diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,25602 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 18257,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.095290251916758e-07,
+      "loss": 1.5925,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.38671875,
+      "learning_rate": 5.47645125958379e-07,
+      "loss": 1.7789,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.095290251916758e-06,
+      "loss": 1.7766,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.642935377875137e-06,
+      "loss": 1.639,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.190580503833516e-06,
+      "loss": 1.6936,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.738225629791895e-06,
+      "loss": 1.7455,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.285870755750274e-06,
+      "loss": 1.6397,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.8335158817086525e-06,
+      "loss": 1.6603,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.381161007667032e-06,
+      "loss": 1.7611,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.928806133625411e-06,
+      "loss": 1.6704,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.3828125,
+      "learning_rate": 5.47645125958379e-06,
+      "loss": 1.7226,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.024096385542169e-06,
+      "loss": 1.7409,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.571741511500548e-06,
+      "loss": 1.6828,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.119386637458927e-06,
+      "loss": 1.677,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.4609375,
+      "learning_rate": 7.667031763417305e-06,
+      "loss": 1.7095,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.214676889375684e-06,
+      "loss": 1.5529,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.762322015334064e-06,
+      "loss": 1.6196,
+      "step": 80
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.453125,
+      "learning_rate": 9.309967141292443e-06,
+      "loss": 1.6254,
+      "step": 85
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.857612267250823e-06,
+      "loss": 1.6312,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.0405257393209202e-05,
+      "loss": 1.5719,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.095290251916758e-05,
+      "loss": 1.5849,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.1500547645125959e-05,
+      "loss": 1.5305,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2048192771084338e-05,
+      "loss": 1.5214,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.2595837897042718e-05,
+      "loss": 1.5621,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.3143483023001096e-05,
+      "loss": 1.4636,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.3691128148959475e-05,
+      "loss": 1.4968,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.4238773274917854e-05,
+      "loss": 1.5992,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.4786418400876234e-05,
+      "loss": 1.4915,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 1.533406352683461e-05,
+      "loss": 1.4276,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.588170865279299e-05,
+      "loss": 1.4851,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.642935377875137e-05,
+      "loss": 1.4907,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.697699890470975e-05,
+      "loss": 1.4977,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.7524644030668127e-05,
+      "loss": 1.3028,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.8072289156626505e-05,
+      "loss": 1.4589,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.8619934282584886e-05,
+      "loss": 1.4321,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.9167579408543264e-05,
+      "loss": 1.3841,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.9715224534501645e-05,
+      "loss": 1.3932,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.0262869660460023e-05,
+      "loss": 1.3728,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 2.0810514786418404e-05,
+      "loss": 1.3659,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.1358159912376778e-05,
+      "loss": 1.4092,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.190580503833516e-05,
+      "loss": 1.4435,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1953125,
+      "learning_rate": 2.2453450164293537e-05,
+      "loss": 1.3881,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.3001095290251918e-05,
+      "loss": 1.3537,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.3548740416210296e-05,
+      "loss": 1.3825,
+      "step": 215
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 2.4096385542168677e-05,
+      "loss": 1.3656,
+      "step": 220
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 2.4644030668127055e-05,
+      "loss": 1.3531,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 2.5191675794085436e-05,
+      "loss": 1.3594,
+      "step": 230
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 2.5739320920043813e-05,
+      "loss": 1.3433,
+      "step": 235
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 2.628696604600219e-05,
+      "loss": 1.3442,
+      "step": 240
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 2.6834611171960572e-05,
+      "loss": 1.4246,
+      "step": 245
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 2.738225629791895e-05,
+      "loss": 1.3524,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 2.792990142387733e-05,
+      "loss": 1.3018,
+      "step": 255
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.847754654983571e-05,
+      "loss": 1.4327,
+      "step": 260
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 2.902519167579409e-05,
+      "loss": 1.2796,
+      "step": 265
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.9572836801752468e-05,
+      "loss": 1.4626,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.12890625,
+      "learning_rate": 3.012048192771085e-05,
+      "loss": 1.3257,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 3.066812705366922e-05,
+      "loss": 1.3463,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 3.12157721796276e-05,
+      "loss": 1.3423,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.12890625,
+      "learning_rate": 3.176341730558598e-05,
+      "loss": 1.3721,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 3.231106243154436e-05,
+      "loss": 1.2935,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.14453125,
+      "learning_rate": 3.285870755750274e-05,
+      "loss": 1.3651,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.14453125,
+      "learning_rate": 3.3406352683461115e-05,
+      "loss": 1.3716,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 3.39539978094195e-05,
+      "loss": 1.237,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.13671875,
+      "learning_rate": 3.450164293537788e-05,
+      "loss": 1.3898,
+      "step": 315
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.130859375,
+      "learning_rate": 3.5049288061336255e-05,
+      "loss": 1.2465,
+      "step": 320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.138671875,
+      "learning_rate": 3.559693318729463e-05,
+      "loss": 1.308,
+      "step": 325
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 3.614457831325301e-05,
+      "loss": 1.2948,
+      "step": 330
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.12890625,
+      "learning_rate": 3.6692223439211395e-05,
+      "loss": 1.3287,
+      "step": 335
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.130859375,
+      "learning_rate": 3.723986856516977e-05,
+      "loss": 1.3124,
+      "step": 340
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 3.778751369112815e-05,
+      "loss": 1.3206,
+      "step": 345
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.140625,
+      "learning_rate": 3.833515881708653e-05,
+      "loss": 1.354,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 3.888280394304491e-05,
+      "loss": 1.2987,
+      "step": 355
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1328125,
+      "learning_rate": 3.943044906900329e-05,
+      "loss": 1.2752,
+      "step": 360
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.134765625,
+      "learning_rate": 3.997809419496167e-05,
+      "loss": 1.3191,
+      "step": 365
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 4.0525739320920046e-05,
+      "loss": 1.3768,
+      "step": 370
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.126953125,
+      "learning_rate": 4.107338444687843e-05,
+      "loss": 1.2676,
+      "step": 375
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.142578125,
+      "learning_rate": 4.162102957283681e-05,
+      "loss": 1.2931,
+      "step": 380
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 4.2168674698795186e-05,
+      "loss": 1.2615,
+      "step": 385
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 4.2716319824753556e-05,
+      "loss": 1.311,
+      "step": 390
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 4.326396495071194e-05,
+      "loss": 1.2629,
+      "step": 395
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 4.381161007667032e-05,
+      "loss": 1.3424,
+      "step": 400
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.140625,
+      "learning_rate": 4.4359255202628696e-05,
+      "loss": 1.2941,
+      "step": 405
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 4.4906900328587074e-05,
+      "loss": 1.2161,
+      "step": 410
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 1.3718,
+      "step": 415
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.134765625,
+      "learning_rate": 4.6002190580503836e-05,
+      "loss": 1.3224,
+      "step": 420
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.14453125,
+      "learning_rate": 4.6549835706462214e-05,
+      "loss": 1.3083,
+      "step": 425
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.142578125,
+      "learning_rate": 4.709748083242059e-05,
+      "loss": 1.2505,
+      "step": 430
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 4.764512595837897e-05,
+      "loss": 1.2964,
+      "step": 435
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 4.8192771084337354e-05,
+      "loss": 1.264,
+      "step": 440
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 4.874041621029573e-05,
+      "loss": 1.2452,
+      "step": 445
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 4.928806133625411e-05,
+      "loss": 1.3098,
+      "step": 450
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 4.983570646221249e-05,
+      "loss": 1.2268,
+      "step": 455
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.158203125,
+      "learning_rate": 5.038335158817087e-05,
+      "loss": 1.3172,
+      "step": 460
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 5.093099671412924e-05,
+      "loss": 1.2184,
+      "step": 465
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.138671875,
+      "learning_rate": 5.147864184008763e-05,
+      "loss": 1.2912,
+      "step": 470
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.146484375,
+      "learning_rate": 5.2026286966046e-05,
+      "loss": 1.2859,
+      "step": 475
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.140625,
+      "learning_rate": 5.257393209200438e-05,
+      "loss": 1.2723,
+      "step": 480
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.14453125,
+      "learning_rate": 5.312157721796276e-05,
+      "loss": 1.2359,
+      "step": 485
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 5.3669222343921145e-05,
+      "loss": 1.3044,
+      "step": 490
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 5.4216867469879516e-05,
+      "loss": 1.2649,
+      "step": 495
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 5.47645125958379e-05,
+      "loss": 1.323,
+      "step": 500
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 5.531215772179628e-05,
+      "loss": 1.2744,
+      "step": 505
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.154296875,
+      "learning_rate": 5.585980284775466e-05,
+      "loss": 1.2393,
+      "step": 510
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 5.640744797371303e-05,
+      "loss": 1.2021,
+      "step": 515
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1484375,
