386073649106e-05,
+      "loss": 0.9497,
+      "step": 56500
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.1084006672034727e-05,
+      "loss": 1.1865,
+      "step": 56505
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.106863221696468e-05,
+      "loss": 0.9653,
+      "step": 56510
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.10532627094027e-05,
+      "loss": 0.9337,
+      "step": 56515
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.1037898150312087e-05,
+      "loss": 0.8808,
+      "step": 56520
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.1022538540655955e-05,
+      "loss": 0.9955,
+      "step": 56525
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.1007183881397075e-05,
+      "loss": 1.0113,
+      "step": 56530
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.0991834173497848e-05,
+      "loss": 0.9069,
+      "step": 56535
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.0976489417920443e-05,
+      "loss": 0.8756,
+      "step": 56540
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.0961149615626706e-05,
+      "loss": 0.9911,
+      "step": 56545
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.094581476757813e-05,
+      "loss": 0.896,
+      "step": 56550
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.09304848747359e-05,
+      "loss": 0.87,
+      "step": 56555
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.48828125,
+      "learning_rate": 2.0915159938060926e-05,
+      "loss": 0.9616,
+      "step": 56560
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.0899839958513812e-05,
+      "loss": 0.8914,
+      "step": 56565
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.0884524937054805e-05,
+      "loss": 0.8992,
+      "step": 56570
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.086921487464387e-05,
+      "loss": 0.9412,
+      "step": 56575
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.08539097722407e-05,
+      "loss": 1.0388,
+      "step": 56580
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.0838609630804584e-05,
+      "loss": 1.0003,
+      "step": 56585
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.0823314451294563e-05,
+      "loss": 0.969,
+      "step": 56590
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.0808024234669398e-05,
+      "loss": 0.9585,
+      "step": 56595
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.0792738981887473e-05,
+      "loss": 0.9976,
+      "step": 56600
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.0777458693906837e-05,
+      "loss": 0.8886,
+      "step": 56605
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.0762183371685328e-05,
+      "loss": 1.0712,
+      "step": 56610
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.0746913016180435e-05,
+      "loss": 0.9209,
+      "step": 56615
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.0731647628349273e-05,
+      "loss": 0.9952,
+      "step": 56620
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.0716387209148737e-05,
+      "loss": 0.9298,
+      "step": 56625
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.070113175953532e-05,
+      "loss": 0.9394,
+      "step": 56630
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.068588128046528e-05,
+      "loss": 1.0072,
+      "step": 56635
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.0670635772894553e-05,
+      "loss": 0.9542,
+      "step": 56640
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.0655395237778708e-05,
+      "loss": 0.9364,
+      "step": 56645
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.064015967607308e-05,
+      "loss": 0.9768,
+      "step": 56650
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.0624929088732592e-05,
+      "loss": 1.021,
+      "step": 56655
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.0609703476711984e-05,
+      "loss": 1.0188,
+      "step": 56660
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.0594482840965547e-05,
+      "loss": 0.791,
+      "step": 56665
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.057926718244737e-05,
+      "loss": 1.0041,
+      "step": 56670
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.0564056502111195e-05,
+      "loss": 1.1466,
+      "step": 56675
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.0548850800910413e-05,
+      "loss": 1.0553,
+      "step": 56680
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.053365007979814e-05,
+      "loss": 0.9049,
+      "step": 56685
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.051845433972721e-05,
+      "loss": 0.8279,
+      "step": 56690
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.0503263581650067e-05,
+      "loss": 1.0301,
+      "step": 56695
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.0488077806518902e-05,
+      "loss": 1.1201,
+      "step": 56700
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.0472897015285597e-05,
+      "loss": 0.9476,
+      "step": 56705
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.0457721208901682e-05,
+      "loss": 0.9026,
+      "step": 56710
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.044255038831837e-05,
+      "loss": 0.8703,
+      "step": 56715
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.0427384554486595e-05,
+      "loss": 1.0137,
+      "step": 56720
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.0412223708357025e-05,
+      "loss": 0.9533,
+      "step": 56725
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.039706785087988e-05,
+      "loss": 0.9821,
+      "step": 56730
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.038191698300519e-05,
+      "loss": 0.9189,
+      "step": 56735
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.0366771105682637e-05,
+      "loss": 1.097,
+      "step": 56740
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.035163021986154e-05,
+      "loss": 0.8491,
+      "step": 56745
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.0336494326490985e-05,
+      "loss": 0.8807,
+      "step": 56750
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.0321363426519734e-05,
+      "loss": 0.9252,
+      "step": 56755
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.030623752089612e-05,
+      "loss": 0.9657,
+      "step": 56760
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.0291116610568304e-05,
+      "loss": 0.9351,
+      "step": 56765
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.027600069648411e-05,
+      "loss": 1.0192,
+      "step": 56770
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.0260889779590962e-05,
+      "loss": 0.8669,
+      "step": 56775
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.0245783860836053e-05,
+      "loss": 0.7708,
+      "step": 56780
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.023068294116628e-05,
+      "loss": 0.9595,
+      "step": 56785
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.0215587021528116e-05,
+      "loss": 0.996,
+      "step": 56790
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.0200496102867818e-05,
+      "loss": 0.9928,
+      "step": 56795
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.018541018613135e-05,
+      "loss": 0.956,
+      "step": 56800
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.017032927226423e-05,
+      "loss": 0.994,
+      "step": 56805
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.0155253362211822e-05,
+      "loss": 1.0376,
+      "step": 56810
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.0140182456919053e-05,
+      "loss": 1.0575,
+      "step": 56815
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.0125116557330615e-05,
+      "loss": 0.8795,
+      "step": 56820
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.0110055664390813e-05,
+      "loss": 0.876,
+      "step": 56825
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.009499977904372e-05,
+      "loss": 1.074,
+      "step": 56830
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.007994890223306e-05,
+      "loss": 1.0949,
+      "step": 56835
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.0064903034902206e-05,
+      "loss": 0.8713,
+      "step": 56840
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.0049862177994262e-05,
+      "loss": 0.8938,
+      "step": 56845
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.003482633245205e-05,
+      "loss": 0.9963,
+      "step": 56850
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.0019795499217953e-05,
+      "loss": 0.9014,
+      "step": 56855
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.0004769679234203e-05,
+      "loss": 0.8635,
+      "step": 56860
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.998974887344259e-05,
+      "loss": 1.0966,
+      "step": 56865
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.9974733082784623e-05,
+      "loss": 0.8467,
+      "step": 56870
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.995972230820152e-05,
+      "loss": 0.8615,
+      "step": 56875
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.9944716550634214e-05,
+      "loss": 0.8607,
+      "step": 56880
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.992971581102322e-05,
+      "loss": 1.0601,
+      "step": 56885
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9914720090308826e-05,
+      "loss": 0.981,
+      "step": 56890
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.9899729389431022e-05,
+      "loss": 0.9211,
+      "step": 56895
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.988474370932937e-05,
+      "loss": 0.9536,
+      "step": 56900
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.9869763050943234e-05,
+      "loss": 1.1827,
+      "step": 56905
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.985478741521163e-05,
+      "loss": 1.0555,
+      "step": 56910
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.983981680307323e-05,
+      "loss": 1.0235,
+      "step": 56915
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.9824851215466388e-05,
+      "loss": 0.9067,
+      "step": 56920
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.9809890653329178e-05,
+      "loss": 0.8933,
+      "step": 56925
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.979493511759938e-05,
+      "loss": 0.9839,
+      "step": 56930
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.9779984609214363e-05,
+      "loss": 0.934,
+      "step": 56935
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.9765039129111285e-05,
+      "loss": 1.0783,
+      "step": 56940
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.975009867822695e-05,
+      "loss": 0.9056,
+      "step": 56945
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.97351632574978e-05,
+      "loss": 0.9473,
+      "step": 56950
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.9720232867860033e-05,
+      "loss": 1.1167,
+      "step": 56955
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.970530751024954e-05,
+      "loss": 1.1607,
+      "step": 56960
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.96903871856018e-05,
+      "loss": 0.8377,
+      "step": 56965
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.967547189485204e-05,
+      "loss": 0.9362,
+      "step": 56970
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.9660561638935217e-05,
+      "loss": 0.9683,
+      "step": 56975
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.9645656418785852e-05,
+      "loss": 0.8118,
+      "step": 56980
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.9630756235338253e-05,
+      "loss": 0.824,
+      "step": 56985
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9615861089526422e-05,
+      "loss": 0.8573,
+      "step": 56990
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.9600970982283952e-05,
+      "loss": 0.9875,
+      "step": 56995
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.958608591454417e-05,
+      "loss": 0.9674,
+      "step": 57000
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.957120588724014e-05,
+      "loss": 0.9661,
+      "step": 57005
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.95563309013045e-05,
+      "loss": 0.8551,
+      "step": 57010
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.9541460957669656e-05,
+      "loss": 0.8241,
+      "step": 57015
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9526596057267697e-05,
+      "loss": 1.1419,
+      "step": 57020
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.951173620103035e-05,
+      "loss": 0.9015,
+      "step": 57025
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.9496881389889023e-05,
+      "loss": 0.8603,
+      "step": 57030
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.9482031624774855e-05,
+      "loss": 0.9216,
+      "step": 57035
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.9467186906618673e-05,
+      "loss": 0.9155,
+      "step": 57040
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.625,
+      "learning_rate": 1.94523472363509e-05,
+      "loss": 0.9163,
+      "step": 57045
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.9437512614901753e-05,
+      "loss": 0.95,
+      "step": 57050
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.9422683043201086e-05,
+      "loss": 0.854,
+      "step": 57055
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.940785852217839e-05,
+      "loss": 1.1324,
+      "step": 57060
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.9393039052762908e-05,
+      "loss": 0.9357,
+      "step": 57065
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.93782246358836e-05,
+      "loss": 1.1118,
+      "step": 57070
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9363415272468933e-05,
+      "loss": 1.1094,
+      "step": 57075
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.9348610963447235e-05,
+      "loss": 0.9506,
+      "step": 57080
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.9333811709746485e-05,
+      "loss": 0.9978,
+      "step": 57085
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.9319017512294257e-05,
+      "loss": 0.9095,
+      "step": 57090
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9304228372017908e-05,
+      "loss": 1.0865,
+      "step": 57095
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.9289444289844448e-05,
+      "loss": 0.9566,
+      "step": 57100
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.927466526670052e-05,
+      "loss": 0.9192,
+      "step": 57105
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9259891303512512e-05,
+      "loss": 1.0266,
+      "step": 57110
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.9245122401206493e-05,
+      "loss": 0.923,
+      "step": 57115
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.923035856070815e-05,
+      "loss": 0.9628,
+      "step": 57120
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.921559978294295e-05,
+      "loss": 1.0032,
+      "step": 57125
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.920084606883593e-05,
+      "loss": 1.112,
+      "step": 57130
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9186097419311932e-05,
+      "loss": 0.9425,
+      "step": 57135
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.917135383529537e-05,
+      "loss": 1.1636,
+      "step": 57140
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.915661531771039e-05,
+      "loss": 0.9695,
+      "step": 57145
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.914188186748087e-05,
+      "loss": 0.8936,
+      "step": 57150
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.9127153485530246e-05,
+      "loss": 1.08,
+      "step": 57155
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.911243017278176e-05,
+      "loss": 0.9949,
+      "step": 57160
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.9097711930158303e-05,
+      "loss": 0.9987,
+      "step": 57165
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.908299875858237e-05,
+      "loss": 0.8954,
+      "step": 57170
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9068290658976252e-05,
+      "loss": 1.0304,
+      "step": 57175
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.905358763226186e-05,
+      "loss": 0.8348,
+      "step": 57180
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.903888967936075e-05,
+      "loss": 0.9374,
+      "step": 57185
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.902419680119425e-05,
+      "loss": 0.998,
+      "step": 57190
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.9009508998683334e-05,
+      "loss": 0.9541,
+      "step": 57195
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.899482627274861e-05,
+      "loss": 0.9615,
+      "step": 57200
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.8980148624310444e-05,
+      "loss": 0.9093,
+      "step": 57205
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.8965476054288857e-05,
+      "loss": 0.8685,
+      "step": 57210
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.8950808563603485e-05,
+      "loss": 0.963,
+      "step": 57215
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.893614615317375e-05,
+      "loss": 0.8768,
+      "step": 57220
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.892148882391872e-05,
+      "loss": 1.043,
+      "step": 57225
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.8906836576757116e-05,
+      "loss": 0.9636,
+      "step": 57230
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.889218941260732e-05,
+      "loss": 1.0872,
+      "step": 57235
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.8877547332387467e-05,
+      "loss": 1.0531,
+      "step": 57240
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.886291033701537e-05,
+      "loss": 0.8688,
+      "step": 57245
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.8848278427408438e-05,
+      "loss": 1.0525,
+      "step": 57250
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.8833651604483828e-05,
+      "loss": 0.8974,
+      "step": 57255
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.88190298691584e-05,
+      "loss": 0.9064,
+      "step": 57260
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.8804413222348617e-05,
+      "loss": 1.0215,
+      "step": 57265
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.8789801664970686e-05,
+      "loss": 0.9896,
+      "step": 57270
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.877519519794051e-05,
+      "loss": 0.9436,
+      "step": 57275
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.87605938221736e-05,
+      "loss": 1.0005,
+      "step": 57280
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.874599753858517e-05,
+      "loss": 0.8156,
+      "step": 57285
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.8731406348090153e-05,
+      "loss": 1.0979,
+      "step": 57290
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.8716820251603163e-05,
+      "loss": 0.9682,
+      "step": 57295
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.8702239250038433e-05,
+      "loss": 0.917,
+      "step": 57300
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.8687663344309958e-05,
+      "loss": 1.0256,
+      "step": 57305
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.867309253533134e-05,
+      "loss": 0.8435,
+      "step": 57310
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.8658526824015886e-05,
+      "loss": 0.9523,
+      "step": 57315
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.8643966211276655e-05,
+      "loss": 1.0054,
+      "step": 57320
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.862941069802624e-05,
+      "loss": 1.081,
+      "step": 57325
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.8614860285177038e-05,
+      "loss": 0.965,
+      "step": 57330
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.860031497364112e-05,
+      "loss": 0.9633,
+      "step": 57335
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.8585774764330154e-05,
+      "loss": 0.9439,
+      "step": 57340
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.8571239658155526e-05,
+      "loss": 1.0464,
+      "step": 57345
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.855670965602834e-05,
+      "loss": 0.9419,
+      "step": 57350
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.8542184758859372e-05,
+      "loss": 0.9403,
+      "step": 57355
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.8527664967559e-05,
+      "loss": 0.8075,
+      "step": 57360
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.8513150283037395e-05,
+      "loss": 1.0399,
+      "step": 57365
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.849864070620435e-05,
+      "loss": 0.9282,
+      "step": 57370
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.848413623796931e-05,
+      "loss": 0.8879,
+      "step": 57375
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.8469636879241438e-05,
+      "loss": 1.0933,
+      "step": 57380
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.845514263092961e-05,
+      "loss": 0.9499,
+      "step": 57385
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.8440653493942316e-05,
+      "loss": 0.9202,
+      "step": 57390
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.8426169469187726e-05,
+      "loss": 0.9033,
+      "step": 57395
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.8411690557573747e-05,
+      "loss": 1.0041,
+      "step": 57400
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.839721676000794e-05,
+      "loss": 0.9258,
+      "step": 57405
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.8382748077397494e-05,
+      "loss": 0.9638,
+      "step": 57410
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.8368284510649358e-05,
+      "loss": 0.8681,
+      "step": 57415
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.8353826060670153e-05,
+      "loss": 1.0304,
+      "step": 57420
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.8339372728366077e-05,
+      "loss": 0.7835,
+      "step": 57425
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.8324924514643138e-05,
+      "loss": 0.8806,
+      "step": 57430
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.8310481420406967e-05,
+      "loss": 0.9688,
+      "step": 57435
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.829604344656286e-05,
+      "loss": 1.1028,
+      "step": 57440
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.8281610594015775e-05,
+      "loss": 0.8831,
+      "step": 57445
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.826718286367043e-05,
+      "loss": 1.0611,
+      "step": 57450
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.8252760256431123e-05,
+      "loss": 0.9923,
+      "step": 57455
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.823834277320191e-05,
+      "loss": 0.8988,
+      "step": 57460
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.8223930414886515e-05,
+      "loss": 1.0085,
+      "step": 57465
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.8209523182388276e-05,
+      "loss": 0.9161,
+      "step": 57470
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.8195121076610266e-05,
+      "loss": 1.0192,
+      "step": 57475
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.818072409845527e-05,
+      "loss": 0.8755,
+      "step": 57480
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.8166332248825645e-05,
+      "loss": 0.8561,
+      "step": 57485
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.8151945528623536e-05,
+      "loss": 1.0005,
+      "step": 57490
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.8137563938750667e-05,
+      "loss": 0.9342,
+      "step": 57495
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.812318748010856e-05,
+      "loss": 0.9835,
+      "step": 57500
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.8108816153598284e-05,
+      "loss": 1.023,
+      "step": 57505
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.8094449960120673e-05,
+      "loss": 0.9161,
+      "step": 57510
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.808008890057625e-05,
+      "loss": 0.8931,
+      "step": 57515
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.8065732975865134e-05,
+      "loss": 0.8917,
+      "step": 57520
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.805138218688719e-05,
+      "loss": 1.1648,
+      "step": 57525
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.8037036534541963e-05,
+      "loss": 0.8815,
+      "step": 57530
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.8022696019728602e-05,
+      "loss": 1.0268,
+      "step": 57535
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.8008360643346022e-05,
+      "loss": 0.901,
+      "step": 57540
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.7994030406292837e-05,
+      "loss": 1.0224,
+      "step": 57545
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.7979705309467164e-05,
+      "loss": 1.0273,
+      "step": 57550
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.7965385353766985e-05,
+      "loss": 0.8829,
+      "step": 57555
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.7951070540089898e-05,
+      "loss": 1.1264,
+      "step": 57560
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.7936760869333137e-05,
+      "loss": 0.9873,
+      "step": 57565
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.7922456342393658e-05,
+      "loss": 0.9088,
+      "step": 57570
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.790815696016812e-05,
+      "loss": 0.9321,
+      "step": 57575
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.789386272355278e-05,
+      "loss": 1.0373,
+      "step": 57580
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.7879573633443625e-05,
+      "loss": 1.0655,
+      "step": 57585
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.7865289690736364e-05,
+      "loss": 0.8181,
+      "step": 57590
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.785101089632627e-05,
+      "loss": 0.8864,
+      "step": 57595
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.783673725110836e-05,
+      "loss": 0.9482,
+      "step": 57600
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.782246875597733e-05,
+      "loss": 0.834,
+      "step": 57605
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.7808205411827582e-05,
+      "loss": 0.8415,
+      "step": 57610
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.77939472195531e-05,
+      "loss": 0.8928,
+      "step": 57615
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.7779694180047623e-05,
+      "loss": 0.9688,
+      "step": 57620
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5,
+      "learning_rate": 1.7765446294204592e-05,
+      "loss": 1.0195,
+      "step": 57625
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.7751203562917018e-05,
+      "loss": 0.9816,
+      "step": 57630
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5,
+      "learning_rate": 1.773696598707767e-05,
+      "loss": 0.9331,
+      "step": 57635
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.772273356757902e-05,
+      "loss": 0.8634,
+      "step": 57640
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.77085063053131e-05,
+      "loss": 0.9065,
+      "step": 57645
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.7694284201171752e-05,
+      "loss": 0.9567,
+      "step": 57650
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.768006725604642e-05,
+      "loss": 1.1036,
+      "step": 57655
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.7665855470828197e-05,
+      "loss": 1.0417,
+      "step": 57660
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.625,
+      "learning_rate": 1.765164884640792e-05,
+      "loss": 0.9475,
+      "step": 57665
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.763744738367611e-05,
+      "loss": 0.8683,
+      "step": 57670
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.7623251083522863e-05,
+      "loss": 1.0274,
+      "step": 57675
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.760905994683807e-05,
+      "loss": 0.9602,
+      "step": 57680
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.7594873974511263e-05,
+      "loss": 0.9219,
+      "step": 57685
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.7580693167431573e-05,
+      "loss": 0.9376,
+      "step": 57690
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.7566517526487903e-05,
+      "loss": 1.0304,
+      "step": 57695
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.755234705256883e-05,
+      "loss": 0.9291,
+      "step": 57700
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.7538181746562543e-05,
+      "loss": 1.069,
+      "step": 57705
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.752402160935691e-05,
+      "loss": 1.0669,
+      "step": 57710
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.7509866641839534e-05,
+      "loss": 1.0158,
+      "step": 57715
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.7495716844897692e-05,
+      "loss": 0.8566,
+      "step": 57720
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.7481572219418263e-05,
+      "loss": 0.9092,
+      "step": 57725
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.746743276628786e-05,
+      "loss": 0.9913,
+      "step": 57730
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.74532984863928e-05,
+      "loss": 0.8428,
+      "step": 57735
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.743916938061898e-05,
+      "loss": 0.8582,
+      "step": 57740
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.7425045449852053e-05,
+      "loss": 1.0616,
+      "step": 57745
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.7410926694977337e-05,
+      "loss": 0.8835,
+      "step": 57750
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.7396813116879794e-05,
+      "loss": 1.0282,
+      "step": 57755
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.7382704716444075e-05,
+      "loss": 0.9069,
+      "step": 57760
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.7368601494554526e-05,
+      "loss": 0.9493,
+      "step": 57765
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.7354503452095128e-05,
+      "loss": 0.9655,
+      "step": 57770
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.7340410589949572e-05,
+      "loss": 0.9934,
+      "step": 57775
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.7326322909001258e-05,
+      "loss": 0.9102,
+      "step": 57780
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.7312240410133153e-05,
+      "loss": 1.1948,
+      "step": 57785
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.7298163094227982e-05,
+      "loss": 0.8848,
+      "step": 57790
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.7284090962168176e-05,
+      "loss": 0.8271,
+      "step": 57795
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.7270024014835716e-05,
+      "loss": 1.039,
+      "step": 57800
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.7255962253112413e-05,
+      "loss": 0.8683,
+      "step": 57805
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.72419056778796e-05,
+      "loss": 1.0126,
+      "step": 57810
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.722785429001842e-05,
+      "loss": 0.933,
+      "step": 57815
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.7213808090409566e-05,
+      "loss": 0.963,
+      "step": 57820
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.7199767079933527e-05,
+      "loss": 0.8474,
+      "step": 57825
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.7185731259470395e-05,
+      "loss": 1.1649,
+      "step": 57830
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.7171700629899934e-05,
+      "loss": 0.9088,
+      "step": 57835
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.715767519210161e-05,
+      "loss": 1.0726,
+      "step": 57840
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.714365494695457e-05,
+      "loss": 1.0465,
+      "step": 57845
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.712963989533758e-05,
+      "loss": 0.9767,
+      "step": 57850
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.711563003812915e-05,
+      "loss": 0.9515,
+      "step": 57855
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.7101625376207465e-05,
+      "loss": 0.9817,
+      "step": 57860
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.7087625910450277e-05,
+      "loss": 1.0111,
+      "step": 57865
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.7073631641735122e-05,
+      "loss": 0.9127,
+      "step": 57870
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.7059642570939204e-05,
+      "loss": 0.89,
+      "step": 57875
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.7045658698939327e-05,
+      "loss": 0.9156,
+      "step": 57880
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.703168002661204e-05,
+      "loss": 0.839,
+      "step": 57885
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.7017706554833568e-05,
+      "loss": 0.8972,
+      "step": 57890
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.700373828447973e-05,
+      "loss": 0.909,
+      "step": 57895
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.6989775216426106e-05,
+      "loss": 1.077,
+      "step": 57900
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.697581735154793e-05,
+      "loss": 1.0366,
+      "step": 57905
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.6961864690720087e-05,
+      "loss": 1.0014,
+      "step": 57910
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.6947917234817114e-05,
+      "loss": 1.0623,
+      "step": 57915
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.6933974984713263e-05,
+      "loss": 0.9165,
+      "step": 57920
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.692003794128251e-05,
+      "loss": 0.8876,
+      "step": 57925
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.6906106105398356e-05,
+      "loss": 0.8691,
+      "step": 57930
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.6892179477934112e-05,
+      "loss": 0.9261,
+      "step": 57935
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.687825805976274e-05,
+      "loss": 0.9677,
+      "step": 57940
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.686434185175679e-05,
+      "loss": 0.9245,
+      "step": 57945
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.685043085478858e-05,
+      "loss": 0.9576,
+      "step": 57950
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.683652506973008e-05,
+      "loss": 0.9034,
+      "step": 57955
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.6822624497452888e-05,
+      "loss": 0.967,
+      "step": 57960
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.680872913882835e-05,
+      "loss": 0.9726,
+      "step": 57965
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.679483899472739e-05,
+      "loss": 1.0131,
+      "step": 57970
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.6780954066020704e-05,
+      "loss": 1.1013,
+      "step": 57975
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.6767074353578572e-05,
+      "loss": 0.9656,
+      "step": 57980
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.675319985827104e-05,
+      "loss": 0.8297,
+      "step": 57985
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.6739330580967728e-05,
+      "loss": 0.969,
+      "step": 57990
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.6725466522538e-05,
+      "loss": 0.8789,
+      "step": 57995
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.6711607683850905e-05,
+      "loss": 0.8702,
+      "step": 58000
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.669775406577506e-05,
+      "loss": 0.9958,
+      "step": 58005
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.6683905669178866e-05,
+      "loss": 1.0337,
+      "step": 58010
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.667006249493038e-05,
+      "loss": 0.9877,
+      "step": 58015
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.665622454389729e-05,
+      "loss": 1.002,
+      "step": 58020
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.6642391816946934e-05,
+      "loss": 0.9392,
+      "step": 58025
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.6628564314946393e-05,
+      "loss": 0.8782,
+      "step": 58030
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.661474203876242e-05,
+      "loss": 0.9426,
+      "step": 58035
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.6600924989261368e-05,
+      "loss": 1.0012,
+      "step": 58040
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.658711316730931e-05,
+      "loss": 1.0189,
+      "step": 58045
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.657330657377202e-05,
+      "loss": 0.9001,
+      "step": 58050
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.655950520951487e-05,
+      "loss": 0.9475,
+      "step": 58055
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.6545709075402972e-05,
+      "loss": 0.8767,
+      "step": 58060
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.6531918172301087e-05,
+      "loss": 0.8723,
+      "step": 58065
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.6518132501073634e-05,
+      "loss": 1.0125,
+      "step": 58070
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.6504352062584692e-05,
+      "loss": 0.9769,
+      "step": 58075
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.649057685769806e-05,
+      "loss": 0.991,
+      "step": 58080
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.6476806887277208e-05,
+      "loss": 0.9772,
+      "step": 58085
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.6463042152185193e-05,
+      "loss": 1.0976,
+      "step": 58090
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.6449282653284836e-05,
+      "loss": 1.1015,
+      "step": 58095
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.6435528391438626e-05,
+      "loss": 0.9803,
+      "step": 58100
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.642177936750865e-05,
+      "loss": 1.1817,
+      "step": 58105
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.640803558235672e-05,
+      "loss": 0.9719,
+      "step": 58110
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.6394297036844353e-05,
+      "loss": 1.0251,
+      "step": 58115
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5,
+      "learning_rate": 1.6380563731832664e-05,
+      "loss": 1.0194,
+      "step": 58120
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.6366835668182455e-05,
+      "loss": 1.0195,
+      "step": 58125
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.6353112846754247e-05,
+      "loss": 0.8529,
+      "step": 58130
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.6339395268408188e-05,
+      "loss": 0.9441,
+      "step": 58135
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.6325682934004104e-05,
+      "loss": 0.9558,
+      "step": 58140
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.6311975844401528e-05,
+      "loss": 0.9949,
+      "step": 58145
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.6298274000459612e-05,
+      "loss": 0.9409,
+      "step": 58150
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.6284577403037193e-05,
+      "loss": 0.9242,
+      "step": 58155
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.627088605299284e-05,
+      "loss": 1.1643,
+      "step": 58160
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.6257199951184686e-05,
+      "loss": 1.0059,
+      "step": 58165
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.6243519098470606e-05,
+      "loss": 0.9065,
+      "step": 58170