+      "learning_rate": 5.695509309967142e-05,
+      "loss": 1.341,
+      "step": 520
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.146484375,
+      "learning_rate": 5.7502738225629795e-05,
+      "loss": 1.2942,
+      "step": 525
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 5.805038335158818e-05,
+      "loss": 1.2963,
+      "step": 530
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 5.859802847754655e-05,
+      "loss": 1.355,
+      "step": 535
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.16015625,
+      "learning_rate": 5.9145673603504935e-05,
+      "loss": 1.2284,
+      "step": 540
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 5.969331872946331e-05,
+      "loss": 1.2243,
+      "step": 545
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 6.02409638554217e-05,
+      "loss": 1.3001,
+      "step": 550
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.146484375,
+      "learning_rate": 6.078860898138007e-05,
+      "loss": 1.238,
+      "step": 555
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.15234375,
+      "learning_rate": 6.133625410733844e-05,
+      "loss": 1.2305,
+      "step": 560
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 6.188389923329682e-05,
+      "loss": 1.2445,
+      "step": 565
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.279296875,
+      "learning_rate": 6.24315443592552e-05,
+      "loss": 1.2829,
+      "step": 570
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 6.297918948521358e-05,
+      "loss": 1.2107,
+      "step": 575
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.146484375,
+      "learning_rate": 6.352683461117196e-05,
+      "loss": 1.2449,
+      "step": 580
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 6.407447973713035e-05,
+      "loss": 1.2306,
+      "step": 585
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.462212486308872e-05,
+      "loss": 1.2696,
+      "step": 590
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1484375,
+      "learning_rate": 6.51697699890471e-05,
+      "loss": 1.3035,
+      "step": 595
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.158203125,
+      "learning_rate": 6.571741511500547e-05,
+      "loss": 1.2613,
+      "step": 600
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 6.626506024096386e-05,
+      "loss": 1.2877,
+      "step": 605
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 6.681270536692223e-05,
+      "loss": 1.229,
+      "step": 610
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.15625,
+      "learning_rate": 6.736035049288061e-05,
+      "loss": 1.2112,
+      "step": 615
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 6.7907995618839e-05,
+      "loss": 1.265,
+      "step": 620
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.15625,
+      "learning_rate": 6.845564074479738e-05,
+      "loss": 1.2641,
+      "step": 625
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1484375,
+      "learning_rate": 6.900328587075575e-05,
+      "loss": 1.2364,
+      "step": 630
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.16015625,
+      "learning_rate": 6.955093099671414e-05,
+      "loss": 1.2265,
+      "step": 635
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.009857612267251e-05,
+      "loss": 1.2454,
+      "step": 640
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.06462212486309e-05,
+      "loss": 1.2541,
+      "step": 645
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 7.119386637458927e-05,
+      "loss": 1.3241,
+      "step": 650
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.158203125,
+      "learning_rate": 7.174151150054765e-05,
+      "loss": 1.2416,
+      "step": 655
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 7.228915662650602e-05,
+      "loss": 1.2588,
+      "step": 660
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 7.28368017524644e-05,
+      "loss": 1.2438,
+      "step": 665
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.338444687842279e-05,
+      "loss": 1.3034,
+      "step": 670
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 7.393209200438116e-05,
+      "loss": 1.2278,
+      "step": 675
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 7.447973713033955e-05,
+      "loss": 1.2721,
+      "step": 680
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.502738225629792e-05,
+      "loss": 1.2244,
+      "step": 685
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.16015625,
+      "learning_rate": 7.55750273822563e-05,
+      "loss": 1.3325,
+      "step": 690
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.612267250821467e-05,
+      "loss": 1.2812,
+      "step": 695
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.667031763417306e-05,
+      "loss": 1.2739,
+      "step": 700
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.721796276013144e-05,
+      "loss": 1.2815,
+      "step": 705
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.162109375,
+      "learning_rate": 7.776560788608982e-05,
+      "loss": 1.2886,
+      "step": 710
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 7.83132530120482e-05,
+      "loss": 1.2802,
+      "step": 715
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.886089813800658e-05,
+      "loss": 1.3426,
+      "step": 720
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.940854326396495e-05,
+      "loss": 1.3236,
+      "step": 725
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 7.995618838992334e-05,
+      "loss": 1.214,
+      "step": 730
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.166015625,
+      "learning_rate": 8.05038335158817e-05,
+      "loss": 1.3208,
+      "step": 735
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 8.105147864184009e-05,
+      "loss": 1.2268,
+      "step": 740
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 8.159912376779846e-05,
+      "loss": 1.2339,
+      "step": 745
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 8.214676889375686e-05,
+      "loss": 1.1939,
+      "step": 750
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 8.269441401971523e-05,
+      "loss": 1.2931,
+      "step": 755
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 8.324205914567362e-05,
+      "loss": 1.2586,
+      "step": 760
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.166015625,
+      "learning_rate": 8.378970427163199e-05,
+      "loss": 1.2386,
+      "step": 765
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.150390625,
+      "learning_rate": 8.433734939759037e-05,
+      "loss": 1.1571,
+      "step": 770
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 8.488499452354874e-05,
+      "loss": 1.2219,
+      "step": 775
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 8.543263964950711e-05,
+      "loss": 1.238,
+      "step": 780
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1640625,
+      "learning_rate": 8.59802847754655e-05,
+      "loss": 1.2028,
+      "step": 785
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.162109375,
+      "learning_rate": 8.652792990142388e-05,
+      "loss": 1.1656,
+      "step": 790
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.15234375,
+      "learning_rate": 8.707557502738227e-05,
+      "loss": 1.2504,
+      "step": 795
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 8.762322015334064e-05,
+      "loss": 1.2424,
+      "step": 800
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 8.817086527929902e-05,
+      "loss": 1.2736,
+      "step": 805
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 8.871851040525739e-05,
+      "loss": 1.1999,
+      "step": 810
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.926615553121578e-05,
+      "loss": 1.2618,
+      "step": 815
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 8.981380065717415e-05,
+      "loss": 1.2386,
+      "step": 820
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.036144578313253e-05,
+      "loss": 1.3859,
+      "step": 825
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.15625,
+      "learning_rate": 9.090909090909092e-05,
+      "loss": 1.2344,
+      "step": 830
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 9.14567360350493e-05,
+      "loss": 1.2679,
+      "step": 835
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 9.200438116100767e-05,
+      "loss": 1.2153,
+      "step": 840
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 9.255202628696606e-05,
+      "loss": 1.3274,
+      "step": 845
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 9.309967141292443e-05,
+      "loss": 1.2584,
+      "step": 850
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.162109375,
+      "learning_rate": 9.364731653888281e-05,
+      "loss": 1.2726,
+      "step": 855
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 9.419496166484118e-05,
+      "loss": 1.2299,
+      "step": 860
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.15625,
+      "learning_rate": 9.474260679079957e-05,
+      "loss": 1.189,
+      "step": 865
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 9.529025191675794e-05,
+      "loss": 1.1814,
+      "step": 870
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1640625,
+      "learning_rate": 9.583789704271632e-05,
+      "loss": 1.2205,
+      "step": 875
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 9.638554216867471e-05,
+      "loss": 1.3385,
+      "step": 880
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 9.693318729463309e-05,
+      "loss": 1.2303,
+      "step": 885
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.748083242059146e-05,
+      "loss": 1.2501,
+      "step": 890
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.15625,
+      "learning_rate": 9.802847754654983e-05,
+      "loss": 1.2016,
+      "step": 895
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1484375,
+      "learning_rate": 9.857612267250822e-05,
+      "loss": 1.2084,
+      "step": 900
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 9.912376779846659e-05,
+      "loss": 1.2041,
+      "step": 905
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.967141292442497e-05,
+      "loss": 1.2082,
+      "step": 910
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00010021905805038336,
+      "loss": 1.2389,
+      "step": 915
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00010076670317634174,
+      "loss": 1.2297,
+      "step": 920
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00010131434830230011,
+      "loss": 1.1597,
+      "step": 925
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00010186199342825848,
+      "loss": 1.2344,
+      "step": 930
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010240963855421688,
+      "loss": 1.173,
+      "step": 935
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00010295728368017525,
+      "loss": 1.2133,
+      "step": 940
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00010350492880613362,
+      "loss": 1.2304,
+      "step": 945
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.189453125,
+      "learning_rate": 0.000104052573932092,
+      "loss": 1.2685,
+      "step": 950
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001046002190580504,
+      "loss": 1.2245,
+      "step": 955
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00010514786418400876,
+      "loss": 1.2806,
+      "step": 960
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00010569550930996715,
+      "loss": 1.2308,
+      "step": 965
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00010624315443592552,
+      "loss": 1.2814,
+      "step": 970
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00010679079956188392,
+      "loss": 1.2158,
+      "step": 975
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00010733844468784229,
+      "loss": 1.254,
+      "step": 980
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00010788608981380066,
+      "loss": 1.3504,
+      "step": 985
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00010843373493975903,
+      "loss": 1.2993,
+      "step": 990
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010898138006571743,
+      "loss": 1.2117,
+      "step": 995
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001095290251916758,
+      "loss": 1.2644,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00011007667031763418,
+      "loss": 1.1637,
+      "step": 1005
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00011062431544359256,
+      "loss": 1.2364,
+      "step": 1010
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 0.00011117196056955093,
+      "loss": 1.1681,
+      "step": 1015
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00011171960569550932,
+      "loss": 1.2646,
+      "step": 1020
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001122672508214677,
+      "loss": 1.2545,
+      "step": 1025
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00011281489594742607,
+      "loss": 1.3127,
+      "step": 1030
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00011336254107338444,
+      "loss": 1.1493,
+      "step": 1035
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011391018619934284,
+      "loss": 1.2099,
+      "step": 1040
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001144578313253012,
+      "loss": 1.1963,
+      "step": 1045
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00011500547645125959,
+      "loss": 1.2605,
+      "step": 1050
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00011555312157721796,
+      "loss": 1.1419,
+      "step": 1055
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00011610076670317636,
+      "loss": 1.1865,
+      "step": 1060
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00011664841182913473,
+      "loss": 1.2134,
+      "step": 1065
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001171960569550931,
+      "loss": 1.2468,
+      "step": 1070
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00011774370208105147,
+      "loss": 1.2814,
+      "step": 1075
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00011829134720700987,
+      "loss": 1.2728,
+      "step": 1080
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00011883899233296824,
+      "loss": 1.2128,
+      "step": 1085
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00011938663745892663,
+      "loss": 1.2427,
+      "step": 1090
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.000119934282584885,
+      "loss": 1.2083,
+      "step": 1095
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001204819277108434,
+      "loss": 1.2308,
+      "step": 1100
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00012102957283680177,
+      "loss": 1.2208,
+      "step": 1105
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00012157721796276014,
+      "loss": 1.2535,
+      "step": 1110
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001221248630887185,
+      "loss": 1.2217,
+      "step": 1115
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012267250821467688,
+      "loss": 1.2437,
+      "step": 1120
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00012322015334063528,
+      "loss": 1.2499,
+      "step": 1125
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00012376779846659365,
+      "loss": 1.1709,
+      "step": 1130
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00012431544359255202,
+      "loss": 1.219,
+      "step": 1135
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001248630887185104,
+      "loss": 1.1608,
+      "step": 1140
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001254107338444688,
+      "loss": 1.2019,
+      "step": 1145
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00012595837897042716,
+      "loss": 1.2408,
+      "step": 1150
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00012650602409638556,
+      "loss": 1.2266,
+      "step": 1155
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00012705366922234393,
+      "loss": 1.2009,
+      "step": 1160
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00012760131434830233,
+      "loss": 1.2754,
+      "step": 1165
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001281489594742607,
+      "loss": 1.28,
+      "step": 1170
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012869660460021907,
+      "loss": 1.2804,
+      "step": 1175
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00012924424972617744,
+      "loss": 1.2307,
+      "step": 1180
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00012979189485213584,
+      "loss": 1.2555,
+      "step": 1185
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001303395399780942,
+      "loss": 1.2674,
+      "step": 1190
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00013088718510405258,
+      "loss": 1.2061,
+      "step": 1195
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00013143483023001095,
+      "loss": 1.2576,
+      "step": 1200
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00013198247535596935,
+      "loss": 1.2024,
+      "step": 1205
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00013253012048192772,
+      "loss": 1.2608,
+      "step": 1210
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001330777656078861,
+      "loss": 1.1914,
+      "step": 1215
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00013362541073384446,
+      "loss": 1.2109,
+      "step": 1220
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00013417305585980283,
+      "loss": 1.1982,
+      "step": 1225
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00013472070098576123,
+      "loss": 1.3028,
+      "step": 1230
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001352683461117196,
+      "loss": 1.1142,
+      "step": 1235
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.000135815991237678,
+      "loss": 1.2338,
+      "step": 1240
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00013636363636363637,
+      "loss": 1.2409,
+      "step": 1245
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00013691128148959477,
+      "loss": 1.2334,
+      "step": 1250
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00013745892661555314,
+      "loss": 1.2972,
+      "step": 1255
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001380065717415115,
+      "loss": 1.2492,
+      "step": 1260
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00013855421686746988,
+      "loss": 1.2353,
+      "step": 1265
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00013910186199342828,
+      "loss": 1.1976,
+      "step": 1270
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00013964950711938665,
+      "loss": 1.1801,
+      "step": 1275
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00014019715224534502,
+      "loss": 1.246,
+      "step": 1280
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001407447973713034,
+      "loss": 1.1764,
+      "step": 1285
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0001412924424972618,
+      "loss": 1.2766,
+      "step": 1290
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00014184008762322016,
+      "loss": 1.3015,
+      "step": 1295
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00014238773274917853,
+      "loss": 1.2957,
+      "step": 1300
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001429353778751369,
+      "loss": 1.2891,
+      "step": 1305
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001434830230010953,
+      "loss": 1.2687,
+      "step": 1310
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00014403066812705367,
+      "loss": 1.1833,
+      "step": 1315
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00014457831325301204,
+      "loss": 1.2054,
+      "step": 1320
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00014512595837897044,
+      "loss": 1.2631,
+      "step": 1325
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001456736035049288,
+      "loss": 1.1603,
+      "step": 1330
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001462212486308872,
+      "loss": 1.2929,
+      "step": 1335
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00014676889375684558,
+      "loss": 1.1819,
+      "step": 1340
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00014731653888280395,
+      "loss": 1.1963,
+      "step": 1345
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00014786418400876232,
+      "loss": 1.2173,
+      "step": 1350
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00014841182913472072,
+      "loss": 1.2552,
+      "step": 1355
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001489594742606791,
+      "loss": 1.1746,
+      "step": 1360
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00014950711938663746,
+      "loss": 1.2428,
+      "step": 1365
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00015005476451259583,
+      "loss": 1.2926,
+      "step": 1370
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00015060240963855423,
+      "loss": 1.2575,
+      "step": 1375
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001511500547645126,
+      "loss": 1.1596,
+      "step": 1380
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00015169769989047097,
+      "loss": 1.1738,
+      "step": 1385
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00015224534501642934,
+      "loss": 1.2195,
+      "step": 1390
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00015279299014238774,
+      "loss": 1.2544,
+      "step": 1395
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001533406352683461,
+      "loss": 1.1298,
+      "step": 1400
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00015388828039430448,
+      "loss": 1.1252,
+      "step": 1405
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00015443592552026288,
+      "loss": 1.2897,
+      "step": 1410
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00015498357064622128,
+      "loss": 1.2853,
+      "step": 1415
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00015553121577217965,
+      "loss": 1.2858,
+      "step": 1420
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00015607886089813802,
+      "loss": 1.1699,
+      "step": 1425
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001566265060240964,
+      "loss": 1.2723,
+      "step": 1430
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001571741511500548,
+      "loss": 1.2586,
+      "step": 1435
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00015772179627601316,
+      "loss": 1.2875,
+      "step": 1440
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00015826944140197153,
+      "loss": 1.2107,
+      "step": 1445
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001588170865279299,
+      "loss": 1.2168,
+      "step": 1450
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00015936473165388827,
+      "loss": 1.2014,
+      "step": 1455
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00015991237677984667,
+      "loss": 1.1969,
+      "step": 1460
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00016046002190580504,
+      "loss": 1.2384,
+      "step": 1465
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001610076670317634,
+      "loss": 1.2022,
+      "step": 1470
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00016155531215772178,
+      "loss": 1.2074,
+      "step": 1475
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00016210295728368018,
+      "loss": 1.2065,
+      "step": 1480
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00016265060240963855,
+      "loss": 1.2527,
+      "step": 1485
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00016319824753559692,
+      "loss": 1.1705,
+      "step": 1490
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00016374589266155532,
+      "loss": 1.2375,
+      "step": 1495
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00016429353778751372,
+      "loss": 1.2421,
+      "step": 1500
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001648411829134721,
+      "loss": 1.2498,
+      "step": 1505
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00016538882803943046,
+      "loss": 1.2472,
+      "step": 1510
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00016593647316538883,
+      "loss": 1.2729,
+      "step": 1515
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00016648411829134723,
+      "loss": 1.1838,
+      "step": 1520
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001670317634173056,
+      "loss": 1.2131,
+      "step": 1525
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00016757940854326397,
+      "loss": 1.2797,
+      "step": 1530
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00016812705366922234,