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.6229843495708163e-05,
+      "loss": 0.9551,
+      "step": 58175
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.6216173143754544e-05,
+      "loss": 1.0303,
+      "step": 58180
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.620250804346659e-05,
+      "loss": 0.7746,
+      "step": 58185
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.618884819570087e-05,
+      "loss": 0.8097,
+      "step": 58190
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.617519360131361e-05,
+      "loss": 1.0209,
+      "step": 58195
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.6161544261160676e-05,
+      "loss": 0.8574,
+      "step": 58200
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5,
+      "learning_rate": 1.614790017609762e-05,
+      "loss": 0.9302,
+      "step": 58205
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.6134261346979707e-05,
+      "loss": 1.0741,
+      "step": 58210
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.6120627774661788e-05,
+      "loss": 0.9656,
+      "step": 58215
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.610699945999844e-05,
+      "loss": 0.8366,
+      "step": 58220
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.6093376403843973e-05,
+      "loss": 0.9303,
+      "step": 58225
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.6079758607052176e-05,
+      "loss": 0.9636,
+      "step": 58230
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.6066146070476696e-05,
+      "loss": 0.9858,
+      "step": 58235
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.6052538794970795e-05,
+      "loss": 0.8403,
+      "step": 58240
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.6038936781387337e-05,
+      "loss": 0.9222,
+      "step": 58245
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.602534003057895e-05,
+      "loss": 0.7898,
+      "step": 58250
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.601174854339792e-05,
+      "loss": 1.1416,
+      "step": 58255
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.5998162320696118e-05,
+      "loss": 0.9055,
+      "step": 58260
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.5984581363325168e-05,
+      "loss": 0.9214,
+      "step": 58265
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.5971005672136375e-05,
+      "loss": 0.9064,
+      "step": 58270
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.595743524798061e-05,
+      "loss": 1.0478,
+      "step": 58275
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.5943870091708558e-05,
+      "loss": 1.0135,
+      "step": 58280
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.5930310204170427e-05,
+      "loss": 0.9129,
+      "step": 58285
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.5916755586216236e-05,
+      "loss": 0.782,
+      "step": 58290
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.5903206238695535e-05,
+      "loss": 0.8294,
+      "step": 58295
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.588966216245764e-05,
+      "loss": 0.9915,
+      "step": 58300
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.5876123358351547e-05,
+      "loss": 0.9133,
+      "step": 58305
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.5862589827225827e-05,
+      "loss": 1.0046,
+      "step": 58310
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.5849061569928802e-05,
+      "loss": 1.0208,
+      "step": 58315
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.5835538587308453e-05,
+      "loss": 0.836,
+      "step": 58320
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.5822020880212383e-05,
+      "loss": 0.9109,
+      "step": 58325
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.5808508449487935e-05,
+      "loss": 0.8614,
+      "step": 58330
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.579500129598207e-05,
+      "loss": 0.9432,
+      "step": 58335
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.57814994205414e-05,
+      "loss": 0.956,
+      "step": 58340
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.5768002824012263e-05,
+      "loss": 1.0127,
+      "step": 58345
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.5754511507240666e-05,
+      "loss": 0.9024,
+      "step": 58350
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.5741025471072202e-05,
+      "loss": 1.0308,
+      "step": 58355
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.5727544716352228e-05,
+      "loss": 1.0369,
+      "step": 58360
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.5714069243925754e-05,
+      "loss": 0.8761,
+      "step": 58365
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.57005990546374e-05,
+      "loss": 0.9829,
+      "step": 58370
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.5687134149331496e-05,
+      "loss": 0.9662,
+      "step": 58375
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.5673674528852065e-05,
+      "loss": 1.0646,
+      "step": 58380
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.5660220194042774e-05,
+      "loss": 0.855,
+      "step": 58385
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.5646771145746897e-05,
+      "loss": 0.9553,
+      "step": 58390
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.5633327384807473e-05,
+      "loss": 0.9051,
+      "step": 58395
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.5619888912067205e-05,
+      "loss": 1.0056,
+      "step": 58400
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.5606455728368375e-05,
+      "loss": 0.9086,
+      "step": 58405
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.559302783455302e-05,
+      "loss": 0.9692,
+      "step": 58410
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.5579605231462835e-05,
+      "loss": 1.0364,
+      "step": 58415
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.5566187919939114e-05,
+      "loss": 0.8547,
+      "step": 58420
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.555277590082289e-05,
+      "loss": 0.9497,
+      "step": 58425
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.5539369174954887e-05,
+      "loss": 0.955,
+      "step": 58430
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.552596774317541e-05,
+      "loss": 0.9568,
+      "step": 58435
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.5512571606324454e-05,
+      "loss": 0.8585,
+      "step": 58440
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.549918076524176e-05,
+      "loss": 0.9049,
+      "step": 58445
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.5485795220766642e-05,
+      "loss": 0.9563,
+      "step": 58450
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.5472414973738115e-05,
+      "loss": 0.9304,
+      "step": 58455
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.545904002499492e-05,
+      "loss": 0.9151,
+      "step": 58460
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.5445670375375354e-05,
+      "loss": 0.8703,
+      "step": 58465
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.543230602571747e-05,
+      "loss": 0.9275,
+      "step": 58470
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.5418946976858983e-05,
+      "loss": 0.8973,
+      "step": 58475
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.5405593229637206e-05,
+      "loss": 0.9844,
+      "step": 58480
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.5392244784889188e-05,
+      "loss": 0.8972,
+      "step": 58485
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.537890164345165e-05,
+      "loss": 0.8179,
+      "step": 58490
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.536556380616093e-05,
+      "loss": 0.9699,
+      "step": 58495
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.5352231273853046e-05,
+      "loss": 0.9691,
+      "step": 58500
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.53389040473637e-05,
+      "loss": 0.9526,
+      "step": 58505
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.5325582127528303e-05,
+      "loss": 0.9293,
+      "step": 58510
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.5312265515181824e-05,
+      "loss": 0.9441,
+      "step": 58515
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.5298954211159e-05,
+      "loss": 0.9324,
+      "step": 58520
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.5285648216294213e-05,
+      "loss": 0.8834,
+      "step": 58525
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.527234753142145e-05,
+      "loss": 0.9087,
+      "step": 58530
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.5259052157374442e-05,
+      "loss": 0.9629,
+      "step": 58535
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.5245762094986581e-05,
+      "loss": 0.9918,
+      "step": 58540
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.523247734509089e-05,
+      "loss": 1.027,
+      "step": 58545
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.5219197908520023e-05,
+      "loss": 1.0379,
+      "step": 58550
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.5205923786106414e-05,
+      "loss": 1.0219,
+      "step": 58555
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.5192654978682052e-05,
+      "loss": 1.1855,
+      "step": 58560
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.5179391487078664e-05,
+      "loss": 0.8573,
+      "step": 58565
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.516613331212765e-05,
+      "loss": 1.0287,
+      "step": 58570
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.515288045465999e-05,
+      "loss": 0.9057,
+      "step": 58575
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.5139632915506407e-05,
+      "loss": 0.9238,
+      "step": 58580
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.5126390695497317e-05,
+      "loss": 0.8945,
+      "step": 58585
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.5113153795462687e-05,
+      "loss": 0.9769,
+      "step": 58590
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.5099922216232288e-05,
+      "loss": 0.8387,
+      "step": 58595
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.5086695958635432e-05,
+      "loss": 0.8441,
+      "step": 58600
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.5073475023501204e-05,
+      "loss": 1.0256,
+      "step": 58605
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.5060259411658261e-05,
+      "loss": 1.1083,
+      "step": 58610
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.5047049123934987e-05,
+      "loss": 1.0134,
+      "step": 58615
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.503384416115946e-05,
+      "loss": 0.8517,
+      "step": 58620
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.5020644524159333e-05,
+      "loss": 0.8647,
+      "step": 58625
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.5007450213761976e-05,
+      "loss": 0.953,
+      "step": 58630
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.499426123079447e-05,
+      "loss": 1.1321,
+      "step": 58635
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.4981077576083457e-05,
+      "loss": 1.0327,
+      "step": 58640
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5,
+      "learning_rate": 1.496789925045533e-05,
+      "loss": 1.0104,
+      "step": 58645
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.4954726254736174e-05,
+      "loss": 0.9281,
+      "step": 58650
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.4941558589751581e-05,
+      "loss": 0.9496,
+      "step": 58655
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.4928396256326971e-05,
+      "loss": 0.9333,
+      "step": 58660
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.4915239255287395e-05,
+      "loss": 0.9894,
+      "step": 58665
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.4902087587457502e-05,
+      "loss": 0.9979,
+      "step": 58670
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.4888941253661681e-05,
+      "loss": 0.957,
+      "step": 58675
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.4875800254723982e-05,
+      "loss": 0.9186,
+      "step": 58680
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.4862664591468034e-05,
+      "loss": 0.7844,
+      "step": 58685
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.4849534264717246e-05,
+      "loss": 0.8935,
+      "step": 58690
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.483640927529465e-05,
+      "loss": 1.0858,
+      "step": 58695
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.4823289624022907e-05,
+      "loss": 1.0458,
+      "step": 58700
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.4810175311724361e-05,
+      "loss": 1.0108,
+      "step": 58705
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.4797066339221044e-05,
+      "loss": 0.938,
+      "step": 58710
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.4783962707334675e-05,
+      "loss": 1.0258,
+      "step": 58715
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.4770864416886555e-05,
+      "loss": 0.9121,
+      "step": 58720
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5,
+      "learning_rate": 1.4757771468697713e-05,
+      "loss": 1.0701,
+      "step": 58725
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.474468386358887e-05,
+      "loss": 0.9293,
+      "step": 58730
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.4731601602380307e-05,
+      "loss": 0.9501,
+      "step": 58735
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.4718524685892077e-05,
+      "loss": 0.9337,
+      "step": 58740
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.4705453114943868e-05,
+      "loss": 0.9834,
+      "step": 58745
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.4692386890354981e-05,
+      "loss": 0.9413,
+      "step": 58750
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.467932601294446e-05,
+      "loss": 1.1198,
+      "step": 58755
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.466627048353093e-05,
+      "loss": 0.9578,
+      "step": 58760
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.465322030293278e-05,
+      "loss": 0.9431,
+      "step": 58765
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.4640175471967965e-05,
+      "loss": 0.964,
+      "step": 58770
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.4627135991454155e-05,
+      "loss": 0.9486,
+      "step": 58775
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.4614101862208729e-05,
+      "loss": 1.0664,
+      "step": 58780
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.460107308504861e-05,
+      "loss": 0.9023,
+      "step": 58785
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.4588049660790527e-05,
+      "loss": 1.0412,
+      "step": 58790
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.4575031590250732e-05,
+      "loss": 0.9061,
+      "step": 58795
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.4562018874245254e-05,
+      "loss": 0.939,
+      "step": 58800
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.4549011513589762e-05,
+      "loss": 0.8717,
+      "step": 58805
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.4536009509099547e-05,
+      "loss": 0.9185,
+      "step": 58810
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.4523012861589568e-05,
+      "loss": 0.9918,
+      "step": 58815
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.4510021571874498e-05,
+      "loss": 0.9828,
+      "step": 58820
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.4497035640768664e-05,
+      "loss": 0.8807,
+      "step": 58825
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.4484055069085989e-05,
+      "loss": 0.9587,
+      "step": 58830
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.4471079857640134e-05,
+      "loss": 0.9532,
+      "step": 58835
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.445811000724443e-05,
+      "loss": 0.8421,
+      "step": 58840
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.4445145518711789e-05,
+      "loss": 0.9329,
+      "step": 58845
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.4432186392854862e-05,
+      "loss": 0.9504,
+      "step": 58850
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.441923263048598e-05,
+      "loss": 0.8915,
+      "step": 58855
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.4406284232417056e-05,
+      "loss": 1.0894,
+      "step": 58860
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.4393341199459698e-05,
+      "loss": 0.9599,
+      "step": 58865
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.4380403532425212e-05,
+      "loss": 0.8874,
+      "step": 58870
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.4367471232124575e-05,
+      "loss": 1.1097,
+      "step": 58875
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.4354544299368333e-05,
+      "loss": 0.9817,
+      "step": 58880
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.4341622734966797e-05,
+      "loss": 0.9076,
+      "step": 58885
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.4328706539729941e-05,
+      "loss": 0.8973,
+      "step": 58890
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.4315795714467294e-05,
+      "loss": 1.0058,
+      "step": 58895
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.4302890259988167e-05,
+      "loss": 0.8475,
+      "step": 58900
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.4289990177101497e-05,
+      "loss": 1.0302,
+      "step": 58905
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.4277095466615864e-05,
+      "loss": 0.9015,
+      "step": 58910
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.4264206129339486e-05,
+      "loss": 0.984,
+      "step": 58915
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.4251322166080339e-05,
+      "loss": 1.0828,
+      "step": 58920
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.4238443577645955e-05,
+      "loss": 0.8563,
+      "step": 58925
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4225570364843599e-05,
+      "loss": 1.044,
+      "step": 58930
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.4212702528480214e-05,
+      "loss": 1.1616,
+      "step": 58935
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.4199840069362302e-05,
+      "loss": 0.9844,
+      "step": 58940
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.418698298829615e-05,
+      "loss": 0.9174,
+      "step": 58945
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.4174131286087645e-05,
+      "loss": 1.0788,
+      "step": 58950
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.4161284963542321e-05,
+      "loss": 1.0316,
+      "step": 58955
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.4148444021465413e-05,
+      "loss": 0.859,
+      "step": 58960
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.625,
+      "learning_rate": 1.4135608460661842e-05,
+      "loss": 0.9537,
+      "step": 58965
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5,
+      "learning_rate": 1.412277828193611e-05,
+      "loss": 0.9823,
+      "step": 58970
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.410995348609243e-05,
+      "loss": 0.8578,
+      "step": 58975
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.4097134073934693e-05,
+      "loss": 1.0997,
+      "step": 58980
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.408432004626643e-05,
+      "loss": 1.0181,
+      "step": 58985
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.4071511403890824e-05,
+      "loss": 1.0344,
+      "step": 58990
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.4058708147610755e-05,
+      "loss": 0.9378,
+      "step": 58995
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.4045910278228757e-05,
+      "loss": 0.7372,
+      "step": 59000
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.4033117796546969e-05,
+      "loss": 0.9252,
+      "step": 59005
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.4020330703367268e-05,
+      "loss": 0.9874,
+      "step": 59010
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.4007548999491204e-05,
+      "loss": 0.9918,
+      "step": 59015
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.399477268571986e-05,
+      "loss": 0.9145,
+      "step": 59020
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.3982001762854114e-05,
+      "loss": 0.9774,
+      "step": 59025
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.3969236231694494e-05,
+      "loss": 0.9343,
+      "step": 59030
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.3956476093041105e-05,
+      "loss": 0.8558,
+      "step": 59035
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.3943721347693783e-05,
+      "loss": 0.8362,
+      "step": 59040
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.3930971996452048e-05,
+      "loss": 0.8882,
+      "step": 59045
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3918228040114977e-05,
+      "loss": 0.8783,
+      "step": 59050
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.3905489479481425e-05,
+      "loss": 0.8499,
+      "step": 59055
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.3892756315349853e-05,
+      "loss": 0.9287,
+      "step": 59060
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.3880028548518376e-05,
+      "loss": 1.0537,
+      "step": 59065
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.3867306179784811e-05,
+      "loss": 1.0329,
+      "step": 59070
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.3854589209946567e-05,
+      "loss": 0.9988,
+      "step": 59075
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.3841877639800805e-05,
+      "loss": 0.9891,
+      "step": 59080
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.3829171470144254e-05,
+      "loss": 0.893,
+      "step": 59085
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.3816470701773377e-05,
+      "loss": 1.0603,
+      "step": 59090
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.3803775335484292e-05,
+      "loss": 0.7809,
+      "step": 59095
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.379108537207272e-05,
+      "loss": 0.9883,
+      "step": 59100
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.3778400812334102e-05,
+      "loss": 1.0164,
+      "step": 59105
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.3765721657063535e-05,
+      "loss": 0.9677,
+      "step": 59110
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.3753047907055727e-05,
+      "loss": 0.9704,
+      "step": 59115
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.3740379563105122e-05,
+      "loss": 0.8938,
+      "step": 59120
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.372771662600577e-05,
+      "loss": 1.0333,
+      "step": 59125
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.3715059096551386e-05,
+      "loss": 0.9952,
+      "step": 59130
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.3702406975535353e-05,
+      "loss": 0.8836,
+      "step": 59135
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.3689760263750762e-05,
+      "loss": 0.8332,
+      "step": 59140
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.3677118961990276e-05,
+      "loss": 0.9347,
+      "step": 59145
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5,
+      "learning_rate": 1.3664483071046286e-05,
+      "loss": 0.8312,
+      "step": 59150
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.3651852591710845e-05,
+      "loss": 0.9411,
+      "step": 59155
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.363922752477561e-05,
+      "loss": 0.9272,
+      "step": 59160
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.3626607871031939e-05,
+      "loss": 0.9742,
+      "step": 59165
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.3613993631270894e-05,
+      "loss": 0.9807,
+      "step": 59170
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.3601384806283101e-05,
+      "loss": 0.7687,
+      "step": 59175
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.3588781396858896e-05,
+      "loss": 1.0959,
+      "step": 59180
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.3576183403788279e-05,
+      "loss": 1.0129,
+      "step": 59185
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.3563590827860939e-05,
+      "loss": 0.9644,
+      "step": 59190
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.3551003669866146e-05,
+      "loss": 0.9295,
+      "step": 59195
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.3538421930592903e-05,
+      "loss": 0.9419,
+      "step": 59200
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.3525845610829868e-05,
+      "loss": 1.0253,
+      "step": 59205
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.3513274711365298e-05,
+      "loss": 0.8688,
+      "step": 59210
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.3500709232987163e-05,
+      "loss": 0.9537,
+      "step": 59215
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.3488149176483112e-05,
+      "loss": 0.9867,
+      "step": 59220
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.3475594542640391e-05,
+      "loss": 0.9123,
+      "step": 59225
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5,
+      "learning_rate": 1.3463045332245939e-05,
+      "loss": 1.1345,
+      "step": 59230
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.345050154608637e-05,
+      "loss": 0.9323,
+      "step": 59235
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.343796318494792e-05,
+      "loss": 0.78,
+      "step": 59240
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.3425430249616522e-05,
+      "loss": 0.9499,
+      "step": 59245
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.3412902740877786e-05,
+      "loss": 0.9186,
+      "step": 59250
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.3400380659516898e-05,
+      "loss": 1.0809,
+      "step": 59255
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.3387864006318773e-05,
+      "loss": 0.9911,
+      "step": 59260
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.3375352782067996e-05,
+      "loss": 0.9153,
+      "step": 59265
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.3362846987548738e-05,
+      "loss": 0.9067,
+      "step": 59270
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.3350346623544918e-05,
+      "loss": 0.9464,
+      "step": 59275
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.3337851690840064e-05,
+      "loss": 0.998,
+      "step": 59280
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.3325362190217371e-05,
+      "loss": 0.8617,
+      "step": 59285
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.331287812245967e-05,
+      "loss": 0.9716,
+      "step": 59290
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.3300399488349491e-05,
+      "loss": 0.8445,
+      "step": 59295
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.328792628866905e-05,
+      "loss": 0.9536,
+      "step": 59300
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.3275458524200112e-05,
+      "loss": 0.9947,
+      "step": 59305
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.3262996195724209e-05,
+      "loss": 1.0093,
+      "step": 59310
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.3250539304022503e-05,
+      "loss": 0.9308,
+      "step": 59315
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.323808784987578e-05,
+      "loss": 0.9925,
+      "step": 59320
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.3225641834064529e-05,
+      "loss": 1.017,
+      "step": 59325
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.3213201257368913e-05,
+      "loss": 0.7214,
+      "step": 59330
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.320076612056863e-05,
+      "loss": 0.9199,
+      "step": 59335
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.3188336424443193e-05,
+      "loss": 0.9226,
+      "step": 59340
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.317591216977172e-05,
+      "loss": 0.8979,
+      "step": 59345
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.3163493357332934e-05,
+      "loss": 1.0728,
+      "step": 59350
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.315107998790529e-05,
+      "loss": 0.7297,
+      "step": 59355
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.3138672062266876e-05,
+      "loss": 0.9474,
+      "step": 59360
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.3126269581195416e-05,
+      "loss": 0.9479,
+      "step": 59365
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.3113872545468308e-05,
+      "loss": 0.8623,
+      "step": 59370
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3101480955862644e-05,
+      "loss": 0.8414,
+      "step": 59375
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.3089094813155112e-05,
+      "loss": 0.9636,
+      "step": 59380
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.3076714118122114e-05,
+      "loss": 0.9034,
+      "step": 59385
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.306433887153965e-05,
+      "loss": 0.8748,
+      "step": 59390
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.3051969074183457e-05,
+      "loss": 1.0434,
+      "step": 59395
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.3039604726828847e-05,
+      "loss": 0.9884,
+      "step": 59400
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.3027245830250856e-05,
+      "loss": 0.8497,
+      "step": 59405
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.3014892385224175e-05,
+      "loss": 0.9093,
+      "step": 59410
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.3002544392523076e-05,
+      "loss": 1.004,
+      "step": 59415
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.299020185292158e-05,
+      "loss": 0.9488,
+      "step": 59420
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.297786476719336e-05,
+      "loss": 0.7838,
+      "step": 59425
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.2965533136111652e-05,
+      "loss": 0.9595,
+      "step": 59430
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.2953206960449471e-05,
+      "loss": 1.0603,
+      "step": 59435
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.29408862409794e-05,
+      "loss": 0.9718,
+      "step": 59440
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.2928570978473753e-05,
+      "loss": 1.0246,
+      "step": 59445
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2916261173704425e-05,
+      "loss": 0.9285,
+      "step": 59450
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.290395682744302e-05,
+      "loss": 0.8797,
+      "step": 59455
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5,
+      "learning_rate": 1.2891657940460822e-05,
+      "loss": 0.8116,
+      "step": 59460
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.287936451352868e-05,
+      "loss": 1.02,
+      "step": 59465
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.2867076547417223e-05,
+      "loss": 0.8766,
+      "step": 59470
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.2854794042896623e-05,
+      "loss": 1.0155,
+      "step": 59475
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.2842517000736776e-05,
+      "loss": 0.9822,
+      "step": 59480
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.2830245421707255e-05,
+      "loss": 0.8895,
+      "step": 59485
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.2817979306577232e-05,
+      "loss": 1.0055,
+      "step": 59490
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.2805718656115528e-05,
+      "loss": 0.9778,
+      "step": 59495
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.2793463471090695e-05,
+      "loss": 1.0549,
+      "step": 59500
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.2781213752270904e-05,
+      "loss": 1.0148,
+      "step": 59505
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.2768969500423954e-05,
+      "loss": 0.9994,
+      "step": 59510
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.2756730716317333e-05,
+      "loss": 1.0061,
+      "step": 59515
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.2744497400718224e-05,
+      "loss": 0.9772,
+      "step": 59520
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.2732269554393373e-05,
+      "loss": 0.9542,
+      "step": 59525
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2720047178109252e-05,
+      "loss": 1.0453,
+      "step": 59530
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.2707830272631994e-05,
+      "loss": 1.0213,
+      "step": 59535
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2695618838727352e-05,
+      "loss": 0.9817,
+      "step": 59540
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.268341287716074e-05,
+      "loss": 1.0595,
+      "step": 59545
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.267121238869725e-05,
+      "loss": 0.9901,
+      "step": 59550
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.2659017374101646e-05,
+      "loss": 0.9364,
+      "step": 59555
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2646827834138286e-05,
+      "loss": 0.8643,
+      "step": 59560
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.2634643769571242e-05,
+      "loss": 0.9698,
+      "step": 59565
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.2622465181164256e-05,
+      "loss": 0.8793,
+      "step": 59570
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.261029206968064e-05,
+      "loss": 0.8717,
+      "step": 59575
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.2598124435883452e-05,
+      "loss": 0.7989,
+      "step": 59580
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.258596228053538e-05,
+      "loss": 0.9763,
+      "step": 59585
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.2573805604398725e-05,
+      "loss": 0.8852,
+      "step": 59590
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.2561654408235535e-05,
+      "loss": 1.0915,
+      "step": 59595
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5,
+      "learning_rate": 1.2549508692807422e-05,
+      "loss": 0.9351,
+      "step": 59600
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.2537368458875687e-05,
+      "loss": 0.9376,
+      "step": 59605
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.2525233707201311e-05,
+      "loss": 0.9712,
+      "step": 59610
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.2513104438544932e-05,
+      "loss": 1.1597,
+      "step": 59615
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2500980653666782e-05,
+      "loss": 0.7914,
+      "step": 59620
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.2488862353326813e-05,
+      "loss": 0.9986,
+      "step": 59625
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.2476749538284648e-05,
+      "loss": 1.0384,
+      "step": 59630
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.246464220929947e-05,
+      "loss": 0.9533,
+      "step": 59635
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.2452540367130228e-05,
+      "loss": 0.9425,
+      "step": 59640
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.2440444012535468e-05,
+      "loss": 0.8041,
+      "step": 59645
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.2428353146273407e-05,
+      "loss": 1.0403,
+      "step": 59650
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.625,
+      "learning_rate": 1.2416267769101898e-05,
+      "loss": 1.0474,
+      "step": 59655
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.2404187881778462e-05,
+      "loss": 1.2028,
+      "step": 59660
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.2392113485060309e-05,
+      "loss": 1.0045,
+      "step": 59665
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2380044579704254e-05,
+      "loss": 1.0195,
+      "step": 59670
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.236798116646678e-05,
+      "loss": 0.9197,
+      "step": 59675
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.2355923246104084e-05,
+      "loss": 0.9245,
+      "step": 59680
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.2343870819371906e-05,
+      "loss": 1.0074,
+      "step": 59685
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.2331823887025739e-05,
+      "loss": 0.9121,
+      "step": 59690
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.2319782449820727e-05,
+      "loss": 0.8504,
+      "step": 59695
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.2307746508511597e-05,
+      "loss": 1.005,
+      "step": 59700
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.229571606385278e-05,
+      "loss": 0.9193,
+      "step": 59705
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.2283691116598384e-05,
+      "loss": 0.8849,
+      "step": 59710
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.2271671667502105e-05,
+      "loss": 0.9522,
+      "step": 59715
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.2259657717317362e-05,
+      "loss": 0.9858,
+      "step": 59720
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5,
+      "learning_rate": 1.2247649266797223e-05,
+      "loss": 0.9662,
+      "step": 59725
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.223564631669435e-05,
+      "loss": 0.9701,
+      "step": 59730
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.2223648867761116e-05,
+      "loss": 1.1265,
+      "step": 59735
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.2211656920749559e-05,
+      "loss": 0.9027,
+      "step": 59740
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.2199670476411319e-05,
+      "loss": 0.9315,
+      "step": 59745
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.2187689535497738e-05,
+      "loss": 1.036,
+      "step": 59750
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.2175714098759772e-05,
+      "loss": 0.9804,
+      "step": 59755
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.2163744166948099e-05,
+      "loss": 0.8822,
+      "step": 59760
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.215177974081294e-05,
+      "loss": 1.0182,
+      "step": 59765
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.2139820821104286e-05,
+      "loss": 0.82,
+      "step": 59770
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.2127867408571757e-05,
+      "loss": 0.862,
+      "step": 59775
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.2115919503964545e-05,
+      "loss": 0.9314,
+      "step": 59780
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.2103977108031595e-05,
+      "loss": 0.9286,
+      "step": 59785
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.2092040221521484e-05,
+      "loss": 0.8684,
+      "step": 59790
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.2080108845182392e-05,
+      "loss": 0.9665,
+      "step": 59795
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.2068182979762198e-05,
+      "loss": 0.8151,
+      "step": 59800
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.2056262626008496e-05,
+      "loss": 0.8913,
+      "step": 59805
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.2044347784668364e-05,
+      "loss": 0.9524,
+      "step": 59810
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.2032438456488692e-05,
+      "loss": 0.9238,
+      "step": 59815
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.2020534642215985e-05,
+      "loss": 0.9864,
+      "step": 59820
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.2008636342596347e-05,
+      "loss": 0.9358,
+      "step": 59825
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.199674355837559e-05,
+      "loss": 1.0017,
+      "step": 59830
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.1984856290299196e-05,
+      "loss": 0.9784,