+      "loss": 1.14,
+      "step": 1535
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00016867469879518074,
+      "loss": 1.1353,
+      "step": 1540
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001692223439211391,
+      "loss": 1.2047,
+      "step": 1545
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00016976998904709748,
+      "loss": 1.2367,
+      "step": 1550
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00017031763417305585,
+      "loss": 1.2328,
+      "step": 1555
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00017086527929901423,
+      "loss": 1.2321,
+      "step": 1560
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00017141292442497262,
+      "loss": 1.2252,
+      "step": 1565
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.000171960569550931,
+      "loss": 1.2612,
+      "step": 1570
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017250821467688937,
+      "loss": 1.2438,
+      "step": 1575
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00017305585980284776,
+      "loss": 1.2371,
+      "step": 1580
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00017360350492880616,
+      "loss": 1.1611,
+      "step": 1585
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017415115005476453,
+      "loss": 1.2399,
+      "step": 1590
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0001746987951807229,
+      "loss": 1.1899,
+      "step": 1595
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00017524644030668127,
+      "loss": 1.2359,
+      "step": 1600
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00017579408543263967,
+      "loss": 1.2184,
+      "step": 1605
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00017634173055859804,
+      "loss": 1.2627,
+      "step": 1610
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00017688937568455641,
+      "loss": 1.2724,
+      "step": 1615
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00017743702081051479,
+      "loss": 1.2299,
+      "step": 1620
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00017798466593647318,
+      "loss": 1.227,
+      "step": 1625
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00017853231106243155,
+      "loss": 1.2011,
+      "step": 1630
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00017907995618838993,
+      "loss": 1.1601,
+      "step": 1635
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001796276013143483,
+      "loss": 1.2511,
+      "step": 1640
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001801752464403067,
+      "loss": 1.209,
+      "step": 1645
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018072289156626507,
+      "loss": 1.2136,
+      "step": 1650
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00018127053669222344,
+      "loss": 1.2127,
+      "step": 1655
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00018181818181818183,
+      "loss": 1.1698,
+      "step": 1660
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0001823658269441402,
+      "loss": 1.2385,
+      "step": 1665
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001829134720700986,
+      "loss": 1.2362,
+      "step": 1670
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00018346111719605697,
+      "loss": 1.2404,
+      "step": 1675
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00018400876232201535,
+      "loss": 1.2084,
+      "step": 1680
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00018455640744797372,
+      "loss": 1.2191,
+      "step": 1685
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018510405257393211,
+      "loss": 1.2111,
+      "step": 1690
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00018565169769989049,
+      "loss": 1.1864,
+      "step": 1695
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00018619934282584886,
+      "loss": 1.2756,
+      "step": 1700
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00018674698795180723,
+      "loss": 1.1807,
+      "step": 1705
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018729463307776562,
+      "loss": 1.1757,
+      "step": 1710
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.000187842278203724,
+      "loss": 1.1995,
+      "step": 1715
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018838992332968237,
+      "loss": 1.2545,
+      "step": 1720
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00018893756845564074,
+      "loss": 1.2862,
+      "step": 1725
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018948521358159914,
+      "loss": 1.2356,
+      "step": 1730
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001900328587075575,
+      "loss": 1.2461,
+      "step": 1735
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019058050383351588,
+      "loss": 1.2673,
+      "step": 1740
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019112814895947428,
+      "loss": 1.2255,
+      "step": 1745
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.00019167579408543265,
+      "loss": 1.1952,
+      "step": 1750
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019222343921139104,
+      "loss": 1.1944,
+      "step": 1755
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019277108433734942,
+      "loss": 1.2916,
+      "step": 1760
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001933187294633078,
+      "loss": 1.216,
+      "step": 1765
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019386637458926618,
+      "loss": 1.2824,
+      "step": 1770
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019441401971522456,
+      "loss": 1.1353,
+      "step": 1775
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00019496166484118293,
+      "loss": 1.1201,
+      "step": 1780
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001955093099671413,
+      "loss": 1.2502,
+      "step": 1785
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00019605695509309967,
+      "loss": 1.1795,
+      "step": 1790
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019660460021905807,
+      "loss": 1.2021,
+      "step": 1795
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019715224534501644,
+      "loss": 1.1568,
+      "step": 1800
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0001976998904709748,
+      "loss": 1.2459,
+      "step": 1805
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019824753559693318,
+      "loss": 1.2203,
+      "step": 1810
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019879518072289158,
+      "loss": 1.2607,
+      "step": 1815
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019934282584884995,
+      "loss": 1.24,
+      "step": 1820
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019989047097480832,
+      "loss": 1.2106,
+      "step": 1825
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00019999997075432027,
+      "loss": 1.1284,
+      "step": 1830
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001999998519437756,
+      "loss": 1.2807,
+      "step": 1835
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0001999996417406196,
+      "loss": 1.185,
+      "step": 1840
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001999993401450443,
+      "loss": 1.2288,
+      "step": 1845
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019999894715732542,
+      "loss": 1.1901,
+      "step": 1850
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.000199998462777822,
+      "loss": 1.2768,
+      "step": 1855
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019999788700697684,
+      "loss": 1.2265,
+      "step": 1860
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001999972198453161,
+      "loss": 1.1219,
+      "step": 1865
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0001999964612934495,
+      "loss": 1.2869,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019999561135207038,
+      "loss": 1.2066,
+      "step": 1875
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019999467002195545,
+      "loss": 1.2709,
+      "step": 1880
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019999363730396503,
+      "loss": 1.1617,
+      "step": 1885
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.000199992513199043,
+      "loss": 1.2292,
+      "step": 1890
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019999129770821662,
+      "loss": 1.1603,
+      "step": 1895
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019998999083259685,
+      "loss": 1.2836,
+      "step": 1900
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019998859257337802,
+      "loss": 1.209,
+      "step": 1905
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00019998710293183804,
+      "loss": 1.211,
+      "step": 1910
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019998552190933838,
+      "loss": 1.2048,
+      "step": 1915
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019998384950732393,
+      "loss": 1.1859,
+      "step": 1920
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00019998208572732321,
+      "loss": 1.2177,
+      "step": 1925
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019998023057094812,
+      "loss": 1.2393,
+      "step": 1930
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019997828403989416,
+      "loss": 1.1509,
+      "step": 1935
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019997624613594034,
+      "loss": 1.1318,
+      "step": 1940
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019997411686094912,
+      "loss": 1.1796,
+      "step": 1945
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001999718962168665,
+      "loss": 1.168,
+      "step": 1950
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019996958420572203,
+      "loss": 1.2048,
+      "step": 1955
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0001999671808296287,
+      "loss": 1.1888,
+      "step": 1960
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.000199964686090783,
+      "loss": 1.2198,
+      "step": 1965
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019996209999146499,
+      "loss": 1.2259,
+      "step": 1970
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019995942253403813,
+      "loss": 1.167,
+      "step": 1975
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019995665372094944,
+      "loss": 1.2167,
+      "step": 1980
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019995379355472942,
+      "loss": 1.1921,
+      "step": 1985
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019995084203799202,
+      "loss": 1.1466,
+      "step": 1990
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019994779917343475,
+      "loss": 1.255,
+      "step": 1995
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019994466496383857,
+      "loss": 1.2633,
+      "step": 2000
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001999414394120679,
+      "loss": 1.1621,
+      "step": 2005
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019993812252107066,
+      "loss": 1.2009,
+      "step": 2010
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019993471429387828,
+      "loss": 1.221,
+      "step": 2015
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0001999312147336056,
+      "loss": 1.253,
+      "step": 2020
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019992762384345096,
+      "loss": 1.188,
+      "step": 2025
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001999239416266962,
+      "loss": 1.2327,
+      "step": 2030
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00019992016808670658,
+      "loss": 1.2273,
+      "step": 2035
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019991630322693086,
+      "loss": 1.2041,
+      "step": 2040
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019991234705090118,
+      "loss": 1.2807,
+      "step": 2045
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019990829956223328,
+      "loss": 1.2737,
+      "step": 2050
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00019990416076462624,
+      "loss": 1.1945,
+      "step": 2055
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019989993066186262,
+      "loss": 1.2268,
+      "step": 2060
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001998956092578084,
+      "loss": 1.2135,
+      "step": 2065
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019989119655641307,
+      "loss": 1.2546,
+      "step": 2070
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001998866925617095,
+      "loss": 1.1937,
+      "step": 2075
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019988209727781403,
+      "loss": 1.2649,
+      "step": 2080
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019987741070892642,
+      "loss": 1.1636,
+      "step": 2085
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001998726328593298,
+      "loss": 1.2095,
+      "step": 2090
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019986776373339082,
+      "loss": 1.2588,
+      "step": 2095
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00019986280333555954,
+      "loss": 1.1869,
+      "step": 2100
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019985775167036932,
+      "loss": 1.1601,
+      "step": 2105
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0001998526087424371,
+      "loss": 1.1809,
+      "step": 2110
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.0001998473745564631,
+      "loss": 1.168,
+      "step": 2115
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.000199842049117231,
+      "loss": 1.1992,
+      "step": 2120
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019983663242960784,
+      "loss": 1.2649,
+      "step": 2125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001998311244985441,
+      "loss": 1.1687,
+      "step": 2130
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019982552532907364,
+      "loss": 1.1899,
+      "step": 2135
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00019981983492631367,
+      "loss": 1.1579,
+      "step": 2140
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001998140532954648,
+      "loss": 1.132,
+      "step": 2145
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019980818044181106,
+      "loss": 1.1908,
+      "step": 2150
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0001998022163707198,
+      "loss": 1.2316,
+      "step": 2155
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00019979616108764173,
+      "loss": 1.1697,
+      "step": 2160
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019979001459811095,
+      "loss": 1.251,
+      "step": 2165
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001997837769077449,
+      "loss": 1.2057,
+      "step": 2170
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019977744802224438,
+      "loss": 1.183,
+      "step": 2175
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019977102794739354,
+      "loss": 1.238,
+      "step": 2180
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019976451668905987,
+      "loss": 1.1808,
+      "step": 2185
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 0.00019975791425319415,
+      "loss": 1.2381,
+      "step": 2190
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019975122064583056,
+      "loss": 1.2191,
+      "step": 2195
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019974443587308656,
+      "loss": 1.2661,
+      "step": 2200
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00019973755994116292,
+      "loss": 1.201,
+      "step": 2205
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0001997305928563438,
+      "loss": 1.1525,
+      "step": 2210
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019972353462499657,
+      "loss": 1.2477,
+      "step": 2215
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019971638525357195,
+      "loss": 1.2477,
+      "step": 2220
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019970914474860393,
+      "loss": 1.2408,
+      "step": 2225
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019970181311670983,
+      "loss": 1.2717,
+      "step": 2230
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019969439036459022,
+      "loss": 1.2187,
+      "step": 2235
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019968687649902898,
+      "loss": 1.1677,
+      "step": 2240
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001996792715268932,
+      "loss": 1.2571,
+      "step": 2245
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00019967157545513326,
+      "loss": 1.2498,
+      "step": 2250
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001996637882907829,
+      "loss": 1.1404,
+      "step": 2255
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0001996559100409589,
+      "loss": 1.2308,
+      "step": 2260
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019964794071286154,
+      "loss": 1.1871,
+      "step": 2265
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019963988031377407,
+      "loss": 1.1729,
+      "step": 2270
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019963172885106322,
+      "loss": 1.2352,
+      "step": 2275
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019962348633217878,
+      "loss": 1.2005,
+      "step": 2280
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019961515276465384,
+      "loss": 1.2379,
+      "step": 2285
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019960672815610465,
+      "loss": 1.1967,
+      "step": 2290
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019959821251423074,
+      "loss": 1.215,
+      "step": 2295
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019958960584681474,
+      "loss": 1.2723,
+      "step": 2300
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00019958090816172254,
+      "loss": 1.318,
+      "step": 2305
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019957211946690321,
+      "loss": 1.1797,
+      "step": 2310
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019956323977038894,
+      "loss": 1.1386,
+      "step": 2315
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019955426908029514,
+      "loss": 1.1652,
+      "step": 2320
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0001995452074048204,
+      "loss": 1.2241,
+      "step": 2325
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001995360547522464,
+      "loss": 1.1903,
+      "step": 2330
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.000199526811130938,
+      "loss": 1.1414,
+      "step": 2335
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019951747654934322,
+      "loss": 1.2083,
+      "step": 2340
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0001995080510159932,
+      "loss": 1.1457,
+      "step": 2345
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019949853453950215,
+      "loss": 1.1902,
+      "step": 2350
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001994889271285675,
+      "loss": 1.2253,
+      "step": 2355
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019947922879196966,
+      "loss": 1.1969,
+      "step": 2360
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019946943953857226,
+      "loss": 1.1755,
+      "step": 2365
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019945955937732194,
+      "loss": 1.2198,
+      "step": 2370
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00019944958831724844,
+      "loss": 1.2239,
+      "step": 2375
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001994395263674646,
+      "loss": 1.1852,
+      "step": 2380
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019942937353716636,
+      "loss": 1.2415,
+      "step": 2385
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0001994191298356326,
+      "loss": 1.1977,
+      "step": 2390
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019940879527222536,
+      "loss": 1.1554,
+      "step": 2395
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019939836985638965,
+      "loss": 1.2109,
+      "step": 2400
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001993878535976536,
+      "loss": 1.1984,
+      "step": 2405
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019937724650562818,
+      "loss": 1.2217,
+      "step": 2410
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019936654859000768,
+      "loss": 1.1562,
+      "step": 2415
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019935575986056907,
+      "loss": 1.2268,
+      "step": 2420
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019934488032717255,
+      "loss": 1.1742,
+      "step": 2425
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019933390999976122,
+      "loss": 1.1728,
+      "step": 2430
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019932284888836112,
+      "loss": 1.191,
+      "step": 2435
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001993116970030813,
+      "loss": 1.2015,
+      "step": 2440
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001993004543541139,
+      "loss": 1.1842,
+      "step": 2445
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00019928912095173372,
+      "loss": 1.2625,
+      "step": 2450
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019927769680629873,
+      "loss": 1.2149,
+      "step": 2455
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001992661819282498,
+      "loss": 1.2789,
+      "step": 2460
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019925457632811064,
+      "loss": 1.1799,
+      "step": 2465
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019924288001648799,
+      "loss": 1.1884,
+      "step": 2470
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00019923109300407137,
+      "loss": 1.2513,
+      "step": 2475
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001992192153016333,
+      "loss": 1.1907,
+      "step": 2480
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001992072469200291,
+      "loss": 1.1897,
+      "step": 2485
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019919518787019704,
+      "loss": 1.2474,
+      "step": 2490
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001991830381631582,
+      "loss": 1.1519,
+      "step": 2495
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019917079781001654,
+      "loss": 1.2137,
+      "step": 2500
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019915846682195884,
+      "loss": 1.1884,
+      "step": 2505
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001991460452102547,
+      "loss": 1.1588,
+      "step": 2510
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019913353298625663,
+      "loss": 1.2153,
+      "step": 2515
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019912093016139987,
+      "loss": 1.1183,
+      "step": 2520
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0001991082367472025,
+      "loss": 1.129,
+      "step": 2525
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019909545275526534,
+      "loss": 1.1752,
+      "step": 2530
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001990825781972721,
+      "loss": 1.1123,
+      "step": 2535
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019906961308498912,
+      "loss": 1.1872,
+      "step": 2540
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001990565574302656,
+      "loss": 1.2054,
+      "step": 2545
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001990434112450335,
+      "loss": 1.1857,
+      "step": 2550
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001990301745413074,
+      "loss": 1.1533,
+      "step": 2555
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019901684733118478,
+      "loss": 1.1818,
+      "step": 2560
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019900342962684566,
+      "loss": 1.2414,
+      "step": 2565
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019898992144055293,
+      "loss": 1.173,
+      "step": 2570
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.000198976322784652,
+      "loss": 1.1529,
+      "step": 2575
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019896263367157112,
+      "loss": 1.2378,
+      "step": 2580
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019894885411382113,
+      "loss": 1.1579,
+      "step": 2585
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019893498412399554,
+      "loss": 1.3038,
+      "step": 2590
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019892102371477056,
+      "loss": 1.2808,
+      "step": 2595
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001989069728989049,
+      "loss": 1.1197,