+      "step": 59835
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.1972974539112225e-05,
+      "loss": 0.9955,
+      "step": 59840
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.1961098305559471e-05,
+      "loss": 1.0199,
+      "step": 59845
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.194922759038537e-05,
+      "loss": 0.8711,
+      "step": 59850
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.193736239433395e-05,
+      "loss": 1.0284,
+      "step": 59855
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.192550271814894e-05,
+      "loss": 0.8833,
+      "step": 59860
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.1913648562573732e-05,
+      "loss": 0.8689,
+      "step": 59865
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.1901799928351365e-05,
+      "loss": 0.926,
+      "step": 59870
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.1889956816224502e-05,
+      "loss": 0.9446,
+      "step": 59875
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.1878119226935481e-05,
+      "loss": 1.0387,
+      "step": 59880
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.1866287161226341e-05,
+      "loss": 0.9219,
+      "step": 59885
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.185446061983867e-05,
+      "loss": 0.9404,
+      "step": 59890
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.1842639603513805e-05,
+      "loss": 0.9646,
+      "step": 59895
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.183082411299271e-05,
+      "loss": 1.0031,
+      "step": 59900
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.181901414901595e-05,
+      "loss": 1.0052,
+      "step": 59905
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.1807209712323842e-05,
+      "loss": 0.9562,
+      "step": 59910
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.1795410803656271e-05,
+      "loss": 0.8217,
+      "step": 59915
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.1783617423752779e-05,
+      "loss": 0.9394,
+      "step": 59920
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.1771829573352622e-05,
+      "loss": 1.0048,
+      "step": 59925
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.1760047253194695e-05,
+      "loss": 1.036,
+      "step": 59930
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.1748270464017475e-05,
+      "loss": 0.9587,
+      "step": 59935
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.1736499206559182e-05,
+      "loss": 0.8339,
+      "step": 59940
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.172473348155766e-05,
+      "loss": 0.9504,
+      "step": 59945
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1712973289750351e-05,
+      "loss": 0.8442,
+      "step": 59950
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.1701218631874434e-05,
+      "loss": 0.9779,
+      "step": 59955
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.1689469508666718e-05,
+      "loss": 1.0213,
+      "step": 59960
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.1677725920863625e-05,
+      "loss": 1.0294,
+      "step": 59965
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.1665987869201245e-05,
+      "loss": 0.96,
+      "step": 59970
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.1654255354415356e-05,
+      "loss": 0.9051,
+      "step": 59975
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.1642528377241369e-05,
+      "loss": 0.8809,
+      "step": 59980
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.1630806938414318e-05,
+      "loss": 0.9754,
+      "step": 59985
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.1619091038668928e-05,
+      "loss": 0.8443,
+      "step": 59990
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.1607380678739598e-05,
+      "loss": 0.949,
+      "step": 59995
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.1595675859360288e-05,
+      "loss": 0.9439,
+      "step": 60000
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.158397658126471e-05,
+      "loss": 0.8939,
+      "step": 60005
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.1572282845186188e-05,
+      "loss": 0.9047,
+      "step": 60010
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.1560594651857692e-05,
+      "loss": 0.9178,
+      "step": 60015
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.1548912002011825e-05,
+      "loss": 0.7978,
+      "step": 60020
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.1537234896380922e-05,
+      "loss": 0.9594,
+      "step": 60025
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.1525563335696854e-05,
+      "loss": 0.887,
+      "step": 60030
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.1513897320691236e-05,
+      "loss": 0.9302,
+      "step": 60035
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1502236852095338e-05,
+      "loss": 0.8694,
+      "step": 60040
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.149058193064001e-05,
+      "loss": 0.9651,
+      "step": 60045
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.1478932557055799e-05,
+      "loss": 1.0108,
+      "step": 60050
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.1467288732072944e-05,
+      "loss": 0.9062,
+      "step": 60055
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.1455650456421241e-05,
+      "loss": 1.0262,
+      "step": 60060
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.1444017730830226e-05,
+      "loss": 1.0358,
+      "step": 60065
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.1432390556029015e-05,
+      "loss": 1.0133,
+      "step": 60070
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.142076893274645e-05,
+      "loss": 0.8855,
+      "step": 60075
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1409152861710959e-05,
+      "loss": 1.0818,
+      "step": 60080
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.1397542343650657e-05,
+      "loss": 1.052,
+      "step": 60085
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.1385937379293343e-05,
+      "loss": 1.0159,
+      "step": 60090
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.1374337969366367e-05,
+      "loss": 0.822,
+      "step": 60095
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.1362744114596823e-05,
+      "loss": 1.0121,
+      "step": 60100
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.1351155815711445e-05,
+      "loss": 0.8235,
+      "step": 60105
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.133957307343657e-05,
+      "loss": 0.9591,
+      "step": 60110
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.1327995888498233e-05,
+      "loss": 1.0327,
+      "step": 60115
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.1316424261622116e-05,
+      "loss": 1.0982,
+      "step": 60120
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.130485819353353e-05,
+      "loss": 1.0098,
+      "step": 60125
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.1293297684957427e-05,
+      "loss": 0.965,
+      "step": 60130
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.1281742736618451e-05,
+      "loss": 0.8862,
+      "step": 60135
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5,
+      "learning_rate": 1.127019334924091e-05,
+      "loss": 0.9905,
+      "step": 60140
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.1258649523548681e-05,
+      "loss": 0.919,
+      "step": 60145
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.1247111260265386e-05,
+      "loss": 0.93,
+      "step": 60150
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.1235578560114212e-05,
+      "loss": 0.87,
+      "step": 60155
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.1224051423818083e-05,
+      "loss": 0.9597,
+      "step": 60160
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.121252985209953e-05,
+      "loss": 0.9734,
+      "step": 60165
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.1201013845680719e-05,
+      "loss": 0.8142,
+      "step": 60170
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.1189503405283486e-05,
+      "loss": 1.0303,
+      "step": 60175
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.117799853162932e-05,
+      "loss": 1.0191,
+      "step": 60180
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.1166499225439387e-05,
+      "loss": 0.956,
+      "step": 60185
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1155005487434433e-05,
+      "loss": 0.9282,
+      "step": 60190
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.1143517318334928e-05,
+      "loss": 0.9121,
+      "step": 60195
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.1132034718860973e-05,
+      "loss": 1.0497,
+      "step": 60200
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.112055768973227e-05,
+      "loss": 0.9809,
+      "step": 60205
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.1109086231668243e-05,
+      "loss": 0.934,
+      "step": 60210
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.1097620345387938e-05,
+      "loss": 0.88,
+      "step": 60215
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.1086160031610038e-05,
+      "loss": 1.0801,
+      "step": 60220
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.10747052910529e-05,
+      "loss": 0.9801,
+      "step": 60225
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.625,
+      "learning_rate": 1.1063256124434496e-05,
+      "loss": 0.926,
+      "step": 60230
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.75,
+      "learning_rate": 1.1051812532472505e-05,
+      "loss": 0.9908,
+      "step": 60235
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.1040374515884189e-05,
+      "loss": 0.9456,
+      "step": 60240
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.1028942075386517e-05,
+      "loss": 1.0099,
+      "step": 60245
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.1017515211696095e-05,
+      "loss": 0.9202,
+      "step": 60250
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.100609392552916e-05,
+      "loss": 0.9572,
+      "step": 60255
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.0994678217601595e-05,
+      "loss": 1.1325,
+      "step": 60260
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.0983268088629006e-05,
+      "loss": 0.9196,
+      "step": 60265
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.0971863539326521e-05,
+      "loss": 1.079,
+      "step": 60270
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.0960464570409045e-05,
+      "loss": 0.9669,
+      "step": 60275
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.0949071182591064e-05,
+      "loss": 1.0241,
+      "step": 60280
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.625,
+      "learning_rate": 1.0937683376586693e-05,
+      "loss": 0.7955,
+      "step": 60285
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.0926301153109774e-05,
+      "loss": 0.9317,
+      "step": 60290
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.0914924512873759e-05,
+      "loss": 0.9039,
+      "step": 60295
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.0903553456591709e-05,
+      "loss": 0.9029,
+      "step": 60300
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.089218798497641e-05,
+      "loss": 0.9796,
+      "step": 60305
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.088082809874027e-05,
+      "loss": 1.0671,
+      "step": 60310
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.0869473798595298e-05,
+      "loss": 0.7776,
+      "step": 60315
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.0858125085253223e-05,
+      "loss": 0.9387,
+      "step": 60320
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.084678195942541e-05,
+      "loss": 0.9069,
+      "step": 60325
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.0835444421822848e-05,
+      "loss": 1.0661,
+      "step": 60330
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0824112473156157e-05,
+      "loss": 0.9517,
+      "step": 60335
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.0812786114135653e-05,
+      "loss": 1.0505,
+      "step": 60340
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.0801465345471318e-05,
+      "loss": 1.0632,
+      "step": 60345
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.0790150167872703e-05,
+      "loss": 0.9453,
+      "step": 60350
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.0778840582049066e-05,
+      "loss": 0.9767,
+      "step": 60355
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.0767536588709349e-05,
+      "loss": 0.9211,
+      "step": 60360
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.625,
+      "learning_rate": 1.075623818856204e-05,
+      "loss": 0.9557,
+      "step": 60365
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.0744945382315364e-05,
+      "loss": 0.9671,
+      "step": 60370
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.0733658170677185e-05,
+      "loss": 1.0677,
+      "step": 60375
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.0722376554354973e-05,
+      "loss": 1.0411,
+      "step": 60380
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.0711100534055862e-05,
+      "loss": 0.9604,
+      "step": 60385
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.0699830110486675e-05,
+      "loss": 1.0111,
+      "step": 60390
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.0688565284353825e-05,
+      "loss": 1.0119,
+      "step": 60395
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.0677306056363424e-05,
+      "loss": 0.9933,
+      "step": 60400
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.066605242722123e-05,
+      "loss": 1.035,
+      "step": 60405
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.065480439763258e-05,
+      "loss": 0.7821,
+      "step": 60410
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.0643561968302552e-05,
+      "loss": 0.9285,
+      "step": 60415
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.063232513993585e-05,
+      "loss": 1.0274,
+      "step": 60420
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.062109391323678e-05,
+      "loss": 0.9299,
+      "step": 60425
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.0609868288909319e-05,
+      "loss": 0.9836,
+      "step": 60430
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.0598648267657151e-05,
+      "loss": 0.9447,
+      "step": 60435
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.0587433850183525e-05,
+      "loss": 0.8362,
+      "step": 60440
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.0576225037191367e-05,
+      "loss": 0.9271,
+      "step": 60445
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.056502182938326e-05,
+      "loss": 0.976,
+      "step": 60450
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.0553824227461461e-05,
+      "loss": 0.9894,
+      "step": 60455
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.0542632232127814e-05,
+      "loss": 1.0547,
+      "step": 60460
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.0531445844083865e-05,
+      "loss": 0.9005,
+      "step": 60465
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.05202650640308e-05,
+      "loss": 0.9292,
+      "step": 60470
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.0509089892669411e-05,
+      "loss": 0.8289,
+      "step": 60475
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5,
+      "learning_rate": 1.0497920330700195e-05,
+      "loss": 0.9028,
+      "step": 60480
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.0486756378823304e-05,
+      "loss": 0.8278,
+      "step": 60485
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.0475598037738433e-05,
+      "loss": 0.9213,
+      "step": 60490
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.0464445308145022e-05,
+      "loss": 0.9051,
+      "step": 60495
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.0453298190742189e-05,
+      "loss": 1.0026,
+      "step": 60500
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.0442156686228576e-05,
+      "loss": 0.9675,
+      "step": 60505
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.0431020795302592e-05,
+      "loss": 1.1873,
+      "step": 60510
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0419890518662256e-05,
+      "loss": 0.9899,
+      "step": 60515
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.0408765857005177e-05,
+      "loss": 1.0228,
+      "step": 60520
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.0397646811028695e-05,
+      "loss": 1.025,
+      "step": 60525
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0386533381429775e-05,
+      "loss": 0.8756,
+      "step": 60530
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0375425568904984e-05,
+      "loss": 1.1357,
+      "step": 60535
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0364323374150619e-05,
+      "loss": 0.9211,
+      "step": 60540
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.0353226797862526e-05,
+      "loss": 0.9579,
+      "step": 60545
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.03421358407363e-05,
+      "loss": 0.9193,
+      "step": 60550
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.0331050503467088e-05,
+      "loss": 1.0952,
+      "step": 60555
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.0319970786749755e-05,
+      "loss": 1.0066,
+      "step": 60560
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.0308896691278812e-05,
+      "loss": 0.8826,
+      "step": 60565
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.029782821774835e-05,
+      "loss": 0.9592,
+      "step": 60570
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.028676536685218e-05,
+      "loss": 0.8765,
+      "step": 60575
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.0275708139283746e-05,
+      "loss": 0.8267,
+      "step": 60580
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.0264656535736106e-05,
+      "loss": 0.9003,
+      "step": 60585
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.0253610556902005e-05,
+      "loss": 0.9266,
+      "step": 60590
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.0242570203473811e-05,
+      "loss": 0.9141,
+      "step": 60595
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.0231535476143516e-05,
+      "loss": 0.9906,
+      "step": 60600
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.0220506375602823e-05,
+      "loss": 0.9491,
+      "step": 60605
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.0209482902543055e-05,
+      "loss": 0.8141,
+      "step": 60610
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.0198465057655137e-05,
+      "loss": 0.9482,
+      "step": 60615
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.0187452841629718e-05,
+      "loss": 0.9945,
+      "step": 60620
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.0176446255157058e-05,
+      "loss": 0.8699,
+      "step": 60625
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.0165445298927023e-05,
+      "loss": 0.8981,
+      "step": 60630
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0154449973629188e-05,
+      "loss": 1.1487,
+      "step": 60635
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.0143460279952777e-05,
+      "loss": 0.8834,
+      "step": 60640
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.0132476218586618e-05,
+      "loss": 1.0051,
+      "step": 60645
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.012149779021917e-05,
+      "loss": 0.8582,
+      "step": 60650
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.0110524995538617e-05,
+      "loss": 1.0623,
+      "step": 60655
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.0099557835232743e-05,
+      "loss": 1.0097,
+      "step": 60660
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0088596309988952e-05,
+      "loss": 1.0686,
+      "step": 60665
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.0077640420494349e-05,
+      "loss": 1.0395,
+      "step": 60670
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.0066690167435677e-05,
+      "loss": 0.8752,
+      "step": 60675
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.0055745551499263e-05,
+      "loss": 0.8251,
+      "step": 60680
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.0044806573371168e-05,
+      "loss": 0.7511,
+      "step": 60685
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.0033873233737056e-05,
+      "loss": 1.025,
+      "step": 60690
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.0022945533282235e-05,
+      "loss": 0.9772,
+      "step": 60695
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.0012023472691655e-05,
+      "loss": 0.993,
+      "step": 60700
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.0001107052649939e-05,
+      "loss": 0.8906,
+      "step": 60705
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.990196273841312e-06,
+      "loss": 0.9415,
+      "step": 60710
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.979291136949708e-06,
+      "loss": 1.1337,
+      "step": 60715
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.96839164265867e-06,
+      "loss": 0.8363,
+      "step": 60720
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.95749779165136e-06,
+      "loss": 0.9499,
+      "step": 60725
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.946609584610644e-06,
+      "loss": 0.8801,
+      "step": 60730
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.935727022219022e-06,
+      "loss": 0.8072,
+      "step": 60735
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.924850105158578e-06,
+      "loss": 0.8885,
+      "step": 60740
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.913978834111126e-06,
+      "loss": 0.9325,
+      "step": 60745
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.903113209758096e-06,
+      "loss": 0.981,
+      "step": 60750
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.892253232780536e-06,
+      "loss": 0.8377,
+      "step": 60755
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.881398903859152e-06,
+      "loss": 0.9308,
+      "step": 60760
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.870550223674323e-06,
+      "loss": 1.1031,
+      "step": 60765
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.859707192906076e-06,
+      "loss": 0.7768,
+      "step": 60770
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.625,
+      "learning_rate": 9.848869812234018e-06,
+      "loss": 1.0453,
+      "step": 60775
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.838038082337474e-06,
+      "loss": 0.9651,
+      "step": 60780
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.827212003895414e-06,
+      "loss": 1.0032,
+      "step": 60785
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.46875,
+      "learning_rate": 9.816391577586381e-06,
+      "loss": 1.1093,
+      "step": 60790
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.805576804088633e-06,
+      "loss": 0.8107,
+      "step": 60795
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.794767684080076e-06,
+      "loss": 1.0138,
+      "step": 60800
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.783964218238206e-06,
+      "loss": 0.9424,
+      "step": 60805
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.773166407240197e-06,
+      "loss": 0.9967,
+      "step": 60810
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.762374251762873e-06,
+      "loss": 0.8841,
+      "step": 60815
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.751587752482728e-06,
+      "loss": 1.014,
+      "step": 60820
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.740806910075829e-06,
+      "loss": 1.0144,
+      "step": 60825
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.73003172521796e-06,
+      "loss": 0.9608,
+      "step": 60830
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.7192621985845e-06,
+      "loss": 0.9867,
+      "step": 60835
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.708498330850502e-06,
+      "loss": 0.9342,
+      "step": 60840
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.697740122690679e-06,
+      "loss": 0.9266,
+      "step": 60845
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.686987574779338e-06,
+      "loss": 1.1289,
+      "step": 60850
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.676240687790484e-06,
+      "loss": 0.8821,
+      "step": 60855
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.66549946239772e-06,
+      "loss": 0.9076,
+      "step": 60860
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.654763899274355e-06,
+      "loss": 0.9685,
+      "step": 60865
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.64403399909326e-06,
+      "loss": 0.9358,
+      "step": 60870
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.625,
+      "learning_rate": 9.63330976252702e-06,
+      "loss": 1.0837,
+      "step": 60875
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.51171875,
+      "learning_rate": 9.622591190247864e-06,
+      "loss": 0.8836,
+      "step": 60880
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.6118782829276e-06,
+      "loss": 1.0149,
+      "step": 60885
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.48046875,
+      "learning_rate": 9.601171041237756e-06,
+      "loss": 0.9759,
+      "step": 60890
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.590469465849484e-06,
+      "loss": 0.9763,
+      "step": 60895
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.579773557433535e-06,
+      "loss": 0.9364,
+      "step": 60900
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5,
+      "learning_rate": 9.569083316660377e-06,
+      "loss": 0.8355,
+      "step": 60905
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.558398744200048e-06,
+      "loss": 0.9732,
+      "step": 60910
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.547719840722301e-06,
+      "loss": 0.9637,
+      "step": 60915
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.537046606896482e-06,
+      "loss": 1.1266,
+      "step": 60920
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.498046875,
+      "learning_rate": 9.52637904339162e-06,
+      "loss": 0.8926,
+      "step": 60925
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.515717150876369e-06,
+      "loss": 0.8966,
+      "step": 60930
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.482421875,
+      "learning_rate": 9.505060930018995e-06,
+      "loss": 0.9604,
+      "step": 60935
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.494410381487474e-06,
+      "loss": 0.8663,
+      "step": 60940
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.483765505949405e-06,
+      "loss": 0.9535,
+      "step": 60945
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.473126304071989e-06,
+      "loss": 0.9091,
+      "step": 60950
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.462492776522114e-06,
+      "loss": 0.9805,
+      "step": 60955
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.451864923966336e-06,
+      "loss": 0.9339,
+      "step": 60960
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.441242747070766e-06,
+      "loss": 0.9978,
+      "step": 60965
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.430626246501228e-06,
+      "loss": 0.9269,
+      "step": 60970
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.420015422923212e-06,
+      "loss": 1.0158,
+      "step": 60975
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.409410277001762e-06,
+      "loss": 0.9552,
+      "step": 60980
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.39881080940166e-06,
+      "loss": 1.0178,
+      "step": 60985
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.625,
+      "learning_rate": 9.388217020787304e-06,
+      "loss": 0.9533,
+      "step": 60990
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4921875,
+      "learning_rate": 9.377628911822678e-06,
+      "loss": 0.9394,
+      "step": 60995
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.367046483171482e-06,
+      "loss": 0.9528,
+      "step": 61000
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.356469735497053e-06,
+      "loss": 1.0007,
+      "step": 61005
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.345898669462338e-06,
+      "loss": 0.9128,
+      "step": 61010
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.33533328572992e-06,
+      "loss": 0.997,
+      "step": 61015
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.324773584962077e-06,
+      "loss": 0.9481,
+      "step": 61020
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.314219567820703e-06,
+      "loss": 0.899,
+      "step": 61025
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.303671234967315e-06,
+      "loss": 1.0267,
+      "step": 61030
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.293128587063115e-06,
+      "loss": 0.9971,
+      "step": 61035
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.282591624768932e-06,
+      "loss": 0.9922,
+      "step": 61040
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.272060348745216e-06,
+      "loss": 1.1262,
+      "step": 61045
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.26153475965208e-06,
+      "loss": 0.9235,
+      "step": 61050
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.251014858149298e-06,
+      "loss": 0.8597,
+      "step": 61055
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.240500644896255e-06,
+      "loss": 1.07,
+      "step": 61060
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.22999212055201e-06,
+      "loss": 0.9589,
+      "step": 61065
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.625,
+      "learning_rate": 9.21948928577524e-06,
+      "loss": 0.9358,
+      "step": 61070
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.208992141224249e-06,
+      "loss": 0.9216,
+      "step": 61075
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.198500687557031e-06,
+      "loss": 0.9737,
+      "step": 61080
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.188014925431233e-06,
+      "loss": 1.0058,
+      "step": 61085
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.177534855504055e-06,
+      "loss": 0.9112,
+      "step": 61090
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.167060478432443e-06,
+      "loss": 0.9753,
+      "step": 61095
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.156591794872938e-06,
+      "loss": 1.0607,
+      "step": 61100
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.146128805481712e-06,
+      "loss": 1.0579,
+      "step": 61105
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49609375,
+      "learning_rate": 9.1356715109146e-06,
+      "loss": 0.8809,
+      "step": 61110
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.125219911827099e-06,
+      "loss": 0.9079,
+      "step": 61115
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.114774008874317e-06,
+      "loss": 1.0232,
+      "step": 61120
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.104333802710997e-06,
+      "loss": 0.8465,
+      "step": 61125
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.093899293991548e-06,
+      "loss": 0.9612,
+      "step": 61130
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.083470483370037e-06,
+      "loss": 1.0004,
+      "step": 61135
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.073047371500133e-06,
+      "loss": 1.0981,
+      "step": 61140
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.062629959035173e-06,
+      "loss": 0.9233,
+      "step": 61145
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.052218246628163e-06,
+      "loss": 0.9419,
+      "step": 61150
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5,
+      "learning_rate": 9.041812234931667e-06,
+      "loss": 0.8708,
+      "step": 61155
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.031411924597977e-06,
+      "loss": 0.9768,
+      "step": 61160
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.021017316279001e-06,
+      "loss": 1.0744,
+      "step": 61165
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.010628410626287e-06,
+      "loss": 0.9773,
+      "step": 61170
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.00024520829098e-06,
+      "loss": 0.9747,
+      "step": 61175
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.625,
+      "learning_rate": 8.989867709923983e-06,
+      "loss": 0.9236,
+      "step": 61180
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.979495916175695e-06,
+      "loss": 0.9793,
+      "step": 61185
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.969129827696265e-06,
+      "loss": 0.9123,
+      "step": 61190
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.958769445135483e-06,
+      "loss": 0.888,
+      "step": 61195
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.948414769142698e-06,
+      "loss": 0.9846,
+      "step": 61200
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.938065800366967e-06,
+      "loss": 0.9033,
+      "step": 61205
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.927722539457006e-06,
+      "loss": 0.9003,
+      "step": 61210
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.917384987061105e-06,
+      "loss": 0.94,
+      "step": 61215
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.90705314382726e-06,
+      "loss": 1.0094,
+      "step": 61220
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.89672701040305e-06,
+      "loss": 1.0216,
+      "step": 61225
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.88640658743577e-06,
+      "loss": 1.0245,
+      "step": 61230
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5,
+      "learning_rate": 8.876091875572279e-06,
+      "loss": 0.9166,
+      "step": 61235
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.865782875459127e-06,
+      "loss": 1.1021,
+      "step": 61240
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.855479587742521e-06,
+      "loss": 0.9594,
+      "step": 61245
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.845182013068231e-06,
+      "loss": 0.8905,
+      "step": 61250
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.834890152081754e-06,
+      "loss": 0.9806,
+      "step": 61255
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.824604005428217e-06,
+      "loss": 0.9909,
+      "step": 61260
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.814323573752315e-06,
+      "loss": 0.8444,
+      "step": 61265
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.804048857698466e-06,
+      "loss": 0.9958,
+      "step": 61270
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.793779857910734e-06,
+      "loss": 0.8591,
+      "step": 61275
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.783516575032713e-06,
+      "loss": 0.9333,
+      "step": 61280
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.48046875,
+      "learning_rate": 8.773259009707769e-06,
+      "loss": 0.9208,
+      "step": 61285
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.763007162578862e-06,
+      "loss": 1.2006,
+      "step": 61290
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.67578125,
+      "learning_rate": 8.75276103428856e-06,
+      "loss": 1.1762,
+      "step": 61295
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.742520625479111e-06,
+      "loss": 0.9718,
+      "step": 61300
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.732285936792427e-06,
+      "loss": 1.0278,
+      "step": 61305
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.722056968869985e-06,
+      "loss": 0.908,
+      "step": 61310
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.71183372235298e-06,
+      "loss": 0.8661,
+      "step": 61315
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.701616197882211e-06,
+      "loss": 1.0178,
+      "step": 61320
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.691404396098124e-06,
+      "loss": 0.9051,
+      "step": 61325
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.681198317640783e-06,
+      "loss": 0.887,
+      "step": 61330
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.67099796314993e-06,
+      "loss": 0.9121,
+      "step": 61335
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.660803333264966e-06,
+      "loss": 0.9048,
+      "step": 61340
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.650614428624848e-06,
+      "loss": 0.9388,
+      "step": 61345
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.640431249868264e-06,
+      "loss": 0.9448,
+      "step": 61350
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.630253797633514e-06,
+      "loss": 0.8843,
+      "step": 61355
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.620082072558499e-06,
+      "loss": 0.9721,
+      "step": 61360
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.609916075280821e-06,
+      "loss": 0.8592,
+      "step": 61365
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.599755806437693e-06,
+      "loss": 0.8063,
+      "step": 61370
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.58960126666596e-06,
+      "loss": 0.9778,
+      "step": 61375
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4921875,
+      "learning_rate": 8.579452456602133e-06,
+      "loss": 0.9136,
+      "step": 61380
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.569309376882362e-06,
+      "loss": 0.9862,
+      "step": 61385
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.559172028142381e-06,
+      "loss": 0.9479,
+      "step": 61390
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.549040411017639e-06,
+      "loss": 1.0091,
+      "step": 61395
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.538914526143216e-06,
+      "loss": 0.9978,
+      "step": 61400
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.528794374153781e-06,
+      "loss": 0.7958,
+      "step": 61405
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.518679955683694e-06,
+      "loss": 1.1249,
+      "step": 61410
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.50857127136695e-06,
+      "loss": 0.9851,
+      "step": 61415
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.498468321837139e-06,
+      "loss": 1.1021,
+      "step": 61420
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.488371107727545e-06,
+      "loss": 0.9302,
+      "step": 61425
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.478279629671094e-06,
+      "loss": 0.9321,
+      "step": 61430
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.468193888300313e-06,
+      "loss": 0.9139,
+      "step": 61435
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.458113884247365e-06,
+      "loss": 0.9128,
+      "step": 61440
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.448039618144089e-06,
+      "loss": 0.8569,
+      "step": 61445
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.437971090621988e-06,
+      "loss": 1.0033,
+      "step": 61450
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.427908302312115e-06,
+      "loss": 0.9779,
+      "step": 61455
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.417851253845244e-06,
+      "loss": 0.7853,
+      "step": 61460
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.40779994585178e-06,
+      "loss": 1.0948,
+      "step": 61465
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.397754378961709e-06,
+      "loss": 1.0884,
+      "step": 61470
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.387714553804716e-06,
+      "loss": 1.2371,
+      "step": 61475
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.377680471010129e-06,
+      "loss": 0.9435,
+      "step": 61480
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.367652131206882e-06,
+      "loss": 1.043,
+      "step": 61485