+      "step": 2600
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019889283168924004,
+      "loss": 1.2699,
+      "step": 2605
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019887860009870002,
+      "loss": 1.1598,
+      "step": 2610
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0001988642781402915,
+      "loss": 1.2068,
+      "step": 2615
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019884986582710366,
+      "loss": 1.125,
+      "step": 2620
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019883536317230838,
+      "loss": 1.2651,
+      "step": 2625
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019882077018915998,
+      "loss": 1.1742,
+      "step": 2630
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019880608689099537,
+      "loss": 1.1599,
+      "step": 2635
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0001987913132912341,
+      "loss": 1.1619,
+      "step": 2640
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001987764494033781,
+      "loss": 1.2287,
+      "step": 2645
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001987614952410119,
+      "loss": 1.1133,
+      "step": 2650
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00019874645081780252,
+      "loss": 1.2409,
+      "step": 2655
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019873131614749952,
+      "loss": 1.1873,
+      "step": 2660
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0001987160912439348,
+      "loss": 1.1572,
+      "step": 2665
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019870077612102287,
+      "loss": 1.2298,
+      "step": 2670
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019868537079276063,
+      "loss": 1.1669,
+      "step": 2675
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00019866987527322746,
+      "loss": 1.1517,
+      "step": 2680
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001986542895765851,
+      "loss": 1.2001,
+      "step": 2685
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019863861371707779,
+      "loss": 1.2124,
+      "step": 2690
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019862284770903212,
+      "loss": 1.2066,
+      "step": 2695
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.000198606991566857,
+      "loss": 1.1659,
+      "step": 2700
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019859104530504392,
+      "loss": 1.1927,
+      "step": 2705
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019857500893816652,
+      "loss": 1.1639,
+      "step": 2710
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019855888248088087,
+      "loss": 1.1809,
+      "step": 2715
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019854266594792548,
+      "loss": 1.1528,
+      "step": 2720
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019852635935412095,
+      "loss": 1.2252,
+      "step": 2725
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019850996271437044,
+      "loss": 1.1844,
+      "step": 2730
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001984934760436592,
+      "loss": 1.2045,
+      "step": 2735
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019847689935705492,
+      "loss": 1.2144,
+      "step": 2740
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019846023266970747,
+      "loss": 1.2032,
+      "step": 2745
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019844347599684895,
+      "loss": 1.224,
+      "step": 2750
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0001984266293537938,
+      "loss": 1.1924,
+      "step": 2755
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019840969275593858,
+      "loss": 1.1363,
+      "step": 2760
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019839266621876216,
+      "loss": 1.2361,
+      "step": 2765
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00019837554975782554,
+      "loss": 1.1856,
+      "step": 2770
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001983583433887719,
+      "loss": 1.1942,
+      "step": 2775
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0001983410471273267,
+      "loss": 1.1392,
+      "step": 2780
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019832366098929735,
+      "loss": 1.1921,
+      "step": 2785
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019830618499057357,
+      "loss": 1.2808,
+      "step": 2790
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019828861914712715,
+      "loss": 1.1448,
+      "step": 2795
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00019827096347501205,
+      "loss": 1.2194,
+      "step": 2800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001982532179903642,
+      "loss": 1.1556,
+      "step": 2805
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019823538270940171,
+      "loss": 1.193,
+      "step": 2810
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019821745764842476,
+      "loss": 1.1494,
+      "step": 2815
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019819944282381554,
+      "loss": 1.1407,
+      "step": 2820
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.0001981813382520383,
+      "loss": 1.248,
+      "step": 2825
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019816314394963924,
+      "loss": 1.2004,
+      "step": 2830
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019814485993324671,
+      "loss": 1.2354,
+      "step": 2835
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019812648621957095,
+      "loss": 1.1944,
+      "step": 2840
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001981080228254042,
+      "loss": 1.2029,
+      "step": 2845
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019808946976762067,
+      "loss": 1.2045,
+      "step": 2850
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001980708270631765,
+      "loss": 1.2151,
+      "step": 2855
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001980520947291098,
+      "loss": 1.1599,
+      "step": 2860
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001980332727825405,
+      "loss": 1.2136,
+      "step": 2865
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00019801436124067053,
+      "loss": 1.1651,
+      "step": 2870
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001979953601207837,
+      "loss": 1.2054,
+      "step": 2875
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0001979762694402456,
+      "loss": 1.2132,
+      "step": 2880
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019795708921650378,
+      "loss": 1.1789,
+      "step": 2885
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001979378194670875,
+      "loss": 1.1421,
+      "step": 2890
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.000197918460209608,
+      "loss": 1.2263,
+      "step": 2895
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001978990114617582,
+      "loss": 1.208,
+      "step": 2900
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019787947324131282,
+      "loss": 1.1271,
+      "step": 2905
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019785984556612836,
+      "loss": 1.2911,
+      "step": 2910
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001978401284541432,
+      "loss": 1.2554,
+      "step": 2915
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019782032192337725,
+      "loss": 1.1166,
+      "step": 2920
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019780042599193225,
+      "loss": 1.2657,
+      "step": 2925
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019778044067799168,
+      "loss": 1.2139,
+      "step": 2930
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019776036599982065,
+      "loss": 1.2246,
+      "step": 2935
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019774020197576596,
+      "loss": 1.1838,
+      "step": 2940
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019771994862425606,
+      "loss": 1.1485,
+      "step": 2945
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019769960596380105,
+      "loss": 1.198,
+      "step": 2950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019767917401299263,
+      "loss": 1.1564,
+      "step": 2955
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019765865279050413,
+      "loss": 1.208,
+      "step": 2960
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00019763804231509046,
+      "loss": 1.2053,
+      "step": 2965
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001976173426055881,
+      "loss": 1.2324,
+      "step": 2970
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001975965536809151,
+      "loss": 1.1634,
+      "step": 2975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.000197575675560071,
+      "loss": 1.1063,
+      "step": 2980
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019755470826213694,
+      "loss": 1.1857,
+      "step": 2985
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019753365180627544,
+      "loss": 1.2372,
+      "step": 2990
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001975125062117306,
+      "loss": 1.2264,
+      "step": 2995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.000197491271497828,
+      "loss": 1.1289,
+      "step": 3000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019746994768397456,
+      "loss": 1.2187,
+      "step": 3005
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019744853478965874,
+      "loss": 1.2276,
+      "step": 3010
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019742703283445035,
+      "loss": 1.1815,
+      "step": 3015
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019740544183800065,
+      "loss": 1.182,
+      "step": 3020
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019738376182004222,
+      "loss": 1.194,
+      "step": 3025
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019736199280038906,
+      "loss": 1.2095,
+      "step": 3030
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001973401347989364,
+      "loss": 1.185,
+      "step": 3035
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019731818783566094,
+      "loss": 1.1828,
+      "step": 3040
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019729615193062058,
+      "loss": 1.1592,
+      "step": 3045
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019727402710395455,
+      "loss": 1.218,
+      "step": 3050
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019725181337588331,
+      "loss": 1.1935,
+      "step": 3055
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019722951076670862,
+      "loss": 1.2958,
+      "step": 3060
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019720711929681347,
+      "loss": 1.2223,
+      "step": 3065
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019718463898666202,
+      "loss": 1.2116,
+      "step": 3070
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00019716206985679962,
+      "loss": 1.2521,
+      "step": 3075
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019713941192785284,
+      "loss": 1.277,
+      "step": 3080
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019711666522052935,
+      "loss": 1.1773,
+      "step": 3085
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019709382975561805,
+      "loss": 1.1902,
+      "step": 3090
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019707090555398887,
+      "loss": 1.2634,
+      "step": 3095
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019704789263659287,
+      "loss": 1.2584,
+      "step": 3100
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019702479102446218,
+      "loss": 1.1253,
+      "step": 3105
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019700160073870998,
+      "loss": 1.2754,
+      "step": 3110
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019697832180053056,
+      "loss": 1.176,
+      "step": 3115
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019695495423119913,
+      "loss": 1.1775,
+      "step": 3120
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019693149805207196,
+      "loss": 1.2001,
+      "step": 3125
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019690795328458632,
+      "loss": 1.1762,
+      "step": 3130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019688431995026044,
+      "loss": 1.2328,
+      "step": 3135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001968605980706934,
+      "loss": 1.2827,
+      "step": 3140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019683678766756538,
+      "loss": 1.1568,
+      "step": 3145
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019681288876263723,
+      "loss": 1.1664,
+      "step": 3150
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019678890137775094,
+      "loss": 1.22,
+      "step": 3155
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019676482553482915,
+      "loss": 1.2122,
+      "step": 3160
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019674066125587549,
+      "loss": 1.2239,
+      "step": 3165
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019671640856297435,
+      "loss": 1.1882,
+      "step": 3170
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019669206747829093,
+      "loss": 1.2303,
+      "step": 3175
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0001966676380240712,
+      "loss": 1.246,
+      "step": 3180
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019664312022264188,
+      "loss": 1.2049,
+      "step": 3185
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00019661851409641054,
+      "loss": 1.2294,
+      "step": 3190
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019659381966786535,
+      "loss": 1.1979,
+      "step": 3195
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 0.00019656903695957522,
+      "loss": 1.1838,
+      "step": 3200
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019654416599418977,
+      "loss": 1.1748,
+      "step": 3205
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019651920679443925,
+      "loss": 1.204,
+      "step": 3210
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019649415938313453,
+      "loss": 1.1967,
+      "step": 3215
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019646902378316717,
+      "loss": 1.1838,
+      "step": 3220
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019644380001750927,
+      "loss": 1.1577,
+      "step": 3225
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019641848810921352,
+      "loss": 1.1423,
+      "step": 3230
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019639308808141318,
+      "loss": 1.1374,
+      "step": 3235
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.000196367599957322,
+      "loss": 1.1429,
+      "step": 3240
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0001963420237602343,
+      "loss": 1.2008,
+      "step": 3245
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019631635951352486,
+      "loss": 1.2106,
+      "step": 3250
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00019629060724064897,
+      "loss": 1.2584,
+      "step": 3255
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00019626476696514228,
+      "loss": 1.1772,
+      "step": 3260
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.000196238838710621,
+      "loss": 1.1017,
+      "step": 3265
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019621282250078165,
+      "loss": 1.2701,
+      "step": 3270
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001961867183594011,
+      "loss": 1.1424,
+      "step": 3275
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019616052631033672,
+      "loss": 1.1661,
+      "step": 3280
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019613424637752612,
+      "loss": 1.1719,
+      "step": 3285
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0001961078785849872,
+      "loss": 1.1828,
+      "step": 3290
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019608142295681827,
+      "loss": 1.1878,
+      "step": 3295
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019605487951719787,
+      "loss": 1.1571,
+      "step": 3300
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019602824829038473,
+      "loss": 1.1789,
+      "step": 3305
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019600152930071788,
+      "loss": 1.1424,
+      "step": 3310
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001959747225726165,
+      "loss": 1.1929,
+      "step": 3315
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001959478281305801,
+      "loss": 1.2009,
+      "step": 3320
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019592084599918814,
+      "loss": 1.1991,
+      "step": 3325
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019589377620310035,
+      "loss": 1.141,
+      "step": 3330
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001958666187670566,
+      "loss": 1.1432,
+      "step": 3335
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019583937371587683,
+      "loss": 1.1486,
+      "step": 3340
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019581204107446096,
+      "loss": 1.2263,
+      "step": 3345
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001957846208677891,
+      "loss": 1.2542,
+      "step": 3350
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019575711312092137,
+      "loss": 1.1678,
+      "step": 3355
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019572951785899775,
+      "loss": 1.1939,
+      "step": 3360
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019570183510723838,
+      "loss": 1.1907,
+      "step": 3365
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019567406489094327,
+      "loss": 1.1974,
+      "step": 3370
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001956462072354924,
+      "loss": 1.2286,
+      "step": 3375
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001956182621663456,
+      "loss": 1.1975,
+      "step": 3380
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019559022970904268,
+      "loss": 1.1666,
+      "step": 3385
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019556210988920327,
+      "loss": 1.2401,
+      "step": 3390
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019553390273252682,
+      "loss": 1.2396,
+      "step": 3395
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00019550560826479268,
+      "loss": 1.2139,
+      "step": 3400
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 0.00019547722651185986,
+      "loss": 1.1671,
+      "step": 3405
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019544875749966727,
+      "loss": 1.2247,
+      "step": 3410
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019542020125423352,
+      "loss": 1.1916,
+      "step": 3415
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019539155780165697,
+      "loss": 1.1933,
+      "step": 3420
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.0001953628271681156,
+      "loss": 1.1717,
+      "step": 3425
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019533400937986716,
+      "loss": 1.1789,
+      "step": 3430
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019530510446324903,
+      "loss": 1.2876,
+      "step": 3435
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001952761124446782,
+      "loss": 1.242,
+      "step": 3440
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00019524703335065124,
+      "loss": 1.1854,
+      "step": 3445
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001952178672077444,
+      "loss": 1.278,
+      "step": 3450
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019518861404261336,
+      "loss": 1.2237,
+      "step": 3455
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019515927388199337,
+      "loss": 1.1598,
+      "step": 3460
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019512984675269928,
+      "loss": 1.2386,
+      "step": 3465
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019510033268162533,
+      "loss": 1.188,
+      "step": 3470
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019507073169574519,
+      "loss": 1.1833,
+      "step": 3475
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001950410438221121,
+      "loss": 1.1666,
+      "step": 3480
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019501126908785853,
+      "loss": 1.1627,
+      "step": 3485
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001949814075201965,
+      "loss": 1.2412,
+      "step": 3490
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001949514591464173,
+      "loss": 1.1465,
+      "step": 3495
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019492142399389154,
+      "loss": 1.1897,
+      "step": 3500
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.0001948913020900692,
+      "loss": 1.2094,
+      "step": 3505
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019486109346247954,
+      "loss": 1.1988,
+      "step": 3510
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.000194830798138731,
+      "loss": 1.1563,
+      "step": 3515
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001948004161465113,
+      "loss": 1.1494,
+      "step": 3520
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00019476994751358744,
+      "loss": 1.1626,
+      "step": 3525
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019473939226780548,
+      "loss": 1.1899,
+      "step": 3530
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00019470875043709076,
+      "loss": 1.1181,
+      "step": 3535
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019467802204944766,
+      "loss": 1.2443,
+      "step": 3540
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001946472071329597,
+      "loss": 1.1897,
+      "step": 3545
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019461630571578943,
+      "loss": 1.1916,
+      "step": 3550
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001945853178261786,
+      "loss": 1.2369,
+      "step": 3555
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001945542434924478,
+      "loss": 1.1876,
+      "step": 3560
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019452308274299677,
+      "loss": 1.2302,
+      "step": 3565
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019449183560630418,
+      "loss": 1.1644,
+      "step": 3570
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001944605021109276,
+      "loss": 1.2396,
+      "step": 3575
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019442908228550364,
+      "loss": 1.2611,
+      "step": 3580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019439757615874768,
+      "loss": 1.1609,
+      "step": 3585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019436598375945405,
+      "loss": 1.2111,
+      "step": 3590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00019433430511649593,
+      "loss": 1.1857,
+      "step": 3595
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019430254025882532,
+      "loss": 1.1455,
+      "step": 3600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0001942706892154729,
+      "loss": 1.1492,
+      "step": 3605
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019423875201554834,
+      "loss": 1.1672,
+      "step": 3610
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019420672868823984,
+      "loss": 1.2263,
+      "step": 3615
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019417461926281442,
+      "loss": 1.1972,
+      "step": 3620
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019414242376861778,
+      "loss": 1.2509,
+      "step": 3625
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019411014223507425,
+      "loss": 1.1969,
+      "step": 3630
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 0.00019407777469168682,
+      "loss": 1.1543,
+      "step": 3635
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019404532116803707,
+      "loss": 1.2066,
+      "step": 3640
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019401278169378517,
+      "loss": 1.0788,
+      "step": 3645
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019398015629866984,
+      "loss": 1.1937,
+      "step": 3650
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019394744501250832,
+      "loss": 1.1784,
+      "step": 3655
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019391464786519636,
+      "loss": 1.144,
+      "step": 3660
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019388176488670814,
+      "loss": 1.1592,
+      "step": 3665
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001938487961070964,