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.357629535023547e-06,
+      "loss": 0.8901,
+      "step": 61490
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.34761268308838e-06,
+      "loss": 0.9678,
+      "step": 61495
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.337601576029219e-06,
+      "loss": 0.8858,
+      "step": 61500
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.327596214473576e-06,
+      "loss": 1.0307,
+      "step": 61505
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.490234375,
+      "learning_rate": 8.317596599048638e-06,
+      "loss": 0.94,
+      "step": 61510
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.307602730381137e-06,
+      "loss": 1.0737,
+      "step": 61515
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.297614609097526e-06,
+      "loss": 0.9876,
+      "step": 61520
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.287632235823884e-06,
+      "loss": 0.946,
+      "step": 61525
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.277655611185897e-06,
+      "loss": 0.9989,
+      "step": 61530
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.267684735808934e-06,
+      "loss": 0.8744,
+      "step": 61535
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.25771961031795e-06,
+      "loss": 0.8663,
+      "step": 61540
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.247760235337598e-06,
+      "loss": 0.9953,
+      "step": 61545
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.237806611492127e-06,
+      "loss": 0.8985,
+      "step": 61550
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.227858739405436e-06,
+      "loss": 1.0169,
+      "step": 61555
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.217916619701093e-06,
+      "loss": 0.8976,
+      "step": 61560
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.207980253002257e-06,
+      "loss": 0.9396,
+      "step": 61565
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.19804963993176e-06,
+      "loss": 0.8884,
+      "step": 61570
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.188124781112082e-06,
+      "loss": 0.8163,
+      "step": 61575
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.178205677165285e-06,
+      "loss": 0.8764,
+      "step": 61580
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.168292328713145e-06,
+      "loss": 1.0038,
+      "step": 61585
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.79296875,
+      "learning_rate": 8.158384736377033e-06,
+      "loss": 0.9548,
+      "step": 61590
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.148482900777976e-06,
+      "loss": 1.0273,
+      "step": 61595
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.138586822536597e-06,
+      "loss": 0.9237,
+      "step": 61600
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.128696502273214e-06,
+      "loss": 0.9631,
+      "step": 61605
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.118811940607795e-06,
+      "loss": 0.9891,
+      "step": 61610
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.108933138159857e-06,
+      "loss": 0.9581,
+      "step": 61615
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.099060095548661e-06,
+      "loss": 0.9396,
+      "step": 61620
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.089192813393042e-06,
+      "loss": 1.2638,
+      "step": 61625
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.079331292311498e-06,
+      "loss": 0.9143,
+      "step": 61630
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.069475532922166e-06,
+      "loss": 1.0492,
+      "step": 61635
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.059625535842807e-06,
+      "loss": 0.8953,
+      "step": 61640
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.049781301690806e-06,
+      "loss": 1.0398,
+      "step": 61645
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.039942831083246e-06,
+      "loss": 0.9835,
+      "step": 61650
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.030110124636814e-06,
+      "loss": 0.8837,
+      "step": 61655
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.020283182967814e-06,
+      "loss": 0.9021,
+      "step": 61660
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.010462006692221e-06,
+      "loss": 0.966,
+      "step": 61665
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.000646596425643e-06,
+      "loss": 0.9157,
+      "step": 61670
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.99083695278331e-06,
+      "loss": 0.9347,
+      "step": 61675
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.981033076380095e-06,
+      "loss": 0.9805,
+      "step": 61680
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.971234967830554e-06,
+      "loss": 0.8568,
+      "step": 61685
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.961442627748793e-06,
+      "loss": 0.903,
+      "step": 61690
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.951656056748658e-06,
+      "loss": 0.9831,
+      "step": 61695
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.941875255443532e-06,
+      "loss": 0.9424,
+      "step": 61700
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.932100224446526e-06,
+      "loss": 0.9975,
+      "step": 61705
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.922330964370316e-06,
+      "loss": 1.0486,
+      "step": 61710
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.912567475827271e-06,
+      "loss": 1.1237,
+      "step": 61715
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.902809759429398e-06,
+      "loss": 0.9946,
+      "step": 61720
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.893057815788273e-06,
+      "loss": 0.9619,
+      "step": 61725
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.88331164551519e-06,
+      "loss": 0.8598,
+      "step": 61730
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.873571249221057e-06,
+      "loss": 0.924,
+      "step": 61735
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.863836627516396e-06,
+      "loss": 0.974,
+      "step": 61740
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.854107781011399e-06,
+      "loss": 1.0555,
+      "step": 61745
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.844384710315867e-06,
+      "loss": 1.0496,
+      "step": 61750
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.834667416039254e-06,
+      "loss": 0.9431,
+      "step": 61755
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.824955898790642e-06,
+      "loss": 0.9346,
+      "step": 61760
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.8152501591788e-06,
+      "loss": 1.0252,
+      "step": 61765
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.80555019781204e-06,
+      "loss": 0.9873,
+      "step": 61770
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.795856015298397e-06,
+      "loss": 0.813,
+      "step": 61775
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.786167612245531e-06,
+      "loss": 0.8898,
+      "step": 61780
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.625,
+      "learning_rate": 7.776484989260691e-06,
+      "loss": 1.0275,
+      "step": 61785
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.7668081469508e-06,
+      "loss": 0.8083,
+      "step": 61790
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.498046875,
+      "learning_rate": 7.757137085922428e-06,
+      "loss": 0.9573,
+      "step": 61795
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.74747180678177e-06,
+      "loss": 0.9706,
+      "step": 61800
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.737812310134617e-06,
+      "loss": 1.0688,
+      "step": 61805
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.728158596586477e-06,
+      "loss": 1.0002,
+      "step": 61810
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.718510666742462e-06,
+      "loss": 0.977,
+      "step": 61815
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5,
+      "learning_rate": 7.70886852120728e-06,
+      "loss": 0.9562,
+      "step": 61820
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.699232160585324e-06,
+      "loss": 1.0195,
+      "step": 61825
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.48046875,
+      "learning_rate": 7.689601585480643e-06,
+      "loss": 1.0175,
+      "step": 61830
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.679976796496846e-06,
+      "loss": 0.8564,
+      "step": 61835
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.670357794237249e-06,
+      "loss": 0.9566,
+      "step": 61840
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.498046875,
+      "learning_rate": 7.6607445793048e-06,
+      "loss": 0.9369,
+      "step": 61845
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.651137152302035e-06,
+      "loss": 1.0571,
+      "step": 61850
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.641535513831165e-06,
+      "loss": 0.995,
+      "step": 61855
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.631939664494048e-06,
+      "loss": 0.9018,
+      "step": 61860
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.622349604892131e-06,
+      "loss": 0.9492,
+      "step": 61865
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.612765335626559e-06,
+      "loss": 1.0655,
+      "step": 61870
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.474609375,
+      "learning_rate": 7.603186857298083e-06,
+      "loss": 0.9456,
+      "step": 61875
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.59361417050708e-06,
+      "loss": 0.72,
+      "step": 61880
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.482421875,
+      "learning_rate": 7.584047275853578e-06,
+      "loss": 1.0067,
+      "step": 61885
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.574486173937256e-06,
+      "loss": 1.0156,
+      "step": 61890
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.564930865357389e-06,
+      "loss": 0.9955,
+      "step": 61895
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.484375,
+      "learning_rate": 7.5553813507129426e-06,
+      "loss": 0.8775,
+      "step": 61900
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.545837630602481e-06,
+      "loss": 0.8675,
+      "step": 61905
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.536299705624217e-06,
+      "loss": 0.9599,
+      "step": 61910
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.526767576375982e-06,
+      "loss": 1.0218,
+      "step": 61915
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.5172412434552756e-06,
+      "loss": 0.8365,
+      "step": 61920
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.507720707459232e-06,
+      "loss": 1.0037,
+      "step": 61925
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.4982059689845726e-06,
+      "loss": 1.0069,
+      "step": 61930
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.488697028627711e-06,
+      "loss": 0.9896,
+      "step": 61935
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.479193886984703e-06,
+      "loss": 0.9226,
+      "step": 61940
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.469696544651184e-06,
+      "loss": 0.944,
+      "step": 61945
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.460205002222464e-06,
+      "loss": 0.9059,
+      "step": 61950
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.450719260293515e-06,
+      "loss": 0.9737,
+      "step": 61955
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.44123931945887e-06,
+      "loss": 0.8882,
+      "step": 61960
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.431765180312744e-06,
+      "loss": 0.797,
+      "step": 61965
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.422296843449028e-06,
+      "loss": 0.9947,
+      "step": 61970
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.412834309461159e-06,
+      "loss": 0.9082,
+      "step": 61975
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.403377578942294e-06,
+      "loss": 0.9802,
+      "step": 61980
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.393926652485195e-06,
+      "loss": 1.0125,
+      "step": 61985
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.384481530682219e-06,
+      "loss": 0.9947,
+      "step": 61990
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.3750422141254275e-06,
+      "loss": 1.0376,
+      "step": 61995
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.3656087034064904e-06,
+      "loss": 1.016,
+      "step": 62000
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.356180999116691e-06,
+      "loss": 0.9243,
+      "step": 62005
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.346759101847e-06,
+      "loss": 1.0494,
+      "step": 62010
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.337343012187947e-06,
+      "loss": 0.8579,
+      "step": 62015
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.32793273072978e-06,
+      "loss": 0.8811,
+      "step": 62020
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.318528258062329e-06,
+      "loss": 0.8636,
+      "step": 62025
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.309129594775077e-06,
+      "loss": 1.0092,
+      "step": 62030
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.299736741457164e-06,
+      "loss": 1.0579,
+      "step": 62035
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.290349698697318e-06,
+      "loss": 0.9998,
+      "step": 62040
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.280968467083937e-06,
+      "loss": 0.7762,
+      "step": 62045
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.271593047205061e-06,
+      "loss": 0.9415,
+      "step": 62050
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.2622234396483306e-06,
+      "loss": 0.8974,
+      "step": 62055
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.462890625,
+      "learning_rate": 7.252859645001075e-06,
+      "loss": 1.0368,
+      "step": 62060
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.243501663850205e-06,
+      "loss": 0.9015,
+      "step": 62065
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.234149496782272e-06,
+      "loss": 1.1338,
+      "step": 62070
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.224803144383496e-06,
+      "loss": 0.8891,
+      "step": 62075
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.215462607239743e-06,
+      "loss": 1.0407,
+      "step": 62080
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.206127885936453e-06,
+      "loss": 0.8616,
+      "step": 62085
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.196798981058739e-06,
+      "loss": 1.0154,
+      "step": 62090
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.187475893191387e-06,
+      "loss": 0.9882,
+      "step": 62095
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.178158622918729e-06,
+      "loss": 0.8684,
+      "step": 62100
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.168847170824811e-06,
+      "loss": 0.9177,
+      "step": 62105
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.49609375,
+      "learning_rate": 7.159541537493286e-06,
+      "loss": 0.9148,
+      "step": 62110
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.150241723507433e-06,
+      "loss": 1.0282,
+      "step": 62115
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.1409477294501645e-06,
+      "loss": 1.0454,
+      "step": 62120
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.1316595559040465e-06,
+      "loss": 0.9332,
+      "step": 62125
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.122377203451292e-06,
+      "loss": 1.0585,
+      "step": 62130
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.113100672673701e-06,
+      "loss": 0.9731,
+      "step": 62135
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.1038299641527416e-06,
+      "loss": 0.869,
+      "step": 62140
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.0945650784695396e-06,
+      "loss": 1.0144,
+      "step": 62145
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.085306016204796e-06,
+      "loss": 0.9836,
+      "step": 62150
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.076052777938891e-06,
+      "loss": 0.9506,
+      "step": 62155
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.066805364251849e-06,
+      "loss": 0.9612,
+      "step": 62160
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.057563775723286e-06,
+      "loss": 0.8578,
+      "step": 62165
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.478515625,
+      "learning_rate": 7.048328012932459e-06,
+      "loss": 0.9426,
+      "step": 62170
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.039098076458306e-06,
+      "loss": 0.8757,
+      "step": 62175
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.47265625,
+      "learning_rate": 7.029873966879352e-06,
+      "loss": 0.9648,
+      "step": 62180
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.02065568477378e-06,
+      "loss": 0.8497,
+      "step": 62185
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.011443230719428e-06,
+      "loss": 0.8719,
+      "step": 62190
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.0022366052936885e-06,
+      "loss": 1.085,
+      "step": 62195
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.993035809073678e-06,
+      "loss": 1.1285,
+      "step": 62200
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.983840842636136e-06,
+      "loss": 1.1114,
+      "step": 62205
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.9746517065573556e-06,
+      "loss": 0.979,
+      "step": 62210
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.965468401413366e-06,
+      "loss": 0.8401,
+      "step": 62215
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.956290927779785e-06,
+      "loss": 1.0541,
+      "step": 62220
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.947119286231851e-06,
+      "loss": 1.0197,
+      "step": 62225
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.93795347734445e-06,
+      "loss": 1.0151,
+      "step": 62230
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.92879350169211e-06,
+      "loss": 1.0242,
+      "step": 62235
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.9196393598490175e-06,
+      "loss": 0.9719,
+      "step": 62240
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.910491052388912e-06,
+      "loss": 0.9262,
+      "step": 62245
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.901348579885258e-06,
+      "loss": 1.1226,
+      "step": 62250
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.892211942911109e-06,
+      "loss": 1.05,
+      "step": 62255
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.625,
+      "learning_rate": 6.88308114203915e-06,
+      "loss": 0.8664,
+      "step": 62260
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.873956177841711e-06,
+      "loss": 1.0119,
+      "step": 62265
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.8648370508907825e-06,
+      "loss": 1.0299,
+      "step": 62270
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.855723761757926e-06,
+      "loss": 0.8667,
+      "step": 62275
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.846616311014386e-06,
+      "loss": 1.0801,
+      "step": 62280
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.837514699231018e-06,
+      "loss": 1.0818,
+      "step": 62285
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.828418926978353e-06,
+      "loss": 0.9123,
+      "step": 62290
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.819328994826491e-06,
+      "loss": 0.9992,
+      "step": 62295
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.810244903345209e-06,
+      "loss": 1.0025,
+      "step": 62300
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.8011666531039185e-06,
+      "loss": 1.0085,
+      "step": 62305
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.7920942446716425e-06,
+      "loss": 0.9108,
+      "step": 62310
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.490234375,
+      "learning_rate": 6.78302767861706e-06,
+      "loss": 0.8411,
+      "step": 62315
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.77396695550846e-06,
+      "loss": 0.8664,
+      "step": 62320
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.764912075913799e-06,
+      "loss": 0.8948,
+      "step": 62325
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.48828125,
+      "learning_rate": 6.755863040400612e-06,
+      "loss": 1.0311,
+      "step": 62330
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.7468198495361564e-06,
+      "loss": 0.9616,
+      "step": 62335
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.7377825038872135e-06,
+      "loss": 0.9264,
+      "step": 62340
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.728751004020284e-06,
+      "loss": 0.8163,
+      "step": 62345
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.7197253505014825e-06,
+      "loss": 0.8381,
+      "step": 62350
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.710705543896512e-06,
+      "loss": 1.0656,
+      "step": 62355
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.701691584770775e-06,
+      "loss": 0.8805,
+      "step": 62360
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.692683473689276e-06,
+      "loss": 1.0133,
+      "step": 62365
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.68368121121663e-06,
+      "loss": 0.8248,
+      "step": 62370
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.6746847979171525e-06,
+      "loss": 0.8632,
+      "step": 62375
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.665694234354691e-06,
+      "loss": 0.8974,
+      "step": 62380
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.451171875,
+      "learning_rate": 6.65670952109283e-06,
+      "loss": 0.8646,
+      "step": 62385
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.647730658694718e-06,
+      "loss": 0.8876,
+      "step": 62390
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.6387576477231595e-06,
+      "loss": 1.0241,
+      "step": 62395
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.629790488740617e-06,
+      "loss": 0.9139,
+      "step": 62400
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.620829182309129e-06,
+      "loss": 1.0862,
+      "step": 62405
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4921875,
+      "learning_rate": 6.611873728990425e-06,
+      "loss": 1.0256,
+      "step": 62410
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.602924129345855e-06,
+      "loss": 1.1855,
+      "step": 62415
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.593980383936349e-06,
+      "loss": 0.8099,
+      "step": 62420
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.585042493322535e-06,
+      "loss": 1.1324,
+      "step": 62425
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4765625,
+      "learning_rate": 6.576110458064677e-06,
+      "loss": 0.9345,
+      "step": 62430
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.567184278722582e-06,
+      "loss": 1.0149,
+      "step": 62435
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.558263955855792e-06,
+      "loss": 1.0576,
+      "step": 62440
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.71875,
+      "learning_rate": 6.549349490023448e-06,
+      "loss": 1.0319,
+      "step": 62445
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.540440881784305e-06,
+      "loss": 0.9467,
+      "step": 62450
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.484375,
+      "learning_rate": 6.531538131696757e-06,
+      "loss": 0.8775,
+      "step": 62455
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.522641240318872e-06,
+      "loss": 0.9048,
+      "step": 62460
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.51375020820828e-06,
+      "loss": 0.8971,
+      "step": 62465
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.5048650359223025e-06,
+      "loss": 0.8988,
+      "step": 62470
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.495985724017872e-06,
+      "loss": 0.9041,
+      "step": 62475
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.487112273051555e-06,
+      "loss": 0.8345,
+      "step": 62480
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.478244683579526e-06,
+      "loss": 1.0131,
+      "step": 62485
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.469382956157633e-06,
+      "loss": 0.9545,
+      "step": 62490
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.46052709134135e-06,
+      "loss": 0.8159,
+      "step": 62495
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.451677089685759e-06,
+      "loss": 0.872,
+      "step": 62500
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.442832951745581e-06,
+      "loss": 0.9304,
+      "step": 62505
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.433994678075195e-06,
+      "loss": 0.9571,
+      "step": 62510
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.42516226922858e-06,
+      "loss": 0.9523,
+      "step": 62515
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.416335725759359e-06,
+      "loss": 0.9097,
+      "step": 62520
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.4075150482208245e-06,
+      "loss": 1.0431,
+      "step": 62525
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.398700237165811e-06,
+      "loss": 0.9239,
+      "step": 62530
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.389891293146899e-06,
+      "loss": 0.942,
+      "step": 62535
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.381088216716202e-06,
+      "loss": 0.9218,
+      "step": 62540
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.3722910084255014e-06,
+      "loss": 0.8891,
+      "step": 62545
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.363499668826245e-06,
+      "loss": 0.9076,
+      "step": 62550
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.35471419846948e-06,
+      "loss": 1.1142,
+      "step": 62555
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.345934597905867e-06,
+      "loss": 0.9014,
+      "step": 62560
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.337160867685743e-06,
+      "loss": 0.9365,
+      "step": 62565
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.328393008359057e-06,
+      "loss": 0.9801,
+      "step": 62570
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.319631020475369e-06,
+      "loss": 0.9142,
+      "step": 62575
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.310874904583897e-06,
+      "loss": 0.9259,
+      "step": 62580
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.302124661233511e-06,
+      "loss": 1.1145,
+      "step": 62585
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.2933802909726615e-06,
+      "loss": 0.938,
+      "step": 62590
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.284641794349433e-06,
+      "loss": 0.8037,
+      "step": 62595
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.275909171911609e-06,
+      "loss": 0.9177,
+      "step": 62600
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.267182424206541e-06,
+      "loss": 0.9361,
+      "step": 62605
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.258461551781225e-06,
+      "loss": 0.9151,
+      "step": 62610
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.24974655518229e-06,
+      "loss": 1.0927,
+      "step": 62615
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.241037434956043e-06,
+      "loss": 1.11,
+      "step": 62620
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.2323341916483254e-06,
+      "loss": 0.8704,
+      "step": 62625
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.223636825804702e-06,
+      "loss": 0.7773,
+      "step": 62630
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.214945337970335e-06,
+      "loss": 0.929,
+      "step": 62635
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4921875,
+      "learning_rate": 6.2062597286900005e-06,
+      "loss": 0.8479,
+      "step": 62640
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.48828125,
+      "learning_rate": 6.197579998508118e-06,
+      "loss": 0.9114,
+      "step": 62645
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.188906147968776e-06,
+      "loss": 0.8494,
+      "step": 62650
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.180238177615616e-06,
+      "loss": 0.8771,
+      "step": 62655
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.171576087991981e-06,
+      "loss": 0.9365,
+      "step": 62660
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.1629198796408276e-06,
+      "loss": 1.0453,
+      "step": 62665
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.154269553104719e-06,
+      "loss": 0.9323,
+      "step": 62670
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.145625108925879e-06,
+      "loss": 0.8692,
+      "step": 62675
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.136986547646151e-06,
+      "loss": 1.0395,
+      "step": 62680
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.128353869807002e-06,
+      "loss": 0.8201,
+      "step": 62685
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.119727075949555e-06,
+      "loss": 0.9716,
+      "step": 62690
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.111106166614522e-06,
+      "loss": 0.8377,
+      "step": 62695
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.102491142342304e-06,
+      "loss": 0.9958,
+      "step": 62700
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.093882003672868e-06,
+      "loss": 1.0683,
+      "step": 62705
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.085278751145851e-06,
+      "loss": 1.0908,
+      "step": 62710
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.076681385300531e-06,
+      "loss": 0.9115,
+      "step": 62715
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.068089906675789e-06,
+      "loss": 1.0243,
+      "step": 62720
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.05950431581015e-06,
+      "loss": 0.9099,
+      "step": 62725
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.0509246132417705e-06,
+      "loss": 1.0154,
+      "step": 62730
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.042350799508434e-06,
+      "loss": 0.9946,
+      "step": 62735
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.0337828751475535e-06,
+      "loss": 0.9308,
+      "step": 62740
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.025220840696211e-06,
+      "loss": 0.9511,
+      "step": 62745
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.48828125,
+      "learning_rate": 6.0166646966910325e-06,
+      "loss": 0.8393,
+      "step": 62750
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.008114443668334e-06,
+      "loss": 0.873,
+      "step": 62755
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.999570082164096e-06,
+      "loss": 0.9249,
+      "step": 62760
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.991031612713849e-06,
+      "loss": 1.0535,
+      "step": 62765
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.47265625,
+      "learning_rate": 5.982499035852795e-06,
+      "loss": 1.0427,
+      "step": 62770
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.466796875,
+      "learning_rate": 5.9739723521158084e-06,
+      "loss": 0.9109,
+      "step": 62775
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.965451562037294e-06,
+      "loss": 0.916,
+      "step": 62780
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.95693666615138e-06,
+      "loss": 0.944,
+      "step": 62785
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.948427664991796e-06,
+      "loss": 0.9449,
+      "step": 62790
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.9399245590918805e-06,
+      "loss": 0.8104,
+      "step": 62795
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.931427348984608e-06,
+      "loss": 1.0679,
+      "step": 62800
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.922936035202598e-06,
+      "loss": 0.8095,
+      "step": 62805
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.9144506182781225e-06,
+      "loss": 0.953,
+      "step": 62810
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.905971098743013e-06,
+      "loss": 0.9414,
+      "step": 62815
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.897497477128811e-06,
+      "loss": 0.8741,
+      "step": 62820
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.889029753966646e-06,
+      "loss": 0.9239,
+      "step": 62825
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.880567929787273e-06,
+      "loss": 0.8671,
+      "step": 62830
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.872112005121089e-06,
+      "loss": 0.9372,
+      "step": 62835
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.863661980498137e-06,
+      "loss": 0.887,
+      "step": 62840
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.855217856448058e-06,
+      "loss": 1.017,
+      "step": 62845
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.846779633500155e-06,
+      "loss": 0.8264,
+      "step": 62850
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.51171875,
+      "learning_rate": 5.8383473121833455e-06,
+      "loss": 0.9504,
+      "step": 62855
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.829920893026142e-06,
+      "loss": 1.0121,
+      "step": 62860
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.625,
+      "learning_rate": 5.8215003765567545e-06,
+      "loss": 0.9497,
+      "step": 62865
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.813085763302994e-06,
+      "loss": 0.8942,
+      "step": 62870
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.45703125,
+      "learning_rate": 5.804677053792284e-06,
+      "loss": 0.9689,
+      "step": 62875
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.796274248551681e-06,
+      "loss": 0.9884,
+      "step": 62880
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.787877348107918e-06,
+      "loss": 0.9302,
+      "step": 62885
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.779486352987285e-06,
+      "loss": 1.1086,
+      "step": 62890
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.771101263715761e-06,
+      "loss": 0.9932,
+      "step": 62895
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.762722080818939e-06,
+      "loss": 1.0333,
+      "step": 62900
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.625,
+      "learning_rate": 5.754348804822018e-06,
+      "loss": 0.9,
+      "step": 62905
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.745981436249847e-06,
+      "loss": 0.8931,
+      "step": 62910
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.737619975626907e-06,
+      "loss": 0.8853,
+      "step": 62915
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.76953125,
+      "learning_rate": 5.7292644234773096e-06,
+      "loss": 0.9668,
+      "step": 62920
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.720914780324771e-06,
+      "loss": 0.9179,
+      "step": 62925
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.712571046692661e-06,
+      "loss": 0.9579,
+      "step": 62930
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.704233223104005e-06,
+      "loss": 1.0306,
+      "step": 62935
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.695901310081386e-06,
+      "loss": 0.9032,
+      "step": 62940
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.687575308147086e-06,
+      "loss": 0.9209,
+      "step": 62945
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.679255217822987e-06,
+      "loss": 0.9953,
+      "step": 62950
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.670941039630595e-06,
+      "loss": 0.9111,
+      "step": 62955
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.662632774091026e-06,
+      "loss": 0.9192,
+      "step": 62960
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.654330421725085e-06,
+      "loss": 0.8808,
+      "step": 62965
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.646033983053178e-06,
+      "loss": 1.0373,
+      "step": 62970
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.637743458595302e-06,
+      "loss": 0.9083,
+      "step": 62975
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.6294588488711385e-06,
+      "loss": 0.9685,
+      "step": 62980
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.621180154399996e-06,
+      "loss": 0.8314,
+      "step": 62985
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.612907375700749e-06,
+      "loss": 0.9279,
+      "step": 62990
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.60464051329197e-06,
+      "loss": 0.7895,
+      "step": 62995
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.596379567691834e-06,
+      "loss": 1.0582,
+      "step": 63000
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.588124539418127e-06,
+      "loss": 1.0279,
+      "step": 63005
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4921875,
+      "learning_rate": 5.579875428988324e-06,
+      "loss": 0.9399,
+      "step": 63010
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.571632236919466e-06,
+      "loss": 1.0019,
+      "step": 63015
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.625,
+      "learning_rate": 5.563394963728219e-06,
+      "loss": 1.0214,
+      "step": 63020
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.555163609930947e-06,
+      "loss": 0.9731,
+      "step": 63025
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.546938176043581e-06,
+      "loss": 0.9859,
+      "step": 63030
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.538718662581699e-06,
+      "loss": 0.9335,
+      "step": 63035
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.53050507006051e-06,
+      "loss": 0.9076,
+      "step": 63040
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.522297398994869e-06,
+      "loss": 0.9375,
+      "step": 63045
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.51409564989922e-06,
+      "loss": 0.9551,
+      "step": 63050
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.505899823287663e-06,
+      "loss": 1.0059,
+      "step": 63055
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.703125,
+      "learning_rate": 5.4977099196739324e-06,
+      "loss": 1.0367,
+      "step": 63060
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.489525939571383e-06,
+      "loss": 0.9067,
+      "step": 63065
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.48134788349296e-06,
+      "loss": 0.9347,
+      "step": 63070
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.47317575195131e-06,
+      "loss": 0.866,
+      "step": 63075
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.498046875,
+      "learning_rate": 5.465009545458666e-06,
+      "loss": 0.8887,
+      "step": 63080
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.456849264526887e-06,
+      "loss": 0.9003,
+      "step": 63085
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.448694909667462e-06,
+      "loss": 1.0918,
+      "step": 63090
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.4405464813915395e-06,
+      "loss": 0.9515,
+      "step": 63095
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.478515625,
+      "learning_rate": 5.4324039802098544e-06,
+      "loss": 1.002,
+      "step": 63100
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.424267406632777e-06,
+      "loss": 0.9115,
+      "step": 63105
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.416136761170354e-06,
+      "loss": 0.9534,
+      "step": 63110
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.4080120443322e-06,
+      "loss": 0.7717,
+      "step": 63115
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.399893256627564e-06,
+      "loss": 1.0774,
+      "step": 63120
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.48828125,
+      "learning_rate": 5.391780398565383e-06,
+      "loss": 0.8528,
+      "step": 63125
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.3836734706541385e-06,
+      "loss": 1.0441,
+      "step": 63130
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.375572473401991e-06,
+      "loss": 1.0177,
+      "step": 63135
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.367477407316745e-06,
+      "loss": 0.927,
+      "step": 63140
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.359388272905785e-06,
+      "loss": 0.9415,
+      "step": 63145
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.35130507067616e-06,
+      "loss": 1.005,
+      "step": 63150
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.343227801134532e-06,
+      "loss": 0.8438,
+      "step": 63155
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.335156464787183e-06,
+      "loss": 0.8493,
+      "step": 63160
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.51171875,
+      "learning_rate": 5.3270910621400435e-06,
+      "loss": 1.1147,
+      "step": 63165
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.319031593698653e-06,
+      "loss": 0.9504,
+      "step": 63170
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.310978059968219e-06,
+      "loss": 0.9684,
+      "step": 63175
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.302930461453492e-06,
+      "loss": 0.8936,
+      "step": 63180
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.294888798658948e-06,
+      "loss": 0.9478,
+      "step": 63185
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.286853072088638e-06,
+      "loss": 1.0088,
+      "step": 63190
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.278823282246237e-06,
+      "loss": 0.9158,
+      "step": 63195
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.270799429635065e-06,
+      "loss": 0.904,
+      "step": 63200
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.262781514758097e-06,
+      "loss": 1.006,
+      "step": 63205
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.254769538117854e-06,
+      "loss": 0.9891,
+      "step": 63210
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.246763500216578e-06,
+      "loss": 0.9617,
+      "step": 63215
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.23876340155609e-06,
+      "loss": 0.9842,
+      "step": 63220
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.2307692426378226e-06,
+      "loss": 0.948,
+      "step": 63225
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.2227810239628635e-06,
+      "loss": 0.9327,
+      "step": 63230
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.214798746031957e-06,
+      "loss": 0.9766,
+      "step": 63235
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.206822409345391e-06,
+      "loss": 0.9396,
+      "step": 63240
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.198852014403166e-06,
+      "loss": 1.098,
+      "step": 63245
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.190887561704871e-06,
+      "loss": 0.8943,
+      "step": 63250
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.182929051749708e-06,
+      "loss": 0.9164,
+      "step": 63255
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.174976485036542e-06,
+      "loss": 0.9736,
+      "step": 63260
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.167029862063865e-06,
+      "loss": 1.0046,
+      "step": 63265
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.159089183329757e-06,
+      "loss": 0.8295,
+      "step": 63270
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.1511544493319515e-06,
+      "loss": 0.925,
+      "step": 63275
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.143225660567796e-06,
+      "loss": 1.0085,
+      "step": 63280
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.135302817534304e-06,
+      "loss": 1.0287,
+      "step": 63285
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.127385920728067e-06,
+      "loss": 0.9187,
+      "step": 63290
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.119474970645322e-06,
+      "loss": 0.8212,
+      "step": 63295
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.111569967781959e-06,
+      "loss": 1.0078,
+      "step": 63300
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.51953125,
+      "learning_rate": 5.103670912633451e-06,
+      "loss": 0.9496,
+      "step": 63305
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.482421875,
+      "learning_rate": 5.095777805694935e-06,
+      "loss": 0.9652,
+      "step": 63310
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5,
+      "learning_rate": 5.0878906474611574e-06,
+      "loss": 0.922,
+      "step": 63315
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.0800094384264694e-06,
+      "loss": 1.0818,
+      "step": 63320
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.07213417908492e-06,
+      "loss": 0.8417,
+      "step": 63325
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.625,
+      "learning_rate": 5.064264869930113e-06,
+      "loss": 0.9057,
+      "step": 63330
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.056401511455288e-06,
+      "loss": 0.9798,
+      "step": 63335
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.048544104153352e-06,
+      "loss": 0.982,
+      "step": 63340
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.040692648516821e-06,
+      "loss": 0.927,
+      "step": 63345
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.032847145037811e-06,
+      "loss": 1.0344,
+      "step": 63350
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.025007594208109e-06,
+      "loss": 0.8876,
+      "step": 63355
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.0171739965191085e-06,
+      "loss": 0.9278,
+      "step": 63360
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.0093463524617965e-06,
+      "loss": 0.8402,
+      "step": 63365
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.001524662526846e-06,
+      "loss": 0.9543,
+      "step": 63370
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.993708927204543e-06,
+      "loss": 1.1113,
+      "step": 63375
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.985899146984762e-06,
+      "loss": 0.9639,
+      "step": 63380
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.978095322357024e-06,
+      "loss": 1.0694,
+      "step": 63385
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.97029745381048e-06,
+      "loss": 0.732,
+      "step": 63390
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.9625055418339505e-06,
+      "loss": 1.0459,
+      "step": 63395
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.954719586915791e-06,
+      "loss": 0.9727,
+      "step": 63400
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.946939589544053e-06,
+      "loss": 0.9419,
+      "step": 63405
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.939165550206415e-06,
+      "loss": 0.7805,
+      "step": 63410
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.931397469390131e-06,
+      "loss": 0.9549,
+      "step": 63415
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.923635347582134e-06,
+      "loss": 1.0289,
+      "step": 63420
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.915879185268968e-06,
+      "loss": 0.8387,
+      "step": 63425
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.908128982936777e-06,
+      "loss": 0.9095,
+      "step": 63430
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.900384741071362e-06,
+      "loss": 0.988,
+      "step": 63435
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.892646460158146e-06,
+      "loss": 0.9183,
+      "step": 63440
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.884914140682151e-06,
+      "loss": 0.9776,
+      "step": 63445
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.877187783128068e-06,
+      "loss": 0.9776,
+      "step": 63450
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.869467387980209e-06,
+      "loss": 1.0015,
+      "step": 63455
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.861752955722454e-06,
+      "loss": 0.7547,
+      "step": 63460
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.8540444868383935e-06,
+      "loss": 1.2453,
+      "step": 63465
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.846341981811187e-06,
+      "loss": 0.8749,
+      "step": 63470
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.838645441123623e-06,
+      "loss": 1.0333,
+      "step": 63475
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.830954865258164e-06,
+      "loss": 0.931,
+      "step": 63480
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.823270254696821e-06,
+      "loss": 0.9601,
+      "step": 63485
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.815591609921322e-06,
+      "loss": 0.8998,
+      "step": 63490
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.807918931412914e-06,
+      "loss": 1.0241,
+      "step": 63495
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.800252219652579e-06,
+      "loss": 0.9631,
+      "step": 63500
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.792591475120867e-06,
+      "loss": 1.09,
+      "step": 63505
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.784936698297937e-06,
+      "loss": 0.9044,
+      "step": 63510
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.777287889663618e-06,
+      "loss": 0.9348,
+      "step": 63515
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.7696450496973464e-06,
+      "loss": 0.9964,
+      "step": 63520
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.762008178878185e-06,
+      "loss": 0.8946,
+      "step": 63525
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.754377277684807e-06,
+      "loss": 1.025,
+      "step": 63530
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.746752346595562e-06,
+      "loss": 0.9254,
+      "step": 63535
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.739133386088345e-06,
+      "loss": 1.1481,
+      "step": 63540
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.7315203966407425e-06,
+      "loss": 1.0514,
+      "step": 63545
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.723913378729949e-06,
+      "loss": 1.0534,
+      "step": 63550
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.716312332832762e-06,
+      "loss": 0.9579,
+      "step": 63555
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.708717259425644e-06,
+      "loss": 0.9386,
+      "step": 63560
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.701128158984658e-06,
+      "loss": 0.8514,
+      "step": 63565
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.693545031985491e-06,
+      "loss": 0.9786,
+      "step": 63570
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.685967878903463e-06,
+      "loss": 0.8708,
+      "step": 63575
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.678396700213539e-06,
+      "loss": 0.9327,
+      "step": 63580
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.670831496390271e-06,
+      "loss": 0.9573,
+      "step": 63585
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.6632722679078366e-06,
+      "loss": 0.9043,
+      "step": 63590
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.65571901524009e-06,
+      "loss": 0.9097,
+      "step": 63595
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.648171738860463e-06,
+      "loss": 1.0028,
+      "step": 63600
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.640630439242022e-06,
+      "loss": 0.8418,
+      "step": 63605
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.633095116857467e-06,
+      "loss": 0.9803,
+      "step": 63610
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.62556577217913e-06,
+      "loss": 0.859,
+      "step": 63615
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.618042405678946e-06,
+      "loss": 0.9136,
+      "step": 63620
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.610525017828504e-06,
+      "loss": 0.9965,
+      "step": 63625
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.603013609098994e-06,
+      "loss": 0.8447,
+      "step": 63630
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.595508179961228e-06,
+      "loss": 1.0054,
+      "step": 63635
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.588008730885685e-06,
+      "loss": 0.9298,
+      "step": 63640
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.5805152623424e-06,
+      "loss": 0.9792,
+      "step": 63645
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.573027774801109e-06,
+      "loss": 0.8879,
+      "step": 63650
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.565546268731102e-06,
+      "loss": 0.9782,
+      "step": 63655
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.55807074460135e-06,
+      "loss": 0.885,
+      "step": 63660
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.550601202880433e-06,
+      "loss": 0.867,
+      "step": 63665
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.543137644036533e-06,
+      "loss": 1.1561,
+      "step": 63670
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.5356800685374955e-06,
+      "loss": 0.9632,
+      "step": 63675
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.5282284768507375e-06,
+      "loss": 0.9658,
+      "step": 63680
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.520782869443352e-06,
+      "loss": 0.9817,
+      "step": 63685
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.513343246782043e-06,
+      "loss": 0.8341,
+      "step": 63690
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.505909609333147e-06,
+      "loss": 0.8791,
+      "step": 63695
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.498481957562573e-06,
+      "loss": 0.9347,
+      "step": 63700
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.491060291935911e-06,
+      "loss": 0.8894,
+      "step": 63705
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.4836446129183914e-06,
+      "loss": 0.8209,
+      "step": 63710
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.476234920974787e-06,
+      "loss": 0.9218,
+      "step": 63715
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.46883121656958e-06,
+      "loss": 0.8734,
+      "step": 63720
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.461433500166834e-06,
+      "loss": 0.9371,
+      "step": 63725
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.454041772230244e-06,
+      "loss": 1.0467,
+      "step": 63730
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.446656033223129e-06,
+      "loss": 0.9309,
+      "step": 63735
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.439276283608451e-06,
+      "loss": 0.9906,
+      "step": 63740
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.431902523848774e-06,
+      "loss": 0.8913,
+      "step": 63745
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.424534754406273e-06,
+      "loss": 0.8362,
+      "step": 63750
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.417172975742789e-06,
+      "loss": 0.8761,
+      "step": 63755
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.409817188319776e-06,
+      "loss": 0.9338,
+      "step": 63760
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.4024673925982755e-06,
+      "loss": 0.9317,
+      "step": 63765
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.395123589038996e-06,
+      "loss": 1.0069,
+      "step": 63770
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.38778577810226e-06,
+      "loss": 1.0037,
+      "step": 63775
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.380453960247999e-06,
+      "loss": 0.8736,
+      "step": 63780
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.373128135935789e-06,
+      "loss": 1.0341,
+      "step": 63785
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.365808305624819e-06,
+      "loss": 1.0015,
+      "step": 63790
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.35849446977391e-06,
+      "loss": 0.9519,
+      "step": 63795
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.351186628841486e-06,
+      "loss": 0.9249,
+      "step": 63800
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.343884783285623e-06,
+      "loss": 0.9506,
+      "step": 63805
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.33658893356399e-06,
+      "loss": 0.9668,
+      "step": 63810
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.3292990801339196e-06,
+      "loss": 0.8758,
+      "step": 63815
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.322015223452358e-06,
+      "loss": 1.2518,
+      "step": 63820
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.314737363975829e-06,
+      "loss": 0.8206,
+      "step": 63825
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.307465502160546e-06,
+      "loss": 0.9852,
+      "step": 63830
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.300199638462321e-06,
+      "loss": 0.9431,
+      "step": 63835
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.292939773336569e-06,
+      "loss": 0.9671,
+      "step": 63840
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.285685907238346e-06,
+      "loss": 1.0574,
+      "step": 63845
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.278438040622346e-06,
+      "loss": 0.9514,
+      "step": 63850
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.271196173942882e-06,
+      "loss": 0.8474,
+      "step": 63855
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.263960307653847e-06,
+      "loss": 0.9748,
+      "step": 63860
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.256730442208812e-06,
+      "loss": 1.079,
+      "step": 63865
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.249506578060969e-06,
+      "loss": 0.8845,
+      "step": 63870
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.625,
+      "learning_rate": 4.2422887156630894e-06,
+      "loss": 1.0111,
+      "step": 63875
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5,
+      "learning_rate": 4.235076855467623e-06,
+      "loss": 1.1052,
+      "step": 63880
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.227870997926609e-06,
+      "loss": 0.8858,
+      "step": 63885
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.220671143491705e-06,
+      "loss": 0.944,
+      "step": 63890
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.213477292614221e-06,
+      "loss": 1.1542,
+      "step": 63895
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.206289445745093e-06,
+      "loss": 0.8764,
+      "step": 63900
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.199107603334818e-06,
+      "loss": 0.9639,
+      "step": 63905
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.19193176583359e-06,
+      "loss": 0.9392,
+      "step": 63910
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.184761933691206e-06,
+      "loss": 1.0341,
+      "step": 63915
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.177598107357061e-06,
+      "loss": 1.0983,
+      "step": 63920
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.170440287280186e-06,
+      "loss": 0.9701,
+      "step": 63925
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.163288473909277e-06,
+      "loss": 0.9325,
+      "step": 63930
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.1561426676925864e-06,
+      "loss": 0.9986,
+      "step": 63935
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.1490028690780225e-06,
+      "loss": 0.9749,
+      "step": 63940
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.1418690785131384e-06,
+      "loss": 0.9181,
+      "step": 63945
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.134741296445055e-06,
+      "loss": 1.0318,
+      "step": 63950
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.127619523320592e-06,
+      "loss": 0.9893,
+      "step": 63955
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.120503759586103e-06,
+      "loss": 1.0783,
+      "step": 63960
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.113394005687654e-06,
+      "loss": 0.9217,
+      "step": 63965
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.106290262070867e-06,
+      "loss": 1.029,
+      "step": 63970
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.099192529181018e-06,
+      "loss": 1.0799,
+      "step": 63975
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.092100807463017e-06,
+      "loss": 0.9321,
+      "step": 63980
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.085015097361344e-06,
+      "loss": 0.9397,
+      "step": 63985
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.0779353993201765e-06,
+      "loss": 0.9828,
+      "step": 63990
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.07086171378327e-06,
+      "loss": 0.8303,
+      "step": 63995
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.0637940411939916e-06,
+      "loss": 0.9768,
+      "step": 64000
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.0567323819953765e-06,
+      "loss": 1.1028,
+      "step": 64005
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.049676736630048e-06,
+      "loss": 0.8675,
+      "step": 64010
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.042627105540253e-06,
+      "loss": 1.008,
+      "step": 64015
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.0355834891678714e-06,
+      "loss": 0.84,
+      "step": 64020
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.028545887954416e-06,
+      "loss": 0.9896,
+      "step": 64025
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.021514302341012e-06,
+      "loss": 1.0693,
+      "step": 64030
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.014488732768385e-06,
+      "loss": 1.0449,
+      "step": 64035
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.007469179676948e-06,
+      "loss": 0.9926,
+      "step": 64040
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.000455643506651e-06,
+      "loss": 0.8867,
+      "step": 64045
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.99344812469713e-06,
+      "loss": 0.8337,
+      "step": 64050
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.986446623687623e-06,
+      "loss": 0.9334,
+      "step": 64055
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.979451140917012e-06,
+      "loss": 1.0358,
+      "step": 64060
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.972461676823735e-06,
+      "loss": 0.996,
+      "step": 64065
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.965478231845932e-06,
+      "loss": 0.9938,
+      "step": 64070
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.958500806421339e-06,
+      "loss": 0.9653,
+      "step": 64075
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.9515294009872865e-06,
+      "loss": 0.9159,
+      "step": 64080
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.9445640159807565e-06,
+      "loss": 0.8874,
+      "step": 64085
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.937604651838367e-06,
+      "loss": 1.082,
+      "step": 64090
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.9306513089963135e-06,
+      "loss": 0.9247,
+      "step": 64095
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.923703987890448e-06,
+      "loss": 0.9998,
+      "step": 64100
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.916762688956255e-06,
+      "loss": 1.0188,
+      "step": 64105
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.909827412628797e-06,
+      "loss": 0.8621,
+      "step": 64110
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.9028981593427935e-06,
+      "loss": 0.9959,
+      "step": 64115
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.895974929532587e-06,
+      "loss": 0.9092,
+      "step": 64120
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.889057723632106e-06,
+      "loss": 0.9389,
+      "step": 64125
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5,
+      "learning_rate": 3.88214654207496e-06,
+      "loss": 0.8261,
+      "step": 64130
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.875241385294337e-06,
+      "loss": 1.281,
+      "step": 64135
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.868342253723056e-06,
+      "loss": 0.9396,
+      "step": 64140
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.861449147793561e-06,
+      "loss": 0.9935,
+      "step": 64145
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.85456206793795e-06,
+      "loss": 0.8586,
+      "step": 64150
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.8476810145878675e-06,
+      "loss": 0.957,
+      "step": 64155
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.840805988174656e-06,
+      "loss": 0.8097,
+      "step": 64160
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.833936989129239e-06,
+      "loss": 0.9352,
+      "step": 64165
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.827074017882193e-06,
+      "loss": 0.8878,
+      "step": 64170
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.820217074863652e-06,
+      "loss": 1.034,
+      "step": 64175
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.625,
+      "learning_rate": 3.813366160503451e-06,
+      "loss": 0.9547,
+      "step": 64180
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.8065212752310232e-06,
+      "loss": 0.8868,
+      "step": 64185
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.79968241947537e-06,
+      "loss": 0.9713,
+      "step": 64190
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.7928495936652043e-06,
+      "loss": 1.0498,
+      "step": 64195
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.7860227982287943e-06,
+      "loss": 1.0994,
+      "step": 64200
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.7792020335940537e-06,
+      "loss": 1.0165,
+      "step": 64205
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.772387300188507e-06,
+      "loss": 1.0315,
+      "step": 64210
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.7655785984393455e-06,
+      "loss": 0.9926,
+      "step": 64215
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.7587759287732948e-06,
+      "loss": 1.0681,
+      "step": 64220
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.47265625,
+      "learning_rate": 3.7519792916167808e-06,
+      "loss": 0.9187,
+      "step": 64225
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.7451886873958285e-06,
+      "loss": 0.9417,
+      "step": 64230
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.738404116536065e-06,
+      "loss": 0.92,
+      "step": 64235
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.731625579462761e-06,
+      "loss": 1.0117,
+      "step": 64240
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.7248530766008315e-06,
+      "loss": 1.0202,
+      "step": 64245
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.7180866083747377e-06,
+      "loss": 0.9853,
+      "step": 64250
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5,
+      "learning_rate": 3.7113261752086294e-06,
+      "loss": 0.934,
+      "step": 64255
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.7045717775262777e-06,
+      "loss": 0.9742,
+      "step": 64260
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.697823415751023e-06,
+      "loss": 0.9932,
+      "step": 64265
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.6910810903058813e-06,
+      "loss": 0.9429,
+      "step": 64270
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.6843448016134596e-06,
+      "loss": 0.9631,
+      "step": 64275
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.6776145500960093e-06,
+      "loss": 0.9709,
+      "step": 64280
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.67089033617537e-06,
+      "loss": 0.8357,
+      "step": 64285
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.6641721602730273e-06,
+      "loss": 1.0281,
+      "step": 64290
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.657460022810111e-06,
+      "loss": 0.8819,
+      "step": 64295
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.6507539242073172e-06,
+      "loss": 0.9193,
+      "step": 64300
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.644053864884989e-06,
+      "loss": 0.9544,
+      "step": 64305
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.637359845263122e-06,
+      "loss": 0.9234,
+      "step": 64310
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.6306718657612816e-06,
+      "loss": 0.9635,
+      "step": 64315
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.6239899267986877e-06,
+      "loss": 1.0724,
+      "step": 64320
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.617314028794161e-06,
+      "loss": 0.8347,
+      "step": 64325
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.6106441721661666e-06,
+      "loss": 0.904,
+      "step": 64330
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.6039803573327704e-06,
+      "loss": 0.9772,
+      "step": 64335
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.597322584711682e-06,
+      "loss": 1.1015,
+      "step": 64340
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.5906708547201907e-06,
+      "loss": 0.8824,
+      "step": 64345
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.5840251677752514e-06,
+      "loss": 0.9461,
+      "step": 64350
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.5773855242934417e-06,
+      "loss": 0.9427,
+      "step": 64355
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.5707519246909073e-06,
+      "loss": 0.8533,
+      "step": 64360
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.56412436938347e-06,
+      "loss": 1.1611,
+      "step": 64365
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.5575028587865546e-06,
+      "loss": 1.0923,
+      "step": 64370
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.550887393315205e-06,
+      "loss": 1.052,
+      "step": 64375
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.544277973384058e-06,
+      "loss": 0.8655,
+      "step": 64380
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.5376745994074257e-06,
+      "loss": 0.8928,
+      "step": 64385
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.5310772717992215e-06,
+      "loss": 0.907,
+      "step": 64390
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.524485990972959e-06,
+      "loss": 0.957,
+      "step": 64395
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.5179007573417853e-06,
+      "loss": 1.0026,
+      "step": 64400
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.5113215713184934e-06,
+      "loss": 1.0308,
+      "step": 64405
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.5047484333154303e-06,
+      "loss": 1.0068,
+      "step": 64410
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.498181343744644e-06,
+      "loss": 0.8248,
+      "step": 64415
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.491620303017773e-06,
+      "loss": 0.9557,
+      "step": 64420
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.485065311546054e-06,
+      "loss": 1.0306,
+      "step": 64425
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.4785163697403477e-06,
+      "loss": 0.9902,
+      "step": 64430
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.471973478011159e-06,
+      "loss": 1.0089,
+      "step": 64435
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.4654366367686263e-06,
+      "loss": 0.9222,
+      "step": 64440
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.4589058464224554e-06,
+      "loss": 0.8768,
+      "step": 64445
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.452381107382019e-06,
+      "loss": 0.9449,
+      "step": 64450
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.4458624200563116e-06,
+      "loss": 0.9348,
+      "step": 64455
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.4393497848538956e-06,
+      "loss": 0.8981,
+      "step": 64460
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.48046875,
+      "learning_rate": 3.4328432021830004e-06,
+      "loss": 0.8987,
+      "step": 64465
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.426342672451499e-06,
+      "loss": 1.2103,
+      "step": 64470
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.4198481960668103e-06,
+      "loss": 0.9515,
+      "step": 64475
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.4133597734360423e-06,
+      "loss": 1.0422,
+      "step": 64480
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.4068774049658913e-06,
+      "loss": 0.9403,
+      "step": 64485
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.400401091062655e-06,
+      "loss": 0.9389,
+      "step": 64490
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.393930832132297e-06,
+      "loss": 0.9773,
+      "step": 64495
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.3874666285803936e-06,
+      "loss": 0.9369,
+      "step": 64500
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.381008480812109e-06,
+      "loss": 0.8346,
+      "step": 64505
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.374556389232253e-06,
+      "loss": 0.8604,
+      "step": 64510
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.368110354245246e-06,
+      "loss": 0.9006,
+      "step": 64515
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.3616703762551437e-06,
+      "loss": 1.0335,
+      "step": 64520
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.355236455665589e-06,
+      "loss": 0.9049,
+      "step": 64525
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.348808592879904e-06,
+      "loss": 0.9769,
+      "step": 64530
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.3423867883009664e-06,
+      "loss": 0.979,
+      "step": 64535
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.466796875,
+      "learning_rate": 3.3359710423312984e-06,
+      "loss": 0.9246,
+      "step": 64540
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.3295613553730566e-06,
+      "loss": 1.0752,
+      "step": 64545
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.323157727828019e-06,
+      "loss": 1.1276,
+      "step": 64550
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.316760160097554e-06,
+      "loss": 1.0467,
+      "step": 64555
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.310368652582674e-06,
+      "loss": 0.9048,
+      "step": 64560
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.303983205684014e-06,
+      "loss": 0.8651,
+      "step": 64565
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.297603819801809e-06,
+      "loss": 0.9898,
+      "step": 64570
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.2912304953359173e-06,
+      "loss": 0.8427,
+      "step": 64575
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.2848632326858643e-06,
+      "loss": 1.0477,
+      "step": 64580
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.27850203225073e-06,
+      "loss": 1.0262,
+      "step": 64585
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.27214689442924e-06,
+      "loss": 0.9684,
+      "step": 64590
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.2657978196197426e-06,
+      "loss": 1.0751,
+      "step": 64595
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.259454808220208e-06,
+      "loss": 1.0147,
+      "step": 64600
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.2531178606282297e-06,
+      "loss": 1.0055,
+      "step": 64605
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.246786977241012e-06,
+      "loss": 1.0449,
+      "step": 64610
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.2404621584553707e-06,
+      "loss": 1.0597,
+      "step": 64615
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.2341434046677554e-06,
+      "loss": 0.9698,
+      "step": 64620
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.2278307162742607e-06,
+      "loss": 1.0122,
+      "step": 64625
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.221524093670536e-06,
+      "loss": 0.989,
+      "step": 64630
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.21522353725191e-06,
+      "loss": 0.9688,
+      "step": 64635
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.2089290474132893e-06,
+      "loss": 0.8625,
+      "step": 64640
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.458984375,
+      "learning_rate": 3.2026406245492467e-06,
+      "loss": 0.8252,
+      "step": 64645
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.1963582690539117e-06,
+      "loss": 0.9317,
+      "step": 64650
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.49609375,
+      "learning_rate": 3.1900819813211025e-06,
+      "loss": 0.9196,
+      "step": 64655
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.1838117617442266e-06,
+      "loss": 0.9512,
+      "step": 64660
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.1775476107162695e-06,
+      "loss": 0.9112,
+      "step": 64665
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.171289528629906e-06,
+      "loss": 0.9008,
+      "step": 64670
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.1650375158774004e-06,
+      "loss": 0.9157,
+      "step": 64675
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.158791572850628e-06,
+      "loss": 0.8422,
+      "step": 64680
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.1525516999410866e-06,
+      "loss": 0.9628,
+      "step": 64685
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.14631789753993e-06,
+      "loss": 1.0034,
+      "step": 64690
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.1400901660378458e-06,
+      "loss": 1.0796,
+      "step": 64695
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.1338685058252216e-06,
+      "loss": 0.8844,
+      "step": 64700
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.1276529172920675e-06,
+      "loss": 0.9305,
+      "step": 64705
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.121443400827928e-06,
+      "loss": 1.0306,
+      "step": 64710
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.115239956822058e-06,
+      "loss": 0.8335,
+      "step": 64715
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.109042585663313e-06,
+      "loss": 0.9649,
+      "step": 64720
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.102851287740105e-06,
+      "loss": 0.9282,
+      "step": 64725
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.0966660634405344e-06,
+      "loss": 0.681,
+      "step": 64730
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.0904869131523128e-06,
+      "loss": 0.8902,
+      "step": 64735
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.0843138372627424e-06,
+      "loss": 1.1055,
+      "step": 64740
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.078146836158746e-06,
+      "loss": 0.9402,
+      "step": 64745
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.0719859102268933e-06,
+      "loss": 0.8482,
+      "step": 64750
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.0658310598533633e-06,
+      "loss": 1.1807,
+      "step": 64755
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.059682285423926e-06,
+      "loss": 0.9468,
+      "step": 64760
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.053539587324017e-06,
+      "loss": 0.9515,
+      "step": 64765
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.047402965938673e-06,
+      "loss": 0.9918,
+      "step": 64770
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.041272421652508e-06,
+      "loss": 1.0043,
+      "step": 64775
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.035147954849826e-06,
+      "loss": 0.98,
+      "step": 64780
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.0290295659145094e-06,
+      "loss": 1.0729,
+      "step": 64785
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.02291725523004e-06,
+      "loss": 0.9641,
+      "step": 64790
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.0168110231795887e-06,
+      "loss": 1.0132,
+      "step": 64795
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.010710870145872e-06,
+      "loss": 0.9996,
+      "step": 64800
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.0046167965112504e-06,
+      "loss": 1.2168,
+      "step": 64805
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5,
+      "learning_rate": 2.9985288026577184e-06,
+      "loss": 0.8434,
+      "step": 64810
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.9924468889668824e-06,
+      "loss": 0.8563,
+      "step": 64815
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.9863710558199474e-06,
+      "loss": 1.0859,
+      "step": 64820
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.980301303597777e-06,
+      "loss": 1.0357,
+      "step": 64825
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.9742376326808095e-06,
+      "loss": 0.8976,
+      "step": 64830
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.9681800434491315e-06,
+      "loss": 1.0682,
+      "step": 64835
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.9621285362824268e-06,
+      "loss": 0.8746,
+      "step": 64840
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.9560831115600486e-06,
+      "loss": 0.9363,
+      "step": 64845
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.9500437696608928e-06,
+      "loss": 0.9036,
+      "step": 64850
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.944010510963513e-06,
+      "loss": 0.9912,
+      "step": 64855
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5,
+      "learning_rate": 2.937983335846095e-06,
+      "loss": 0.797,
+      "step": 64860
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.931962244686437e-06,
+      "loss": 1.0168,
+      "step": 64865
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.925947237861926e-06,
+      "loss": 1.0282,
+      "step": 64870
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.919938315749604e-06,
+      "loss": 0.974,
+      "step": 64875
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.9139354787261265e-06,
+      "loss": 1.0569,
+      "step": 64880
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.625,
+      "learning_rate": 2.907938727167725e-06,
+      "loss": 1.0904,
+      "step": 64885
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.48828125,
+      "learning_rate": 2.901948061450299e-06,
+      "loss": 1.0113,
+      "step": 64890
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.895963481949371e-06,
+      "loss": 0.9074,
+      "step": 64895
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.889984989040051e-06,
+      "loss": 0.9813,
+      "step": 64900
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.884012583097051e-06,
+      "loss": 0.8772,
+      "step": 64905
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.8780462644947603e-06,
+      "loss": 0.923,
+      "step": 64910
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.8720860336071355e-06,
+      "loss": 1.0726,
+      "step": 64915
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.866131890807777e-06,
+      "loss": 0.9847,
+      "step": 64920
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.8601838364699097e-06,
+      "loss": 0.9785,
+      "step": 64925
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.8542418709663343e-06,
+      "loss": 0.9507,
+      "step": 64930
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.84830599466952e-06,
+      "loss": 0.8207,
+      "step": 64935
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.842376207951547e-06,
+      "loss": 0.9632,
+      "step": 64940
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.8364525111840735e-06,
+      "loss": 1.0299,
+      "step": 64945
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.8305349047384355e-06,
+      "loss": 0.9159,
+      "step": 64950
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.8246233889855145e-06,
+      "loss": 0.9085,
+      "step": 64955
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.8187179642958916e-06,
+      "loss": 0.9428,
+      "step": 64960
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.812818631039682e-06,
+      "loss": 0.9958,
+      "step": 64965
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5,
+      "learning_rate": 2.806925389586701e-06,
+      "loss": 0.8781,
+      "step": 64970
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.801038240306342e-06,
+      "loss": 0.8886,
+      "step": 64975
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.795157183567587e-06,
+      "loss": 0.9409,
+      "step": 64980
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.625,
+      "learning_rate": 2.7892822197390978e-06,
+      "loss": 1.0129,
+      "step": 64985
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.7834133491891123e-06,
+      "loss": 0.8587,
+      "step": 64990
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.777550572285492e-06,
+      "loss": 0.9335,
+      "step": 64995
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.7716938893957323e-06,
+      "loss": 1.0092,
+      "step": 65000
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.765843300886939e-06,
+      "loss": 0.9026,
+      "step": 65005
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.75999880712583e-06,
+      "loss": 0.9924,
+      "step": 65010
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.7541604084787343e-06,
+      "loss": 0.8775,
+      "step": 65015
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.748328105311626e-06,
+      "loss": 0.7592,
+      "step": 65020
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.7425018979900574e-06,
+      "loss": 1.1202,
+      "step": 65025
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.736681786879247e-06,
+      "loss": 1.1027,
+      "step": 65030
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.7308677723440033e-06,
+      "loss": 1.0419,
+      "step": 65035
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.725059854748735e-06,
+      "loss": 1.0767,
+      "step": 65040
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.7192580344574948e-06,
+      "loss": 0.9814,
+      "step": 65045
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.7134623118339807e-06,
+      "loss": 1.1068,
+      "step": 65050
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.7076726872414358e-06,
+      "loss": 0.924,
+      "step": 65055
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.70188916104277e-06,
+      "loss": 0.9105,
+      "step": 65060
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.6961117336005036e-06,
+      "loss": 0.968,
+      "step": 65065
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.690340405276781e-06,
+      "loss": 0.9132,
+      "step": 65070
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.6845751764333347e-06,
+      "loss": 0.9445,
+      "step": 65075
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.678816047431554e-06,
+      "loss": 0.9016,
+      "step": 65080
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.6730630186324378e-06,
+      "loss": 1.0218,
+      "step": 65085
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.6673160903965656e-06,
+      "loss": 0.9595,
+      "step": 65090
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.6615752630841707e-06,
+      "loss": 1.0249,
+      "step": 65095
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.486328125,
+      "learning_rate": 2.6558405370551097e-06,
+      "loss": 0.867,
+      "step": 65100
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.650111912668818e-06,
+      "loss": 1.1201,
+      "step": 65105
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.6443893902843853e-06,
+      "loss": 0.765,
+      "step": 65110
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.6386729702605027e-06,
+      "loss": 0.8427,
+      "step": 65115
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.632962652955495e-06,
+      "loss": 0.9295,
+      "step": 65120
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.6272584387272757e-06,
+      "loss": 0.8598,
+      "step": 65125
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.621560327933381e-06,
+      "loss": 0.9254,
+      "step": 65130
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.6158683209310142e-06,
+      "loss": 0.9867,
+      "step": 65135
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.610182418076912e-06,
+      "loss": 1.1187,
+      "step": 65140
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.6045026197275e-06,
+      "loss": 0.8453,
+      "step": 65145
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.625,
+      "learning_rate": 2.5988289262387944e-06,
+      "loss": 0.8491,
+      "step": 65150
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.59316133796641e-06,
+      "loss": 1.0447,
+      "step": 65155
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.5874998552656293e-06,
+      "loss": 0.9154,
+      "step": 65160
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.581844478491291e-06,
+      "loss": 1.0599,
+      "step": 65165
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.576195207997889e-06,
+      "loss": 1.068,
+      "step": 65170
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.5705520441395183e-06,
+      "loss": 0.9493,
+      "step": 65175
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.5649149872699287e-06,
+      "loss": 1.04,
+      "step": 65180
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.5592840377424153e-06,
+      "loss": 1.0012,
+      "step": 65185
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.5536591959099634e-06,
+      "loss": 0.9686,
+      "step": 65190
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.5480404621251453e-06,
+      "loss": 0.8808,
+      "step": 65195
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.5424278367401244e-06,
+      "loss": 0.8381,
+      "step": 65200
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.5368213201067192e-06,
+      "loss": 0.9821,
+      "step": 65205
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.531220912576371e-06,
+      "loss": 0.9479,
+      "step": 65210
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.5256266145000986e-06,
+      "loss": 0.9484,
+      "step": 65215
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.5200384262285658e-06,
+      "loss": 0.9905,
+      "step": 65220
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.5144563481120263e-06,
+      "loss": 0.9472,
+      "step": 65225
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.508880380500411e-06,
+      "loss": 1.0241,
+      "step": 65230
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.5033105237431964e-06,
+      "loss": 0.9847,
+      "step": 65235
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.4977467781895138e-06,
+      "loss": 1.1087,
+      "step": 65240
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.4921891441881286e-06,
+      "loss": 0.9616,
+      "step": 65245
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.486637622087362e-06,
+      "loss": 1.0369,
+      "step": 65250
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.4810922122352032e-06,
+      "loss": 0.9011,
+      "step": 65255
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.475552914979273e-06,
+      "loss": 1.0724,
+      "step": 65260
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.4700197306667393e-06,
+      "loss": 1.0771,
+      "step": 65265
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.4644926596444464e-06,
+      "loss": 0.9856,
+      "step": 65270
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.625,
+      "learning_rate": 2.45897170225885e-06,
+      "loss": 1.0245,
+      "step": 65275
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.4534568588559847e-06,
+      "loss": 0.9473,
+      "step": 65280
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.447948129781541e-06,
+      "loss": 1.0451,
+      "step": 65285
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5,
+      "learning_rate": 2.4424455153808202e-06,
+      "loss": 0.8536,
+      "step": 65290
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.4369490159987018e-06,
+      "loss": 0.8815,
+      "step": 65295
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.431458631979744e-06,
+      "loss": 1.0601,
+      "step": 65300
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.425974363668093e-06,
+      "loss": 0.8539,
+      "step": 65305
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.420496211407486e-06,
+      "loss": 0.9943,
+      "step": 65310
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.4150241755413026e-06,
+      "loss": 0.9937,
+      "step": 65315
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.4095582564125584e-06,
+      "loss": 0.9439,
+      "step": 65320
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.4040984543638567e-06,
+      "loss": 0.8602,
+      "step": 65325
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.3986447697374013e-06,
+      "loss": 0.8506,
+      "step": 65330
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.3931972028750414e-06,
+      "loss": 0.8138,
+      "step": 65335
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.3877557541182704e-06,
+      "loss": 0.8287,
+      "step": 65340
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.3823204238081153e-06,
+      "loss": 0.9755,
+      "step": 65345
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.376891212285304e-06,
+      "loss": 0.9008,
+      "step": 65350
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.371468119890141e-06,
+      "loss": 1.1037,
+      "step": 65355
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.3660511469625336e-06,
+      "loss": 0.9623,
+      "step": 65360
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.3606402938420425e-06,
+      "loss": 0.8876,
+      "step": 65365
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.35523556086783e-06,
+      "loss": 1.1445,
+      "step": 65370
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4921875,
+      "learning_rate": 2.3498369483786476e-06,
+      "loss": 0.9246,
+      "step": 65375
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.3444444567128907e-06,
+      "loss": 1.0409,
+      "step": 65380
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.339058086208601e-06,
+      "loss": 1.1015,
+      "step": 65385
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.333677837203352e-06,
+      "loss": 0.8732,
+      "step": 65390
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.3283037100344187e-06,
+      "loss": 0.7876,
+      "step": 65395
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.3229357050386536e-06,
+      "loss": 0.8576,
+      "step": 65400
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.31757382255251e-06,
+      "loss": 0.9901,
+      "step": 65405
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.312218062912097e-06,
+      "loss": 0.8474,
+      "step": 65410
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.3068684264531236e-06,
+      "loss": 0.8593,
+      "step": 65415
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.3015249135108883e-06,
+      "loss": 0.9145,
+      "step": 65420
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.296187524420346e-06,
+      "loss": 1.052,
+      "step": 65425
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.290856259516039e-06,
+      "loss": 0.948,
+      "step": 65430
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.625,
+      "learning_rate": 2.285531119132145e-06,
+      "loss": 1.0625,
+      "step": 65435
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.2802121036024528e-06,
+      "loss": 0.9666,
+      "step": 65440
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.274899213260351e-06,
+      "loss": 1.0447,
+      "step": 65445
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.269592448438873e-06,
+      "loss": 1.0886,
+      "step": 65450
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.2642918094706423e-06,
+      "loss": 0.8001,
+      "step": 65455
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.2589972966879037e-06,
+      "loss": 0.9675,
+      "step": 65460
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.253708910422547e-06,
+      "loss": 0.8762,
+      "step": 65465
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.2484266510060416e-06,
+      "loss": 0.9948,
+      "step": 65470
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.2431505187694767e-06,
+      "loss": 0.8915,
+      "step": 65475
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.2378805140435778e-06,
+      "loss": 0.9509,
+      "step": 65480
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.232616637158669e-06,
+      "loss": 0.9404,
+      "step": 65485
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.227358888444686e-06,
+      "loss": 0.9962,
+      "step": 65490
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.2221072682312103e-06,
+      "loss": 0.9881,
+      "step": 65495
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.216861776847412e-06,
+      "loss": 1.0974,
+      "step": 65500
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.211622414622072e-06,
+      "loss": 1.0143,
+      "step": 65505
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.206389181883628e-06,
+      "loss": 0.9016,
+      "step": 65510
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.201162078960073e-06,
+      "loss": 1.0309,
+      "step": 65515
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.195941106179067e-06,
+      "loss": 0.9086,
+      "step": 65520
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.1907262638678705e-06,
+      "loss": 0.9846,
+      "step": 65525
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.1855175523533443e-06,
+      "loss": 0.8841,
+      "step": 65530
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.180314971961972e-06,
+      "loss": 0.8873,
+      "step": 65535
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.175118523019859e-06,
+      "loss": 0.9228,
+      "step": 65540
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.169928205852745e-06,
+      "loss": 0.8389,
+      "step": 65545
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.164744020785936e-06,
+      "loss": 0.9474,
+      "step": 65550
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.159565968144406e-06,
+      "loss": 0.7416,
+      "step": 65555
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.154394048252717e-06,
+      "loss": 0.8988,
+      "step": 65560
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.1492282614350323e-06,
+      "loss": 0.9856,
+      "step": 65565
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.1440686080151708e-06,
+      "loss": 0.9949,
+      "step": 65570
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.1389150883165397e-06,
+      "loss": 0.9206,
+      "step": 65575
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.1337677026621707e-06,
+      "loss": 0.8943,
+      "step": 65580
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.1286264513746822e-06,
+      "loss": 0.9125,
+      "step": 65585
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.486328125,
+      "learning_rate": 2.123491334776373e-06,
+      "loss": 0.9924,
+      "step": 65590
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4921875,
+      "learning_rate": 2.1183623531890963e-06,
+      "loss": 0.9117,
+      "step": 65595
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.625,
+      "learning_rate": 2.113239506934328e-06,
+      "loss": 0.9869,
+      "step": 65600
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.1081227963332116e-06,
+      "loss": 0.9346,
+      "step": 65605
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.103012221706435e-06,
+      "loss": 0.9464,
+      "step": 65610
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.0979077833743532e-06,
+      "loss": 0.9183,
+      "step": 65615
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.09280948165691e-06,
+      "loss": 0.9272,
+      "step": 65620
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.087717316873683e-06,
+      "loss": 0.9441,
+      "step": 65625
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.082631289343828e-06,
+      "loss": 0.8887,
+      "step": 65630
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.0775513993861905e-06,
+      "loss": 0.941,
+      "step": 65635
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.0724776473191374e-06,
+      "loss": 0.9125,
+      "step": 65640
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.067410033460726e-06,
+      "loss": 0.9141,
+      "step": 65645
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.0623485581285795e-06,
+      "loss": 1.0082,
+      "step": 65650
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.0572932216399665e-06,
+      "loss": 1.0309,
+      "step": 65655
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.052244024311767e-06,
+      "loss": 0.931,
+      "step": 65660
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.0472009664604606e-06,
+      "loss": 1.0014,
+      "step": 65665
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.042164048402162e-06,
+      "loss": 0.996,
+      "step": 65670
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.0371332704525847e-06,
+      "loss": 0.9044,
+      "step": 65675
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.0321086329270547e-06,
+      "loss": 0.854,
+      "step": 65680
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.0270901361405412e-06,
+      "loss": 1.0305,
+      "step": 65685
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.0220777804076053e-06,
+      "loss": 0.9537,
+      "step": 65690
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.0170715660424166e-06,
+      "loss": 0.7839,
+      "step": 65695
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.0120714933587803e-06,
+      "loss": 1.1057,
+      "step": 65700
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.00707756267009e-06,
+      "loss": 0.9495,
+      "step": 65705
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.002089774289384e-06,
+      "loss": 0.9803,
+      "step": 65710
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.9971081285293013e-06,
+      "loss": 0.9487,
+      "step": 65715
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.9921326257021035e-06,
+      "loss": 1.0803,
+      "step": 65720
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9871632661196406e-06,
+      "loss": 0.9914,
+      "step": 65725
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.98220005009343e-06,
+      "loss": 0.8314,
+      "step": 65730
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9772429779345457e-06,
+      "loss": 0.8738,
+      "step": 65735
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.9722920499537168e-06,
+      "loss": 0.9574,
+      "step": 65740
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.96734726646125e-06,
+      "loss": 1.0062,
+      "step": 65745
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.9624086277671206e-06,
+      "loss": 0.8644,
+      "step": 65750
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.957476134180869e-06,
+      "loss": 1.0087,
+      "step": 65755
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.9525497860116703e-06,
+      "loss": 0.7952,
+      "step": 65760
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.9476295835683223e-06,
+      "loss": 0.9089,
+      "step": 65765
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.9427155271592224e-06,
+      "loss": 0.8125,
+      "step": 65770
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.937807617092391e-06,
+      "loss": 1.1625,
+      "step": 65775
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9329058536754598e-06,
+      "loss": 0.9517,
+      "step": 65780
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9280102372156827e-06,
+      "loss": 0.8917,
+      "step": 65785
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.923120768019926e-06,
+      "loss": 0.9509,
+      "step": 65790
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.918237446394644e-06,
+      "loss": 1.0978,
+      "step": 65795
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.913360272645959e-06,
+      "loss": 0.8706,
+      "step": 65800
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.908489247079548e-06,
+      "loss": 0.9537,
+      "step": 65805
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.9036243700007562e-06,
+      "loss": 0.8734,
+      "step": 65810
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.8987656417145173e-06,
+      "loss": 1.0161,
+      "step": 65815
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.8939130625253653e-06,
+      "loss": 1.0725,
+      "step": 65820
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.8890666327374795e-06,
+      "loss": 0.8345,
+      "step": 65825
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.884226352654639e-06,
+      "loss": 0.912,
+      "step": 65830
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.879392222580234e-06,
+      "loss": 1.0161,
+      "step": 65835
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.8745642428172783e-06,
+      "loss": 0.7901,
+      "step": 65840
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.8697424136683961e-06,
+      "loss": 1.0204,
+      "step": 65845
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.8649267354358012e-06,
+      "loss": 0.9864,
+      "step": 65850
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.8601172084213747e-06,
+      "loss": 0.8536,
+      "step": 65855
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.855313832926575e-06,
+      "loss": 0.9017,
+      "step": 65860
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.850516609252484e-06,
+      "loss": 1.0248,
+      "step": 65865
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.8457255376997828e-06,
+      "loss": 0.915,
+      "step": 65870
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.8409406185688093e-06,
+      "loss": 0.8835,
+      "step": 65875
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.8361618521594571e-06,
+      "loss": 1.0339,
+      "step": 65880
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.831389238771286e-06,
+      "loss": 1.0657,
+      "step": 65885
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.8266227787034463e-06,
+      "loss": 0.9807,
+      "step": 65890
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.8218624722547095e-06,
+      "loss": 0.9648,
+      "step": 65895
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.8171083197234262e-06,
+      "loss": 0.9703,
+      "step": 65900
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.8123603214076246e-06,
+      "loss": 0.8993,
+      "step": 65905
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.807618477604911e-06,
+      "loss": 1.0617,
+      "step": 65910
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.8028827886124922e-06,
+      "loss": 0.9137,
+      "step": 65915
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.7981532547272195e-06,
+      "loss": 0.8972,
+      "step": 65920
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.793429876245545e-06,
+      "loss": 0.8466,
+      "step": 65925
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.7887126534635311e-06,
+      "loss": 1.1436,
+      "step": 65930
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.7840015866768645e-06,
+      "loss": 0.8009,
+      "step": 65935
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.7792966761808416e-06,
+      "loss": 0.9147,
+      "step": 65940
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.7745979222703602e-06,
+      "loss": 0.9281,
+      "step": 65945
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.769905325239951e-06,
+      "loss": 1.0265,
+      "step": 65950
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.765218885383757e-06,
+      "loss": 0.8913,
+      "step": 65955
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.7605386029955205e-06,
+      "loss": 0.9417,
+      "step": 65960
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.7558644783685962e-06,
+      "loss": 0.9741,
+      "step": 65965
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.751196511795994e-06,
+      "loss": 0.8871,
+      "step": 65970
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.7465347035702795e-06,
+      "loss": 0.8722,
+      "step": 65975
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.741879053983675e-06,
+      "loss": 0.9352,
+      "step": 65980
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.7372295633280023e-06,
+      "loss": 0.8443,
+      "step": 65985
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.7325862318946951e-06,
+      "loss": 1.0828,
+      "step": 65990
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.727949059974787e-06,
+      "loss": 0.8578,
+      "step": 65995
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.7233180478589683e-06,
+      "loss": 0.886,
+      "step": 66000
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.7186931958374953e-06,
+      "loss": 1.0852,
+      "step": 66005
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.7140745042002694e-06,
+      "loss": 0.9812,
+      "step": 66010
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.7094619732367812e-06,
+      "loss": 0.923,
+      "step": 66015
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.7048556032361773e-06,
+      "loss": 0.9416,
+      "step": 66020
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.7002553944871602e-06,
+      "loss": 0.9781,
+      "step": 66025
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.6956613472780881e-06,
+      "loss": 0.8714,
+      "step": 66030
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.6910734618969304e-06,
+      "loss": 0.8347,
+      "step": 66035
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.6864917386312462e-06,
+      "loss": 0.8215,
+      "step": 66040
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.681916177768228e-06,
+      "loss": 0.9532,
+      "step": 66045
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.6773467795946796e-06,
+      "loss": 0.9491,
+      "step": 66050
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.6727835443970275e-06,
+      "loss": 1.0347,
+      "step": 66055
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.6682264724612762e-06,
+      "loss": 0.9638,
+      "step": 66060
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.6636755640730862e-06,
+      "loss": 0.7433,
+      "step": 66065
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.6591308195176958e-06,
+      "loss": 0.9629,
+      "step": 66070
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.6545922390799994e-06,
+      "loss": 0.9322,
+      "step": 66075
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5,
+      "learning_rate": 1.6500598230444586e-06,
+      "loss": 0.7957,
+      "step": 66080
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.6455335716951793e-06,
+      "loss": 0.9208,
+      "step": 66085
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.6410134853158786e-06,
+      "loss": 0.9595,
+      "step": 66090
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.636499564189875e-06,
+      "loss": 0.8989,
+      "step": 66095
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.6319918086000974e-06,
+      "loss": 0.8626,
+      "step": 66100
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.6274902188291197e-06,
+      "loss": 0.8508,
+      "step": 66105
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.6229947951590718e-06,
+      "loss": 0.9814,
+      "step": 66110
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.6185055378717728e-06,
+      "loss": 0.954,
+      "step": 66115
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.6140224472485755e-06,
+      "loss": 1.0494,
+      "step": 66120
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.6095455235705104e-06,
+      "loss": 0.913,
+      "step": 66125
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.6050747671181864e-06,
+      "loss": 0.9894,
+      "step": 66130
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.6006101781718463e-06,
+      "loss": 0.7572,
+      "step": 66135
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.5961517570113215e-06,
+      "loss": 1.012,
+      "step": 66140
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.5916995039160775e-06,
+      "loss": 0.8833,
+      "step": 66145
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.58725341916518e-06,
+      "loss": 1.0355,
+      "step": 66150
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.5828135030373282e-06,
+      "loss": 0.8036,
+      "step": 66155
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.5783797558108326e-06,
+      "loss": 0.9307,
+      "step": 66160
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.5739521777635601e-06,
+      "loss": 0.9648,
+      "step": 66165
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.5695307691730665e-06,
+      "loss": 0.9466,
+      "step": 66170
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.5651155303164967e-06,
+      "loss": 1.1032,
+      "step": 66175
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.5607064614705846e-06,
+      "loss": 0.763,
+      "step": 66180
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.5563035629117095e-06,
+      "loss": 0.9432,
+      "step": 66185
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.5519068349158505e-06,
+      "loss": 0.8592,
+      "step": 66190
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.547516277758576e-06,
+      "loss": 1.0989,
+      "step": 66195
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.5431318917151216e-06,
+      "loss": 0.9865,
+      "step": 66200
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.5387536770603007e-06,
+      "loss": 0.7946,
+      "step": 66205
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.5343816340685268e-06,
+      "loss": 0.9329,
+      "step": 66210
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.530015763013859e-06,
+      "loss": 1.0905,
+      "step": 66215
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.5256560641699446e-06,
+      "loss": 1.2853,
+      "step": 66220
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.5213025378100654e-06,
+      "loss": 0.9044,
+      "step": 66225
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.5169551842071028e-06,
+      "loss": 0.838,
+      "step": 66230
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.5126140036335502e-06,
+      "loss": 0.9923,
+      "step": 66235
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.5082789963615228e-06,
+      "loss": 1.1155,
+      "step": 66240
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.5039501626627372e-06,
+      "loss": 0.8375,
+      "step": 66245
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.4996275028085205e-06,
+      "loss": 0.9562,
+      "step": 66250
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.4953110170698558e-06,
+      "loss": 0.9143,
+      "step": 66255
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.4910007057172714e-06,
+      "loss": 0.8973,
+      "step": 66260
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.4866965690209622e-06,
+      "loss": 0.8352,
+      "step": 66265
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.4823986072507123e-06,
+      "loss": 1.0024,
+      "step": 66270
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.4781068206759063e-06,
+      "loss": 0.8752,
+      "step": 66275
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.4738212095655846e-06,
+      "loss": 1.0038,
+      "step": 66280
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.4695417741883543e-06,
+      "loss": 0.8395,
+      "step": 66285
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.4652685148124678e-06,
+      "loss": 0.925,
+      "step": 66290
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.4610014317057664e-06,
+      "loss": 0.999,
+      "step": 66295
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.4567405251357246e-06,
+      "loss": 0.9159,
+      "step": 66300
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.4524857953694183e-06,
+      "loss": 0.8985,
+      "step": 66305
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.4482372426735337e-06,
+      "loss": 0.8518,
+      "step": 66310
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.4439948673143799e-06,
+      "loss": 0.967,
+      "step": 66315
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.439758669557889e-06,
+      "loss": 0.9035,
+      "step": 66320
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.4355286496695598e-06,
+      "loss": 0.9379,
+      "step": 66325
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.4313048079145463e-06,
+      "loss": 1.0857,
+      "step": 66330
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4270871445576262e-06,
+      "loss": 0.9745,
+      "step": 66335
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.4228756598631322e-06,
+      "loss": 0.8944,
+      "step": 66340
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.418670354095064e-06,
+      "loss": 0.9383,
+      "step": 66345
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.4144712275170224e-06,
+      "loss": 0.984,
+      "step": 66350
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.4102782803921966e-06,
+      "loss": 0.8358,
+      "step": 66355
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.4060915129834096e-06,
+      "loss": 0.9024,
+      "step": 66360
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.401910925553107e-06,
+      "loss": 0.9139,
+      "step": 66365
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.397736518363324e-06,
+      "loss": 0.9026,
+      "step": 66370
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.3935682916756953e-06,
+      "loss": 0.9661,
+      "step": 66375
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.3894062457515233e-06,
+      "loss": 0.8531,
+      "step": 66380
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.3852503808516771e-06,
+      "loss": 0.9482,
+      "step": 66385
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.625,
+      "learning_rate": 1.381100697236637e-06,
+      "loss": 0.934,
+      "step": 66390
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.3769571951665393e-06,
+      "loss": 1.1681,
+      "step": 66395
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.3728198749010657e-06,
+      "loss": 0.8876,
+      "step": 66400
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.368688736699586e-06,
+      "loss": 0.8038,
+      "step": 66405
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.3645637808210266e-06,
+      "loss": 0.9515,
+      "step": 66410
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.360445007523936e-06,
+      "loss": 0.914,
+      "step": 66415
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.3563324170665082e-06,
+      "loss": 0.891,
+      "step": 66420
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.3522260097064921e-06,
+      "loss": 1.0436,
+      "step": 66425
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.3481257857013152e-06,
+      "loss": 1.1661,
+      "step": 66430
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.3440317453079609e-06,
+      "loss": 0.8857,
+      "step": 66435
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.3399438887830572e-06,
+      "loss": 0.807,
+      "step": 66440
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.3358622163828438e-06,
+      "loss": 0.9729,
+      "step": 66445
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.331786728363149e-06,
+      "loss": 0.8429,
+      "step": 66450
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.3277174249794355e-06,
+      "loss": 0.7673,
+      "step": 66455
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.3236543064867769e-06,
+      "loss": 0.9777,
+      "step": 66460
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.319597373139836e-06,
+      "loss": 0.9414,
+      "step": 66465
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.3155466251929316e-06,
+      "loss": 0.8902,
+      "step": 66470
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.3115020628999498e-06,
+      "loss": 0.8715,
+      "step": 66475
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.3074636865144206e-06,
+      "loss": 0.9018,
+      "step": 66480
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.3034314962894756e-06,
+      "loss": 0.8096,
+      "step": 66485
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.2994054924778342e-06,
+      "loss": 1.0186,
+      "step": 66490
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.2953856753318838e-06,
+      "loss": 0.9389,
+      "step": 66495
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.291372045103556e-06,
+      "loss": 1.0381,
+      "step": 66500
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.2873646020444608e-06,
+      "loss": 0.8876,
+      "step": 66505
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2833633464057748e-06,
+      "loss": 0.9844,
+      "step": 66510
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.2793682784382977e-06,
+      "loss": 0.9653,
+      "step": 66515
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.2753793983924622e-06,
+      "loss": 0.9561,
+      "step": 66520
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.2713967065182797e-06,
+      "loss": 0.9512,
+      "step": 66525
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.2674202030653836e-06,
+      "loss": 1.0748,
+      "step": 66530
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.2634498882830415e-06,
+      "loss": 0.9455,
+      "step": 66535
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.2594857624201207e-06,
+      "loss": 1.1208,
+      "step": 66540
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.255527825725078e-06,
+      "loss": 0.8845,
+      "step": 66545
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.251576078446015e-06,
+      "loss": 1.1462,
+      "step": 66550
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.2476305208306226e-06,
+      "loss": 0.9087,
+      "step": 66555
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.243691153126225e-06,
+      "loss": 1.0049,
+      "step": 66560
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.2397579755797361e-06,
+      "loss": 0.9512,
+      "step": 66565
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.2358309884377029e-06,
+      "loss": 0.9347,
+      "step": 66570
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.2319101919462616e-06,
+      "loss": 0.9896,
+      "step": 66575
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2279955863511826e-06,
+      "loss": 0.9109,
+      "step": 66580
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.2240871718978253e-06,
+      "loss": 0.9015,
+      "step": 66585
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.2201849488311822e-06,
+      "loss": 0.9944,
+      "step": 66590
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.216288917395847e-06,
+      "loss": 1.1017,
+      "step": 66595
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.625,
+      "learning_rate": 1.2123990778360238e-06,
+      "loss": 0.9436,
+      "step": 66600
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.2085154303955515e-06,
+      "loss": 0.9388,
+      "step": 66605
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.2046379753178238e-06,
+      "loss": 0.912,
+      "step": 66610
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.2007667128459133e-06,
+      "loss": 0.9379,
+      "step": 66615
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.196901643222481e-06,
+      "loss": 0.9229,
+      "step": 66620
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.193042766689767e-06,
+      "loss": 1.0321,
+      "step": 66625
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.1891900834896775e-06,
+      "loss": 0.9575,
+      "step": 66630
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1853435938636858e-06,
+      "loss": 1.1775,
+      "step": 66635
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.1815032980528885e-06,
+      "loss": 0.9579,
+      "step": 66640
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.1776691962980146e-06,
+      "loss": 0.9174,
+      "step": 66645
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.1738412888393835e-06,
+      "loss": 0.9915,
+      "step": 66650
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.1700195759169364e-06,
+      "loss": 0.9272,
+      "step": 66655
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.1662040577702148e-06,
+      "loss": 0.9615,
+      "step": 66660
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.1623947346383946e-06,
+      "loss": 0.9528,
+      "step": 66665
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.1585916067602286e-06,
+      "loss": 0.8361,
+      "step": 66670
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.1547946743741155e-06,
+      "loss": 1.0168,
+      "step": 66675
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.1510039377180536e-06,
+      "loss": 0.9082,
+      "step": 66680
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.1472193970296419e-06,
+      "loss": 0.9601,
+      "step": 66685
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.1434410525461014e-06,
+      "loss": 0.9541,
+      "step": 66690
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.139668904504254e-06,
+      "loss": 1.0303,
+      "step": 66695
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1359029531405662e-06,
+      "loss": 0.8633,
+      "step": 66700
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.1321431986910712e-06,
+      "loss": 0.9686,
+      "step": 66705
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.1283896413914362e-06,
+      "loss": 0.9233,
+      "step": 66710
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.1246422814769509e-06,
+      "loss": 1.0196,
+      "step": 66715
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5,
+      "learning_rate": 1.1209011191824935e-06,
+      "loss": 0.9566,
+      "step": 66720
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.1171661547425772e-06,
+      "loss": 0.9021,
+      "step": 66725
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5,
+      "learning_rate": 1.1134373883913029e-06,
+      "loss": 0.8414,
+      "step": 66730
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.1097148203623841e-06,
+      "loss": 1.0286,
+      "step": 66735
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.1059984508891785e-06,
+      "loss": 0.9377,
+      "step": 66740
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.1022882802046218e-06,
+      "loss": 0.9322,
+      "step": 66745
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.0985843085412617e-06,
+      "loss": 0.9119,
+      "step": 66750
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.0948865361312676e-06,
+      "loss": 1.0042,
+      "step": 66755
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.0911949632064434e-06,
+      "loss": 0.9995,
+      "step": 66760
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.0875095899981702e-06,
+      "loss": 1.1144,
+      "step": 66765
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.0838304167374302e-06,
+      "loss": 0.8735,
+      "step": 66770
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.0801574436548722e-06,
+      "loss": 0.8851,
+      "step": 66775
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.076490670980701e-06,
+      "loss": 0.9889,
+      "step": 66780
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.0728300989447548e-06,
+      "loss": 1.0086,
+      "step": 66785
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.0691757277764946e-06,
+      "loss": 0.995,
+      "step": 66790
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.0655275577049706e-06,
+      "loss": 0.9718,
+      "step": 66795
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.0618855889588552e-06,
+      "loss": 0.9534,
+      "step": 66800
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.0582498217664215e-06,
+      "loss": 1.0658,
+      "step": 66805
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.0546202563555874e-06,
+      "loss": 0.9804,
+      "step": 66810
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0509968929538484e-06,
+      "loss": 1.0017,
+      "step": 66815
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.0473797317883005e-06,
+      "loss": 0.9555,
+      "step": 66820
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.043768773085707e-06,
+      "loss": 1.0213,
+      "step": 66825
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.0401640170723758e-06,
+      "loss": 0.9763,
+      "step": 66830
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0365654639742706e-06,
+      "loss": 1.0952,
+      "step": 66835
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.0329731140169663e-06,
+      "loss": 1.2155,
+      "step": 66840
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.0293869674256051e-06,
+      "loss": 0.8288,
+      "step": 66845
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.025807024424985e-06,
+      "loss": 0.9911,
+      "step": 66850
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.0222332852395156e-06,
+      "loss": 0.9902,
+      "step": 66855
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.018665750093184e-06,
+      "loss": 0.83,
+      "step": 66860
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.0151044192096115e-06,
+      "loss": 0.8954,
+      "step": 66865
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.0115492928120306e-06,
+      "loss": 1.0404,
+      "step": 66870
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.0080003711232743e-06,
+      "loss": 0.8523,
+      "step": 66875
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0044576543658092e-06,
+      "loss": 0.97,
+      "step": 66880
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.0009211427616794e-06,
+      "loss": 0.8799,
+      "step": 66885
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.973908365325525e-07,
+      "loss": 1.0422,
+      "step": 66890
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.9386673589974e-07,
+      "loss": 1.0017,
+      "step": 66895
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.903488410840988e-07,
+      "loss": 0.9493,
+      "step": 66900
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.868371523061748e-07,
+      "loss": 0.9286,
+      "step": 66905
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.833316697860474e-07,
+      "loss": 0.9287,
+      "step": 66910
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.486328125,
+      "learning_rate": 9.798323937434739e-07,
+      "loss": 0.9732,
+      "step": 66915
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.763393243977792e-07,
+      "loss": 0.9871,
+      "step": 66920
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.728524619678991e-07,
+      "loss": 0.8952,
+      "step": 66925
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.693718066724256e-07,
+      "loss": 0.9673,
+      "step": 66930
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.658973587295062e-07,
+      "loss": 0.8788,
+      "step": 66935
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.624291183569334e-07,
+      "loss": 0.8784,
+      "step": 66940
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.48828125,
+      "learning_rate": 9.589670857721e-07,
+      "loss": 1.0006,
+      "step": 66945
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.555112611920104e-07,
+      "loss": 0.9158,
+      "step": 66950
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.520616448332686e-07,
+      "loss": 0.9912,
+      "step": 66955
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.486182369121132e-07,
+      "loss": 0.9214,
+      "step": 66960
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.451810376443826e-07,
+      "loss": 1.0211,
+      "step": 66965
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.48828125,
+      "learning_rate": 9.417500472455043e-07,
+      "loss": 0.9691,
+      "step": 66970
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.383252659305619e-07,
+      "loss": 1.106,
+      "step": 66975
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.34906693914217e-07,
+      "loss": 0.9614,
+      "step": 66980
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.314943314107316e-07,
+      "loss": 0.981,
+      "step": 66985
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.280881786340124e-07,
+      "loss": 0.894,
+      "step": 66990
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.246882357975772e-07,
+      "loss": 0.8769,
+      "step": 66995
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.212945031145115e-07,
+      "loss": 1.0055,
+      "step": 67000
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.498046875,
+      "learning_rate": 9.179069807975449e-07,
+      "loss": 0.8045,
+      "step": 67005
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.145256690590076e-07,
+      "loss": 1.0224,
+      "step": 67010
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.111505681108634e-07,
+      "loss": 0.907,
+      "step": 67015
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.077816781646431e-07,
+      "loss": 1.1019,
+      "step": 67020
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.044189994315333e-07,
+      "loss": 0.8189,
+      "step": 67025
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.010625321222987e-07,
+      "loss": 1.0068,
+      "step": 67030
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.625,
+      "learning_rate": 8.977122764473267e-07,
+      "loss": 0.916,
+      "step": 67035
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.943682326166159e-07,
+      "loss": 0.9092,
+      "step": 67040
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.625,
+      "learning_rate": 8.910304008397763e-07,
+      "loss": 0.9385,
+      "step": 67045
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.876987813260407e-07,
+      "loss": 0.968,
+      "step": 67050
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.843733742842086e-07,
+      "loss": 1.0223,
+      "step": 67055
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.810541799227356e-07,
+      "loss": 0.9759,
+      "step": 67060
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.777411984496775e-07,
+      "loss": 0.9314,
+      "step": 67065
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.453125,
+      "learning_rate": 8.744344300726904e-07,
+      "loss": 0.9452,
+      "step": 67070
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.711338749990528e-07,
+      "loss": 0.8704,
+      "step": 67075
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.678395334356437e-07,
+      "loss": 0.8862,
+      "step": 67080
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.645514055889425e-07,
+      "loss": 1.1477,
+      "step": 67085
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.612694916650843e-07,
+      "loss": 0.9368,
+      "step": 67090
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.579937918697489e-07,
+      "loss": 0.9919,
+      "step": 67095
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.547243064082721e-07,
+      "loss": 0.9392,
+      "step": 67100
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.514610354856123e-07,
+      "loss": 0.8963,
+      "step": 67105
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.482039793062835e-07,
+      "loss": 0.9462,
+      "step": 67110
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.449531380744558e-07,
+      "loss": 0.9711,
+      "step": 67115
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.417085119938883e-07,
+      "loss": 0.9606,
+      "step": 67120
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.38470101267963e-07,
+      "loss": 1.0648,
+      "step": 67125
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.352379060996729e-07,
+      "loss": 0.9065,
+      "step": 67130
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.320119266916115e-07,
+      "loss": 0.9925,
+      "step": 67135
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.287921632459838e-07,
+      "loss": 0.8692,
+      "step": 67140
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.255786159646062e-07,
+      "loss": 1.0614,
+      "step": 67145
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.223712850489063e-07,
+      "loss": 0.8723,
+      "step": 67150
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.191701706999344e-07,
+      "loss": 0.9797,
+      "step": 67155
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.159752731183412e-07,
+      "loss": 0.8765,
+      "step": 67160
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.127865925043665e-07,
+      "loss": 0.8819,
+      "step": 67165
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.096041290578948e-07,
+      "loss": 1.06,
+      "step": 67170
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.48046875,
+      "learning_rate": 8.064278829784111e-07,
+      "loss": 1.0899,
+      "step": 67175
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.032578544650004e-07,
+      "loss": 1.0346,
+      "step": 67180
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.000940437163595e-07,
+      "loss": 0.9365,
+      "step": 67185
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.969364509308075e-07,
+      "loss": 0.9604,
+      "step": 67190
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.937850763062527e-07,
+      "loss": 1.0253,
+      "step": 67195
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.90639920040237e-07,
+      "loss": 1.0417,
+      "step": 67200
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.875009823299029e-07,
+      "loss": 0.7986,
+      "step": 67205
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.84368263372004e-07,
+      "loss": 0.8278,
+      "step": 67210
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.812417633628943e-07,
+      "loss": 0.9306,
+      "step": 67215
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.781214824985617e-07,
+      "loss": 0.9969,
+      "step": 67220
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.478515625,
+      "learning_rate": 7.750074209745717e-07,
+      "loss": 0.8955,
+      "step": 67225
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.718995789861238e-07,
+      "loss": 0.9978,
+      "step": 67230
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.68797956728029e-07,
+      "loss": 0.9923,
+      "step": 67235
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.65702554394676e-07,
+      "loss": 1.0853,
+      "step": 67240
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.62613372180121e-07,
+      "loss": 0.9365,
+      "step": 67245
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.595304102779754e-07,
+      "loss": 0.8599,
+      "step": 67250
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.564536688814849e-07,
+      "loss": 1.0178,
+      "step": 67255
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.533831481835174e-07,
+      "loss": 0.9237,
+      "step": 67260
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.439453125,
+      "learning_rate": 7.503188483765189e-07,
+      "loss": 1.0412,
+      "step": 67265
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.490234375,
+      "learning_rate": 7.472607696525801e-07,
+      "loss": 0.848,
+      "step": 67270
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.442089122033702e-07,
+      "loss": 0.8946,
+      "step": 67275
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.411632762201804e-07,
+      "loss": 0.8829,
+      "step": 67280
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.381238618939357e-07,
+      "loss": 1.0094,
+      "step": 67285
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.350906694151283e-07,
+      "loss": 1.258,
+      "step": 67290
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.32063698973895e-07,
+      "loss": 0.8682,
+      "step": 67295
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.290429507599616e-07,
+      "loss": 0.9659,
+      "step": 67300
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.26028424962677e-07,
+      "loss": 0.9875,
+      "step": 67305
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.230201217710009e-07,
+      "loss": 0.8878,
+      "step": 67310
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.200180413734936e-07,
+      "loss": 0.9329,
+      "step": 67315
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.170221839583158e-07,
+      "loss": 0.8614,
+      "step": 67320
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.140325497132616e-07,
+      "loss": 0.9333,
+      "step": 67325
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.110491388257367e-07,
+      "loss": 0.7535,
+      "step": 67330
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.080719514827139e-07,
+      "loss": 1.0076,
+      "step": 67335
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.051009878708326e-07,
+      "loss": 0.916,
+      "step": 67340
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.021362481763106e-07,
+      "loss": 0.8847,
+      "step": 67345
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.991777325849769e-07,
+      "loss": 0.981,
+      "step": 67350
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.962254412822833e-07,
+      "loss": 0.9348,
+      "step": 67355
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.932793744532707e-07,
+      "loss": 0.9446,
+      "step": 67360
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.903395322826134e-07,
+      "loss": 0.9172,
+      "step": 67365
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5,
+      "learning_rate": 6.874059149545753e-07,
+      "loss": 0.7879,
+      "step": 67370
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.844785226530426e-07,
+      "loss": 0.9321,
+      "step": 67375
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.815573555615018e-07,
+      "loss": 0.8774,
+      "step": 67380
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.78642413863062e-07,
+      "loss": 0.8636,
+      "step": 67385
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.757336977404327e-07,
+      "loss": 0.9958,
+      "step": 67390
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.447265625,
+      "learning_rate": 6.728312073759458e-07,
+      "loss": 0.9543,
+      "step": 67395
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.51953125,
+      "learning_rate": 6.699349429515111e-07,
+      "loss": 0.8073,
+      "step": 67400
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.45703125,
+      "learning_rate": 6.670449046486837e-07,
+      "loss": 0.7714,
+      "step": 67405
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.641610926486186e-07,
+      "loss": 1.0004,
+      "step": 67410
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.612835071320711e-07,
+      "loss": 0.9713,
+      "step": 67415
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.584121482794082e-07,
+      "loss": 1.0336,
+      "step": 67420
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.55547016270619e-07,
+      "loss": 0.9656,
+      "step": 67425
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.526881112852711e-07,
+      "loss": 0.8656,
+      "step": 67430
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.498354335025881e-07,
+      "loss": 0.9085,
+      "step": 67435
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.469889831013709e-07,
+      "loss": 1.0413,
+      "step": 67440
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4921875,
+      "learning_rate": 6.441487602600327e-07,
+      "loss": 0.8726,
+      "step": 67445
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.413147651566088e-07,
+      "loss": 1.0768,
+      "step": 67450
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.384869979687347e-07,
+      "loss": 0.9141,
+      "step": 67455
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.54296875,
+      "learning_rate": 6.356654588736688e-07,
+      "loss": 0.9616,
+      "step": 67460
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.328501480482474e-07,
+      "loss": 1.0049,
+      "step": 67465
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.300410656689515e-07,
+      "loss": 0.9068,
+      "step": 67470
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.272382119118625e-07,
+      "loss": 1.0419,
+      "step": 67475
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.244415869526398e-07,
+      "loss": 0.9802,
+      "step": 67480
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.216511909666101e-07,
+      "loss": 1.0195,
+      "step": 67485
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.188670241286665e-07,
+      "loss": 1.1275,
+      "step": 67490
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.16089086613325e-07,
+      "loss": 0.9283,
+      "step": 67495
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.133173785947022e-07,
+      "loss": 1.1244,
+      "step": 67500
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.105519002465365e-07,
+      "loss": 0.9914,
+      "step": 67505
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.462890625,
+      "learning_rate": 6.077926517421784e-07,
+      "loss": 0.7442,
+      "step": 67510
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.050396332545782e-07,
+      "loss": 0.9948,
+      "step": 67515
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.02292844956287e-07,
+      "loss": 0.9302,
+      "step": 67520
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4921875,
+      "learning_rate": 5.995522870194891e-07,
+      "loss": 1.0059,
+      "step": 67525
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.968179596159584e-07,
+      "loss": 0.9251,
+      "step": 67530
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.940898629171021e-07,
+      "loss": 0.9118,
+      "step": 67535
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.913679970938946e-07,
+      "loss": 1.0089,
+      "step": 67540
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.886523623169548e-07,
+      "loss": 1.1221,
+      "step": 67545
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.859429587565136e-07,
+      "loss": 0.9205,
+      "step": 67550
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.832397865823791e-07,
+      "loss": 0.9345,
+      "step": 67555
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.805428459640161e-07,
+      "loss": 0.9954,
+      "step": 67560
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.778521370704448e-07,
+      "loss": 0.9087,
+      "step": 67565
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.751676600703415e-07,
+      "loss": 0.8738,
+      "step": 67570
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.724894151319604e-07,
+      "loss": 0.8994,
+      "step": 67575
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.698174024231895e-07,
+      "loss": 0.9824,
+      "step": 67580
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.625,
+      "learning_rate": 5.671516221114947e-07,
+      "loss": 0.8964,
+      "step": 67585
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.64492074363987e-07,
+      "loss": 0.9816,
+      "step": 67590
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.61838759347355e-07,
+      "loss": 0.9253,
+      "step": 67595
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.591916772279326e-07,
+      "loss": 0.9326,
+      "step": 67600
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.565508281716203e-07,
+      "loss": 0.9426,
+      "step": 67605
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.539162123439634e-07,
+      "loss": 1.0343,
+      "step": 67610
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.512878299100966e-07,
+      "loss": 0.9433,
+      "step": 67615
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.486656810347657e-07,
+      "loss": 0.9195,
+      "step": 67620
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.460497658823393e-07,
+      "loss": 1.1784,
+      "step": 67625
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.434400846167864e-07,
+      "loss": 0.9351,
+      "step": 67630
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.40836637401676e-07,
+      "loss": 0.9793,
+      "step": 67635
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.382394244001998e-07,
+      "loss": 0.9489,
+      "step": 67640
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.356484457751609e-07,
+      "loss": 0.9235,
+      "step": 67645
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.330637016889517e-07,
+      "loss": 0.9436,
+      "step": 67650
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.30485192303587e-07,
+      "loss": 1.1746,
+      "step": 67655
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.51171875,
+      "learning_rate": 5.27912917780704e-07,
+      "loss": 0.9042,
+      "step": 67660
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.253468782815296e-07,
+      "loss": 0.9448,
+      "step": 67665
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.515625,
+      "learning_rate": 5.227870739669017e-07,
+      "loss": 0.8279,
+      "step": 67670
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.51171875,
+      "learning_rate": 5.202335049972806e-07,
+      "loss": 0.969,
+      "step": 67675
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.176861715327163e-07,
+      "loss": 0.8491,
+      "step": 67680
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.151450737328811e-07,
+      "loss": 0.9091,
+      "step": 67685
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.126102117570586e-07,
+      "loss": 0.8733,
+      "step": 67690
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.100815857641439e-07,
+      "loss": 1.0445,
+      "step": 67695
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.075591959126103e-07,
+      "loss": 0.9023,
+      "step": 67700
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.050430423605868e-07,
+      "loss": 0.7164,
+      "step": 67705
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.025331252657806e-07,
+      "loss": 0.7707,
+      "step": 67710
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.000294447855103e-07,
+      "loss": 1.0261,
+      "step": 67715
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.975320010767171e-07,
+      "loss": 0.8371,
+      "step": 67720
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.950407942959534e-07,
+      "loss": 0.9907,
+      "step": 67725
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.9255582459935e-07,
+      "loss": 0.9361,
+      "step": 67730
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.900770921426712e-07,
+      "loss": 0.874,
+      "step": 67735
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.876045970813037e-07,
+      "loss": 0.9589,
+      "step": 67740
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.851383395702125e-07,
+      "loss": 1.0567,
+      "step": 67745
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.82678319763985e-07,
+      "loss": 0.8231,
+      "step": 67750
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.802245378168202e-07,
+      "loss": 1.0487,
+      "step": 67755
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.777769938825283e-07,
+      "loss": 0.9396,
+      "step": 67760
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.753356881145199e-07,
+      "loss": 0.9664,
+      "step": 67765
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.729006206658171e-07,
+      "loss": 1.0031,
+      "step": 67770
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.704717916890533e-07,
+      "loss": 0.8784,
+      "step": 67775
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.6804920133647343e-07,
+      "loss": 0.8679,
+      "step": 67780
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.656328497599338e-07,
+      "loss": 1.0123,
+      "step": 67785
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.6322273711089103e-07,
+      "loss": 0.8113,
+      "step": 67790
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.6081886354040207e-07,
+      "loss": 0.9404,
+      "step": 67795
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.584212291991463e-07,
+      "loss": 0.9276,
+      "step": 67800
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.5602983423742584e-07,
+      "loss": 0.999,
+      "step": 67805
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.5364467880512076e-07,
+      "loss": 0.9367,
+      "step": 67810
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.5126576305174474e-07,
+      "loss": 1.0302,
+      "step": 67815
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.488930871264008e-07,
+      "loss": 0.9275,
+      "step": 67820
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.4652665117782545e-07,
+      "loss": 1.0678,
+      "step": 67825
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.441664553543334e-07,
+      "loss": 0.9375,
+      "step": 67830
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.4181249980388416e-07,
+      "loss": 0.9544,
+      "step": 67835
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.394647846740041e-07,
+      "loss": 0.8992,
+      "step": 67840
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.3712331011186433e-07,
+      "loss": 1.0277,
+      "step": 67845
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.3478807626422536e-07,
+      "loss": 1.0133,
+      "step": 67850
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.3245908327747e-07,
+      "loss": 1.0299,
+      "step": 67855
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.3013633129758146e-07,
+      "loss": 0.9226,
+      "step": 67860
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.2781982047014337e-07,
+      "loss": 0.8341,
+      "step": 67865
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.255095509403617e-07,
+      "loss": 0.9564,
+      "step": 67870
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.2320552285304296e-07,
+      "loss": 0.8604,
+      "step": 67875
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.209077363526159e-07,
+      "loss": 0.9471,
+      "step": 67880
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.1861619158309883e-07,
+      "loss": 0.9266,
+      "step": 67885
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.163308886881323e-07,
+      "loss": 0.8708,
+      "step": 67890
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.140518278109684e-07,
+      "loss": 0.9368,
+      "step": 67895
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5,
+      "learning_rate": 4.1177900909445953e-07,
+      "loss": 0.9397,
+      "step": 67900
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.095124326810473e-07,
+      "loss": 0.9737,
+      "step": 67905
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.072520987128292e-07,
+      "loss": 0.8897,
+      "step": 67910
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.0499800733148074e-07,
+      "loss": 0.895,
+      "step": 67915
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.027501586782778e-07,
+      "loss": 1.0609,
+      "step": 67920
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.005085528941299e-07,
+      "loss": 0.8732,
+      "step": 67925
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.982731901195358e-07,
+      "loss": 0.9002,
+      "step": 67930
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.9604407049461667e-07,
+      "loss": 0.9728,
+      "step": 67935
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.9382119415909413e-07,
+      "loss": 0.8315,
+      "step": 67940
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.916045612523123e-07,
+      "loss": 0.9386,
+      "step": 67945
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.625,
+      "learning_rate": 3.8939417191319327e-07,
+      "loss": 1.23,
+      "step": 67950
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.87190026280293e-07,
+      "loss": 0.8856,
+      "step": 67955
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.8499212449176757e-07,
+      "loss": 0.987,
+      "step": 67960
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.828004666853957e-07,
+      "loss": 0.8913,
+      "step": 67965
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.8061505299854525e-07,
+      "loss": 0.9654,
+      "step": 67970
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.784358835681956e-07,
+      "loss": 1.0292,
+      "step": 67975
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.762629585309374e-07,
+      "loss": 0.9346,
+      "step": 67980
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.740962780229951e-07,
+      "loss": 0.9542,
+      "step": 67985
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.7193584218014886e-07,
+      "loss": 0.9173,
+      "step": 67990
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.697816511378349e-07,
+      "loss": 0.949,
+      "step": 67995
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.6763370503107855e-07,
+      "loss": 0.8603,
+      "step": 68000
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.6549200399451646e-07,
+      "loss": 0.9199,
+      "step": 68005
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.633565481623857e-07,
+      "loss": 1.018,
+      "step": 68010
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.61227337668546e-07,
+      "loss": 0.953,
+      "step": 68015
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.59104372646446e-07,
+      "loss": 0.9271,
+      "step": 68020
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.5698765322917935e-07,
+      "loss": 1.0202,
+      "step": 68025
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.5487717954939547e-07,
+      "loss": 0.9461,
+      "step": 68030
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.527729517394107e-07,
+      "loss": 0.9437,
+      "step": 68035
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.506749699310974e-07,
+      "loss": 0.9535,
+      "step": 68040
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.485832342559725e-07,
+      "loss": 0.8304,
+      "step": 68045
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.464977448451423e-07,
+      "loss": 1.1197,
+      "step": 68050
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.444185018293244e-07,
+      "loss": 1.1238,
+      "step": 68055
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.423455053388591e-07,
+      "loss": 1.0381,
+      "step": 68060
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.402787555036757e-07,
+      "loss": 1.038,
+      "step": 68065
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.382182524533262e-07,
+      "loss": 0.9861,
+      "step": 68070
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.361639963169627e-07,
+      "loss": 1.0142,
+      "step": 68075
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.341159872233379e-07,
+      "loss": 0.8862,
+      "step": 68080
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.32074225300838e-07,
+      "loss": 0.9435,
+      "step": 68085
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.300387106774383e-07,
+      "loss": 0.9784,
+      "step": 68090
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.280094434807257e-07,
+      "loss": 0.9197,
+      "step": 68095
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.2598642383789846e-07,
+      "loss": 0.9625,
+      "step": 68100
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.239696518757662e-07,
+      "loss": 0.9105,
+      "step": 68105
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.21959127720739e-07,
+      "loss": 0.8981,
+      "step": 68110
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.199548514988271e-07,
+      "loss": 0.9995,
+      "step": 68115
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.1795682333567443e-07,
+      "loss": 1.0054,
+      "step": 68120
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.1596504335652533e-07,
+      "loss": 0.9117,
+      "step": 68125
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5,
+      "learning_rate": 3.1397951168620207e-07,
+      "loss": 0.8921,
+      "step": 68130
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.120002284491941e-07,
+      "loss": 0.9583,
+      "step": 68135
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.490234375,
+      "learning_rate": 3.100271937695354e-07,
+      "loss": 1.0138,
+      "step": 68140
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.490234375,
+      "learning_rate": 3.080604077709048e-07,
+      "loss": 0.8712,
+      "step": 68145
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.0609987057660383e-07,
+      "loss": 1.0037,
+      "step": 68150
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.0414558230948963e-07,
+      "loss": 0.9395,
+      "step": 68155
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.021975430920865e-07,
+      "loss": 0.9111,
+      "step": 68160
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.002557530464745e-07,
+      "loss": 0.8204,
+      "step": 68165
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.9832021229438955e-07,
+      "loss": 0.9531,
+      "step": 68170
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.963909209571458e-07,
+      "loss": 0.8763,
+      "step": 68175
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.9446787915565765e-07,
+      "loss": 0.9364,
+      "step": 68180
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.9255108701049525e-07,
+      "loss": 1.0371,
+      "step": 68185
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.906405446417848e-07,
+      "loss": 0.919,
+      "step": 68190
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.887362521692749e-07,
+      "loss": 0.8907,
+      "step": 68195
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.868382097123479e-07,
+      "loss": 0.9184,
+      "step": 68200
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.8494641738996406e-07,
+      "loss": 0.9982,
+      "step": 68205
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.8306087532069535e-07,
+      "loss": 1.0517,
+      "step": 68210
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.8118158362275827e-07,
+      "loss": 0.9018,
+      "step": 68215
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.793085424139141e-07,
+      "loss": 1.1069,
+      "step": 68220
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.7744175181158015e-07,
+      "loss": 0.936,
+      "step": 68225
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.7558121193278496e-07,
+      "loss": 0.9315,
+      "step": 68230
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.737269228941242e-07,
+      "loss": 1.0781,
+      "step": 68235
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.718788848118381e-07,
+      "loss": 0.9661,
+      "step": 68240
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.700370978017785e-07,
+      "loss": 0.9609,
+      "step": 68245
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.682015619793643e-07,
+      "loss": 1.0629,
+      "step": 68250
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.6637227745965886e-07,
+      "loss": 0.9062,
+      "step": 68255
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.645492443573372e-07,
+      "loss": 1.0408,
+      "step": 68260
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.627324627866523e-07,
+      "loss": 0.9313,
+      "step": 68265
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.609219328614909e-07,
+      "loss": 0.9128,
+      "step": 68270
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.591176546953289e-07,
+      "loss": 1.0647,
+      "step": 68275
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.573196284012758e-07,
+      "loss": 0.9276,
+      "step": 68280
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.555278540920192e-07,
+      "loss": 1.0192,
+      "step": 68285
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.537423318798804e-07,
+      "loss": 0.8791,
+      "step": 68290
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5,
+      "learning_rate": 2.51963061876781e-07,
+      "loss": 1.1994,
+      "step": 68295
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.501900441942207e-07,
+      "loss": 0.9692,
+      "step": 68300
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.4842327894336603e-07,
+      "loss": 0.9326,
+      "step": 68305
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.466627662349508e-07,
+      "loss": 1.069,
+      "step": 68310
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.449085061793199e-07,
+      "loss": 0.9998,
+      "step": 68315
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.4316049888643e-07,
+      "loss": 0.9248,
+      "step": 68320
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.4141874446585997e-07,
+      "loss": 0.8767,
+      "step": 68325
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.3968324302677815e-07,
+      "loss": 0.8412,
+      "step": 68330
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.3795399467796408e-07,
+      "loss": 1.0728,
+      "step": 68335
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.3623099952782002e-07,
+      "loss": 0.9397,
+      "step": 68340
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.3451425768432621e-07,
+      "loss": 0.8171,
+      "step": 68345
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.3280376925511881e-07,
+      "loss": 1.077,
+      "step": 68350
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.3109953434737875e-07,
+      "loss": 0.8722,
+      "step": 68355
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.2940155306796497e-07,
+      "loss": 0.8994,
+      "step": 68360
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.2770982552328125e-07,
+      "loss": 1.0112,
+      "step": 68365
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.260243518193761e-07,
+      "loss": 0.9184,
+      "step": 68370
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.2434513206189833e-07,
+      "loss": 0.9368,
+      "step": 68375
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.2267216635609711e-07,
+      "loss": 1.0861,
+      "step": 68380
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.625,
+      "learning_rate": 2.210054548068552e-07,
+      "loss": 0.9625,
+      "step": 68385
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.1934499751862236e-07,
+      "loss": 0.8768,
+      "step": 68390
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.1769079459548204e-07,
+      "loss": 0.9296,
+      "step": 68395
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.1604284614112902e-07,
+      "loss": 0.856,
+      "step": 68400
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.144011522588585e-07,
+      "loss": 0.8969,
+      "step": 68405
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.127657130515548e-07,
+      "loss": 0.8386,
+      "step": 68410
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.1113652862175816e-07,
+      "loss": 0.8712,
+      "step": 68415
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.0951359907157575e-07,
+      "loss": 0.8142,
+      "step": 68420
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.0789692450272624e-07,
+      "loss": 0.9764,
+      "step": 68425
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.062865050165508e-07,
+      "loss": 0.8942,
+      "step": 68430
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.0468234071400194e-07,
+      "loss": 0.95,
+      "step": 68435
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.0308443169561042e-07,
+      "loss": 1.1128,
+      "step": 68440
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.0149277806155164e-07,
+      "loss": 0.9583,
+      "step": 68445
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.9990737991159024e-07,
+      "loss": 1.0166,
+      "step": 68450
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.983282373450801e-07,
+      "loss": 0.8655,
+      "step": 68455
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.9675535046104198e-07,
+      "loss": 0.9587,
+      "step": 68460
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.951887193580304e-07,
+      "loss": 0.9855,
+      "step": 68465
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.936283441342668e-07,
+      "loss": 0.8688,
+      "step": 68470
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.9207422488753957e-07,
+      "loss": 1.0159,
+      "step": 68475
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.9052636171528193e-07,
+      "loss": 1.0702,
+      "step": 68480
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.8898475471449405e-07,
+      "loss": 1.0368,
+      "step": 68485
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.8744940398182088e-07,
+      "loss": 1.0001,
+      "step": 68490
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.8592030961349648e-07,
+      "loss": 1.0321,
+      "step": 68495
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.843974717053776e-07,
+      "loss": 1.0103,
+      "step": 68500
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.828808903528878e-07,
+      "loss": 0.9373,
+      "step": 68505
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.8137056565111777e-07,
+      "loss": 0.9273,
+      "step": 68510
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.7986649769471398e-07,
+      "loss": 1.1433,
+      "step": 68515
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.7836868657797878e-07,
+      "loss": 1.0121,
+      "step": 68520
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.7687713239477043e-07,
+      "loss": 0.9412,
+      "step": 68525
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.7539183523859192e-07,
+      "loss": 0.9708,
+      "step": 68530
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.7391279520254654e-07,
+      "loss": 0.8521,
+      "step": 68535
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.7244001237933793e-07,
+      "loss": 0.8871,
+      "step": 68540
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.709734868612922e-07,
+      "loss": 1.0056,
+      "step": 68545
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.6951321874031367e-07,
+      "loss": 0.9889,
+      "step": 68550
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.6805920810795128e-07,
+      "loss": 0.8824,
+      "step": 68555
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.6661145505533215e-07,
+      "loss": 0.9732,
+      "step": 68560
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.6516995967320593e-07,
+      "loss": 1.0555,
+      "step": 68565
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.6373472205193363e-07,
+      "loss": 0.9849,
+      "step": 68570
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.6230574228146557e-07,
+      "loss": 0.9778,
+      "step": 68575
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.608830204513856e-07,
+      "loss": 1.0558,
+      "step": 68580
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.5946655665086685e-07,
+      "loss": 1.0651,
+      "step": 68585
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.5805635096869387e-07,
+      "loss": 0.8778,
+      "step": 68590
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.566524034932515e-07,
+      "loss": 0.9079,
+      "step": 68595
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.5525471431254713e-07,
+      "loss": 1.0091,
+      "step": 68600
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.5386328351419953e-07,
+      "loss": 0.9609,
+      "step": 68605
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.5247811118541676e-07,
+      "loss": 0.9568,
+      "step": 68610
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.5109919741301826e-07,
+      "loss": 0.9626,
+      "step": 68615
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.4972654228343486e-07,
+      "loss": 0.9385,
+      "step": 68620
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.4836014588271996e-07,
+      "loss": 0.8519,
+      "step": 68625
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.4700000829651617e-07,
+      "loss": 0.9085,
+      "step": 68630
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.456461296100664e-07,
+      "loss": 0.9853,
+      "step": 68635
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.4429850990824723e-07,
+      "loss": 0.9127,
+      "step": 68640
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.429571492755244e-07,
+      "loss": 0.9672,
+      "step": 68645
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.416220477959751e-07,
+      "loss": 0.9399,
+      "step": 68650
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.4029320555327684e-07,
+      "loss": 0.9394,
+      "step": 68655
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.389706226307408e-07,
+      "loss": 1.0412,
+      "step": 68660
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.3765429911124507e-07,
+      "loss": 0.9976,
+      "step": 68665
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.3634423507732362e-07,
+      "loss": 0.8278,
+      "step": 68670
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.3504043061107752e-07,
+      "loss": 1.0012,
+      "step": 68675
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.3374288579424132e-07,
+      "loss": 0.848,
+      "step": 68680
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.3245160070812778e-07,
+      "loss": 1.0536,
+      "step": 68685
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3116657543369436e-07,
+      "loss": 1.0768,
+      "step": 68690
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2988781005147666e-07,
+      "loss": 1.0061,
+      "step": 68695
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.2861530464163273e-07,
+      "loss": 0.88,
+      "step": 68700
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.2734905928393215e-07,
+      "loss": 1.0937,
+      "step": 68705
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.2608907405773362e-07,
+      "loss": 1.0532,
+      "step": 68710
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.2483534904200734e-07,
+      "loss": 0.9389,
+      "step": 68715
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.2358788431536816e-07,
+      "loss": 0.9631,
+      "step": 68720
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.2234667995598693e-07,
+      "loss": 1.0137,
+      "step": 68725
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.2111173604165692e-07,
+      "loss": 0.9438,
+      "step": 68730
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.1988305264980516e-07,
+      "loss": 1.0743,
+      "step": 68735
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.1866062985743664e-07,
+      "loss": 0.9187,
+      "step": 68740
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.1744446774116791e-07,
+      "loss": 0.9324,
+      "step": 68745
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.1623456637723795e-07,
+      "loss": 0.9166,
+      "step": 68750
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.1503092584148611e-07,
+      "loss": 1.0688,
+      "step": 68755
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.1383354620936315e-07,
+      "loss": 0.8598,
+      "step": 68760
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.1264242755590903e-07,
+      "loss": 1.0886,
+      "step": 68765
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.1145756995578626e-07,
+      "loss": 0.9614,
+      "step": 68770
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.1027897348326877e-07,
+      "loss": 0.8253,
+      "step": 68775
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.0910663821221967e-07,
+      "loss": 1.011,
+      "step": 68780
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5,
+      "learning_rate": 1.0794056421614684e-07,
+      "loss": 0.9053,
+      "step": 68785
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.0678075156812517e-07,
+      "loss": 1.0008,
+      "step": 68790
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.0562720034085205e-07,
+      "loss": 0.9327,
+      "step": 68795
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.0447991060663631e-07,
+      "loss": 0.9186,
+      "step": 68800
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.0333888243738709e-07,
+      "loss": 0.856,
+      "step": 68805
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.0220411590463607e-07,
+      "loss": 0.9629,
+      "step": 68810
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.0107561107950414e-07,
+      "loss": 0.9257,
+      "step": 68815
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.995336803272359e-08,
+      "loss": 0.933,
+      "step": 68820
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.883738683464927e-08,
+      "loss": 0.8716,
+      "step": 68825
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.772766755522522e-08,
+      "loss": 0.7858,
+      "step": 68830
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.662421026400692e-08,
+      "loss": 0.8859,
+      "step": 68835
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.552701503016126e-08,
+      "loss": 0.9141,
+      "step": 68840
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.443608192246655e-08,
+      "loss": 0.7986,
+      "step": 68845
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.335141100930145e-08,
+      "loss": 0.8627,
+      "step": 68850
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.51953125,
+      "learning_rate": 9.227300235865599e-08,
+      "loss": 1.0326,
+      "step": 68855
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.120085603812056e-08,
+      "loss": 0.8892,
+      "step": 68860
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.013497211489696e-08,
+      "loss": 0.8939,
+      "step": 68865
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.907535065580952e-08,
+      "loss": 0.8497,
+      "step": 68870
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.802199172726066e-08,
+      "loss": 1.0038,
+      "step": 68875
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5234375,
+      "learning_rate": 8.697489539529757e-08,
+      "loss": 0.9418,
+      "step": 68880
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.593406172552332e-08,
+      "loss": 0.8043,
+      "step": 68885
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.489949078320792e-08,
+      "loss": 0.9,
+      "step": 68890
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.387118263317728e-08,
+      "loss": 0.934,
+      "step": 68895
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.284913733990208e-08,
+      "loss": 1.1524,
+      "step": 68900
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.183335496743105e-08,
+      "loss": 0.9928,
+      "step": 68905
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.082383557944662e-08,
+      "loss": 0.9562,
+      "step": 68910
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.982057923922037e-08,
+      "loss": 1.0204,
+      "step": 68915
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.882358600964646e-08,
+      "loss": 0.9144,
+      "step": 68920
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.783285595320822e-08,
+      "loss": 0.8462,
+      "step": 68925
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.684838913200043e-08,
+      "loss": 0.9862,
+      "step": 68930
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.58701856077515e-08,
+      "loss": 0.8919,
+      "step": 68935
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.489824544175683e-08,
+      "loss": 0.9889,
+      "step": 68940
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.393256869494547e-08,
+      "loss": 1.0822,
+      "step": 68945
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.297315542784678e-08,
+      "loss": 0.8859,
+      "step": 68950
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.478515625,
+      "learning_rate": 7.202000570060153e-08,
+      "loss": 0.9169,
+      "step": 68955
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.625,
+      "learning_rate": 7.107311957293971e-08,
+      "loss": 0.9741,
+      "step": 68960
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.013249710423608e-08,
+      "loss": 0.9086,
+      "step": 68965
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.919813835343237e-08,
+      "loss": 0.8493,
+      "step": 68970
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.827004337910391e-08,
+      "loss": 0.9377,
+      "step": 68975
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.734821223941535e-08,
+      "loss": 0.9256,
+      "step": 68980
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.643264499216483e-08,
+      "loss": 0.9293,
+      "step": 68985
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.552334169472874e-08,
+      "loss": 0.992,
+      "step": 68990
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.462030240409478e-08,
+      "loss": 0.8916,
+      "step": 68995
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.372352717688434e-08,
+      "loss": 1.0083,
+      "step": 69000
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.283301606930803e-08,
+      "loss": 0.8847,
+      "step": 69005
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.194876913716563e-08,
+      "loss": 0.9906,
+      "step": 69010
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.52734375,
+      "learning_rate": 6.107078643590169e-08,
+      "loss": 1.1935,
+      "step": 69015
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.019906802053887e-08,
+      "loss": 0.8415,
+      "step": 69020
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.9333613945722344e-08,
+      "loss": 0.9834,
+      "step": 69025
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.84744242656976e-08,
+      "loss": 0.9222,
+      "step": 69030
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.762149903432157e-08,
+      "loss": 1.1191,
+      "step": 69035
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.677483830505148e-08,
+      "loss": 0.9117,
+      "step": 69040
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.5934442130956e-08,
+      "loss": 0.9156,
+      "step": 69045
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.510031056472631e-08,
+      "loss": 0.877,
+      "step": 69050
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.4272443658631714e-08,
+      "loss": 1.0652,
+      "step": 69055
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.3450841464564027e-08,
+      "loss": 0.9143,
+      "step": 69060
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.263550403402651e-08,
+      "loss": 0.9403,
+      "step": 69065
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.1826431418133816e-08,
+      "loss": 0.9235,
+      "step": 69070
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.1023623667589834e-08,
+      "loss": 0.8941,
+      "step": 69075
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.0227080832720985e-08,
+      "loss": 0.9721,
+      "step": 69080
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.9436802963442886e-08,
+      "loss": 1.0108,
+      "step": 69085
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.865279010930479e-08,
+      "loss": 0.9499,
+      "step": 69090
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.787504231944517e-08,
+      "loss": 1.0272,
+      "step": 69095
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.710355964262503e-08,
+      "loss": 0.9583,
+      "step": 69100
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.633834212717236e-08,
+      "loss": 0.9453,
+      "step": 69105
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.5579389821082117e-08,
+      "loss": 1.0074,
+      "step": 69110
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.482670277190515e-08,
+      "loss": 0.8634,
+      "step": 69115
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.4080281026837036e-08,
+      "loss": 0.8821,
+      "step": 69120
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.3340124632651465e-08,
+      "loss": 0.8397,
+      "step": 69125
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.2606233635755775e-08,
+      "loss": 1.0284,
+      "step": 69130
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.1878608082135397e-08,
+      "loss": 0.9987,
+      "step": 69135
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.115724801740939e-08,
+      "loss": 0.7926,
+      "step": 69140
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.044215348678604e-08,
+      "loss": 0.8805,
+      "step": 69145
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.973332453509615e-08,
+      "loss": 0.9702,
+      "step": 69150
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.903076120677085e-08,
+      "loss": 0.995,
+      "step": 69155
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.833446354584158e-08,
+      "loss": 0.8779,
+      "step": 69160
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.7644431595962314e-08,
+      "loss": 0.8399,
+      "step": 69165
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.6960665400365136e-08,
+      "loss": 0.7854,
+      "step": 69170
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.628316500192686e-08,
+      "loss": 0.9082,
+      "step": 69175
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.56119304431024e-08,
+      "loss": 1.0525,
+      "step": 69180
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.494696176598034e-08,
+      "loss": 0.8938,
+      "step": 69185
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.428825901222732e-08,
+      "loss": 0.9169,
+      "step": 69190
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.363582222314365e-08,
+      "loss": 0.9887,
+      "step": 69195
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.298965143961885e-08,
+      "loss": 0.952,
+      "step": 69200
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.234974670215385e-08,
+      "loss": 1.1259,
+      "step": 69205
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.1716108050861005e-08,
+      "loss": 0.9038,
+      "step": 69210
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.1088735525453e-08,
+      "loss": 1.0463,
+      "step": 69215
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.0467629165265025e-08,
+      "loss": 0.9903,
+      "step": 69220
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.98527890092104e-08,
+      "loss": 1.0275,
+      "step": 69225
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.9244215095847184e-08,
+      "loss": 1.0958,
+      "step": 69230
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.8641907463311524e-08,
+      "loss": 0.9467,
+      "step": 69235
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.804586614936211e-08,
+      "loss": 0.9539,
+      "step": 69240
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.7456091191357945e-08,
+      "loss": 0.9948,
+      "step": 69245
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.6872582626258357e-08,
+      "loss": 1.0394,
+      "step": 69250
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.6295340490645192e-08,
+      "loss": 1.0524,
+      "step": 69255
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.5724364820711722e-08,
+      "loss": 0.9632,
+      "step": 69260
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.515965565222933e-08,
+      "loss": 0.9515,
+      "step": 69265
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4921875,
+      "learning_rate": 2.4601213020591928e-08,
+      "loss": 0.8043,
+      "step": 69270
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.4049036960827053e-08,
+      "loss": 0.8569,
+      "step": 69275
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.3503127507518153e-08,
+      "loss": 0.8868,
+      "step": 69280
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.2963484694904504e-08,
+      "loss": 0.8159,
+      "step": 69285
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.24301085568146e-08,
+      "loss": 1.0896,
+      "step": 69290
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5,
+      "learning_rate": 2.1902999126655054e-08,
+      "loss": 0.9988,
+      "step": 69295
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.138215643748831e-08,
+      "loss": 0.7695,
+      "step": 69300
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.086758052194382e-08,
+      "loss": 1.0433,
+      "step": 69305
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.0359271412295766e-08,
+      "loss": 0.9876,
+      "step": 69310
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.985722914039645e-08,
+      "loss": 0.9633,
+      "step": 69315
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.936145373770959e-08,
+      "loss": 1.1043,
+      "step": 69320
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.887194523532143e-08,
+      "loss": 1.0057,
+      "step": 69325
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.8388703663907436e-08,
+      "loss": 1.0526,
+      "step": 69330
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.791172905375449e-08,
+      "loss": 0.9354,
+      "step": 69335
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.7441021434760896e-08,
+      "loss": 0.8695,
+      "step": 69340
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.6976580836436385e-08,
+      "loss": 0.9798,
+      "step": 69345
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.6518407287902105e-08,
+      "loss": 0.8781,
+      "step": 69350
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.606650081785732e-08,
+      "loss": 0.9757,
+      "step": 69355
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.562086145463493e-08,
+      "loss": 0.9517,
+      "step": 69360
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.5181489226168132e-08,
+      "loss": 0.8726,
+      "step": 69365
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.4748384160001571e-08,
+      "loss": 1.044,
+      "step": 69370
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.4321546283280197e-08,
+      "loss": 0.9414,
+      "step": 69375
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.3900975622760381e-08,
+      "loss": 0.9418,
+      "step": 69380
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.3486672204798822e-08,
+      "loss": 1.0239,
+      "step": 69385
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.3078636055374738e-08,
+      "loss": 1.046,
+      "step": 69390
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2676867200045461e-08,
+      "loss": 0.9348,
+      "step": 69395
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.2281365664013056e-08,
+      "loss": 0.8424,
+      "step": 69400
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1892131472068801e-08,
+      "loss": 1.0332,
+      "step": 69405
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.1509164648593195e-08,
+      "loss": 1.0344,
+      "step": 69410
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.1132465217600363e-08,
+      "loss": 1.0642,
+      "step": 69415
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.0762033202704746e-08,
+      "loss": 0.9868,
+      "step": 69420
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.0397868627121111e-08,
+      "loss": 0.951,
+      "step": 69425
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.0039971513675638e-08,
+      "loss": 0.8531,
+      "step": 69430
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.68834188480594e-09,
+      "loss": 1.0586,
+      "step": 69435
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.486328125,
+      "learning_rate": 9.342979762549942e-09,
+      "loss": 0.8869,
+      "step": 69440
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.003885168556991e-09,
+      "loss": 0.9267,
+      "step": 69445
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.515625,
+      "learning_rate": 8.67105812408786e-09,
+      "loss": 1.0981,
+      "step": 69450
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.344498649981436e-09,
+      "loss": 0.8177,
+      "step": 69455
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.024206766732435e-09,
+      "loss": 1.07,
+      "step": 69460
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.710182494413687e-09,
+      "loss": 0.9687,
+      "step": 69465
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.402425852687245e-09,
+      "loss": 0.9563,
+      "step": 69470
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.100936860870988e-09,
+      "loss": 1.0244,
+      "step": 69475
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5234375,
+      "learning_rate": 6.8057155378498105e-09,
+      "loss": 0.9991,
+      "step": 69480
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.5167619021200274e-09,
+      "loss": 1.0204,
+      "step": 69485
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.2340759718115815e-09,
+      "loss": 1.0564,
+      "step": 69490
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.498046875,
+      "learning_rate": 5.957657764632529e-09,
+      "loss": 0.9994,
+      "step": 69495
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.687507297913452e-09,
+      "loss": 0.9983,
+      "step": 69500
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.42362458857415e-09,
+      "loss": 0.8027,
+      "step": 69505
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.166009653179149e-09,
+      "loss": 0.9545,
+      "step": 69510
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.9146625078599906e-09,
+      "loss": 0.9419,
+      "step": 69515
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.669583168370739e-09,
+      "loss": 1.0382,
+      "step": 69520
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.4307716500768816e-09,
+      "loss": 0.9006,
+      "step": 69525
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.198227967944224e-09,
+      "loss": 0.8615,
+      "step": 69530
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4921875,
+      "learning_rate": 3.9719521365610964e-09,
+      "loss": 1.0192,
+      "step": 69535
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.751944170105048e-09,
+      "loss": 1.0237,
+      "step": 69540
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.5382040823539464e-09,
+      "loss": 0.9387,
+      "step": 69545
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.330731886719285e-09,
+      "loss": 0.8864,
+      "step": 69550
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.1295275961906733e-09,
+      "loss": 0.9047,
+      "step": 69555
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.93459122340245e-09,
+      "loss": 0.8942,
+      "step": 69560
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.745922780555965e-09,
+      "loss": 0.8649,
+      "step": 69565
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.563522279486197e-09,
+      "loss": 1.0669,
+      "step": 69570
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.3873897316173397e-09,
+      "loss": 0.9068,
+      "step": 69575
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.2175251480072156e-09,
+      "loss": 0.8953,
+      "step": 69580
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.0539285392806584e-09,
+      "loss": 0.9438,
+      "step": 69585
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.896599915696129e-09,
+      "loss": 1.0569,
+      "step": 69590
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.74553928712351e-09,
+      "loss": 0.9274,
+      "step": 69595
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.6007466630330036e-09,
+      "loss": 0.8161,
+      "step": 69600
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.4622220524951324e-09,
+      "loss": 1.098,
+      "step": 69605
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.3299654642029425e-09,
+      "loss": 0.9883,
+      "step": 69610
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.2039769064275952e-09,
+      "loss": 0.9939,
+      "step": 69615
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.0842563870738788e-09,
+      "loss": 0.9728,
+      "step": 69620
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.708039136580028e-10,
+      "loss": 0.8537,
+      "step": 69625
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.51171875,
+      "learning_rate": 8.636194932742925e-10,
+      "loss": 0.9281,
+      "step": 69630
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.627031326395973e-10,
+      "loss": 0.9818,
+      "step": 69635
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.680548381043927e-10,
+      "loss": 0.979,
+      "step": 69640
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.796746155750654e-10,
+      "loss": 0.9615,
+      "step": 69645
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.975624706027305e-10,
+      "loss": 0.932,
+      "step": 69650
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.2171840833882257e-10,
+      "loss": 0.9145,
+      "step": 69655
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.5214243353509646e-10,
+      "loss": 1.0162,
+      "step": 69660
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.45703125,
+      "learning_rate": 2.888345505436263e-10,
+      "loss": 1.0425,
+      "step": 69665
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.3179476332790827e-10,
+      "loss": 0.9054,
+      "step": 69670
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.8102307548506503e-10,
+      "loss": 0.9386,
+      "step": 69675
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.3651949017923216e-10,
+      "loss": 0.9464,
+      "step": 69680
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.828401021927392e-11,
+      "loss": 0.9135,
+      "step": 69685
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.631663798106758e-11,
+      "loss": 0.8707,
+      "step": 69690
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.061737547411681e-11,
+      "loss": 0.9513,
+      "step": 69695
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.1186224308245018e-11,
+      "loss": 0.9132,
+      "step": 69700
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.023185715799742e-12,
+      "loss": 1.0332,
+      "step": 69705
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.1282605072437947e-12,
+      "loss": 0.8443,
+      "step": 69710
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.9567210078239441,
+      "eval_runtime": 6401.306,
+      "eval_samples_per_second": 2.411,
+      "eval_steps_per_second": 1.205,
+      "step": 69713
+    },
+    {
+      "epoch": 1.0,
+      "step": 69713,
+      "total_flos": 1.225431715317547e+19,
+      "train_loss": 0.9623549959262275,
+      "train_runtime": 219590.8488,
+      "train_samples_per_second": 0.635,
+      "train_steps_per_second": 0.317
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 69713,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "total_flos": 1.225431715317547e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}