+      "loss": 1.2506,
+      "step": 3670
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019381574155649212,
+      "loss": 1.2269,
+      "step": 3675
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019378260126510483,
+      "loss": 1.2301,
+      "step": 3680
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001937493752632223,
+      "loss": 1.1464,
+      "step": 3685
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00019371606358121072,
+      "loss": 1.2216,
+      "step": 3690
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019368266624951455,
+      "loss": 1.185,
+      "step": 3695
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001936491832986565,
+      "loss": 1.1557,
+      "step": 3700
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019361561475923758,
+      "loss": 1.1541,
+      "step": 3705
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.00019358196066193703,
+      "loss": 1.1495,
+      "step": 3710
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001935482210375122,
+      "loss": 1.1735,
+      "step": 3715
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00019351439591679869,
+      "loss": 1.255,
+      "step": 3720
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001934804853307102,
+      "loss": 1.2078,
+      "step": 3725
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019344648931023852,
+      "loss": 1.0816,
+      "step": 3730
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.0001934124078864536,
+      "loss": 1.1158,
+      "step": 3735
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019337824109050335,
+      "loss": 1.2213,
+      "step": 3740
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019334398895361375,
+      "loss": 1.2004,
+      "step": 3745
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019330965150708883,
+      "loss": 1.2926,
+      "step": 3750
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001932752287823104,
+      "loss": 1.1325,
+      "step": 3755
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.193359375,
+      "learning_rate": 0.00019324072081073842,
+      "loss": 1.205,
+      "step": 3760
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019320612762391067,
+      "loss": 1.183,
+      "step": 3765
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019317144925344278,
+      "loss": 1.1753,
+      "step": 3770
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001931366857310283,
+      "loss": 1.0924,
+      "step": 3775
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019310183708843854,
+      "loss": 1.211,
+      "step": 3780
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019306690335752264,
+      "loss": 1.1834,
+      "step": 3785
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019303188457020753,
+      "loss": 1.1375,
+      "step": 3790
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019299678075849778,
+      "loss": 1.1875,
+      "step": 3795
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001929615919544758,
+      "loss": 1.1639,
+      "step": 3800
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001929263181903015,
+      "loss": 1.2174,
+      "step": 3805
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019289095949821266,
+      "loss": 1.1335,
+      "step": 3810
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019285551591052447,
+      "loss": 1.1638,
+      "step": 3815
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019281998745962988,
+      "loss": 1.2058,
+      "step": 3820
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019278437417799923,
+      "loss": 1.1903,
+      "step": 3825
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019274867609818052,
+      "loss": 1.2149,
+      "step": 3830
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019271289325279925,
+      "loss": 1.2248,
+      "step": 3835
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019267702567455825,
+      "loss": 1.1595,
+      "step": 3840
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00019264107339623797,
+      "loss": 1.0764,
+      "step": 3845
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001926050364506961,
+      "loss": 1.1806,
+      "step": 3850
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019256891487086787,
+      "loss": 1.1941,
+      "step": 3855
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019253270868976573,
+      "loss": 1.2064,
+      "step": 3860
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00019249641794047957,
+      "loss": 1.2227,
+      "step": 3865
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019246004265617645,
+      "loss": 1.2342,
+      "step": 3870
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019242358287010076,
+      "loss": 1.1305,
+      "step": 3875
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001923870386155741,
+      "loss": 1.1803,
+      "step": 3880
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019235040992599523,
+      "loss": 1.2234,
+      "step": 3885
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019231369683484018,
+      "loss": 1.192,
+      "step": 3890
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.000192276899375662,
+      "loss": 1.2201,
+      "step": 3895
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001922400175820909,
+      "loss": 1.1795,
+      "step": 3900
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001922030514878342,
+      "loss": 1.1543,
+      "step": 3905
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00019216600112667624,
+      "loss": 1.2045,
+      "step": 3910
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0001921288665324783,
+      "loss": 1.1899,
+      "step": 3915
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.00019209164773917876,
+      "loss": 1.1393,
+      "step": 3920
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00019205434478079288,
+      "loss": 1.1836,
+      "step": 3925
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019201695769141284,
+      "loss": 1.2185,
+      "step": 3930
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019197948650520776,
+      "loss": 1.1442,
+      "step": 3935
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019194193125642354,
+      "loss": 1.1998,
+      "step": 3940
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.000191904291979383,
+      "loss": 1.1879,
+      "step": 3945
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0001918665687084857,
+      "loss": 1.1614,
+      "step": 3950
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019182876147820794,
+      "loss": 1.2431,
+      "step": 3955
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00019179087032310284,
+      "loss": 1.2027,
+      "step": 3960
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019175289527780014,
+      "loss": 1.1537,
+      "step": 3965
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00019171483637700626,
+      "loss": 1.1182,
+      "step": 3970
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001916766936555043,
+      "loss": 1.1563,
+      "step": 3975
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00019163846714815392,
+      "loss": 1.1452,
+      "step": 3980
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001916001568898914,
+      "loss": 1.17,
+      "step": 3985
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00019156176291572954,
+      "loss": 1.2051,
+      "step": 3990
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001915232852607576,
+      "loss": 1.1872,
+      "step": 3995
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001914847239601414,
+      "loss": 1.1855,
+      "step": 4000
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019144607904912316,
+      "loss": 1.1743,
+      "step": 4005
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019140735056302158,
+      "loss": 1.1908,
+      "step": 4010
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001913685385372316,
+      "loss": 1.2331,
+      "step": 4015
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019132964300722468,
+      "loss": 1.1411,
+      "step": 4020
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019129066400854842,
+      "loss": 1.1171,
+      "step": 4025
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019125160157682688,
+      "loss": 1.1484,
+      "step": 4030
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019121245574776026,
+      "loss": 1.1838,
+      "step": 4035
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.000191173226557125,
+      "loss": 1.1931,
+      "step": 4040
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019113391404077375,
+      "loss": 1.1914,
+      "step": 4045
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00019109451823463527,
+      "loss": 1.1788,
+      "step": 4050
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019105503917471453,
+      "loss": 1.1897,
+      "step": 4055
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019101547689709243,
+      "loss": 1.1927,
+      "step": 4060
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001909758314379261,
+      "loss": 1.2086,
+      "step": 4065
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0001909361028334486,
+      "loss": 1.2443,
+      "step": 4070
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00019089629111996898,
+      "loss": 1.2293,
+      "step": 4075
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001908563963338723,
+      "loss": 1.2378,
+      "step": 4080
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00019081641851161941,
+      "loss": 1.1847,
+      "step": 4085
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019077635768974724,
+      "loss": 1.2248,
+      "step": 4090
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00019073621390486843,
+      "loss": 1.1274,
+      "step": 4095
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019069598719367146,
+      "loss": 1.2443,
+      "step": 4100
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00019065567759292065,
+      "loss": 1.1931,
+      "step": 4105
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001906152851394561,
+      "loss": 1.1672,
+      "step": 4110
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00019057480987019347,
+      "loss": 1.1768,
+      "step": 4115
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019053425182212434,
+      "loss": 1.1816,
+      "step": 4120
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00019049361103231575,
+      "loss": 1.1541,
+      "step": 4125
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001904528875379105,
+      "loss": 1.2016,
+      "step": 4130
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001904120813761268,
+      "loss": 1.1801,
+      "step": 4135
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019037119258425862,
+      "loss": 1.1839,
+      "step": 4140
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001903302211996753,
+      "loss": 1.1676,
+      "step": 4145
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00019028916725982178,
+      "loss": 1.2568,
+      "step": 4150
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019024803080221832,
+      "loss": 1.1593,
+      "step": 4155
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00019020681186446064,
+      "loss": 1.1637,
+      "step": 4160
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00019016551048421992,
+      "loss": 1.1169,
+      "step": 4165
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00019012412669924263,
+      "loss": 1.2356,
+      "step": 4170
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00019008266054735046,
+      "loss": 1.1956,
+      "step": 4175
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00019004111206644055,
+      "loss": 1.1518,
+      "step": 4180
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018999948129448516,
+      "loss": 1.2697,
+      "step": 4185
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00018995776826953182,
+      "loss": 1.1719,
+      "step": 4190
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001899159730297032,
+      "loss": 1.2134,
+      "step": 4195
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001898740956131971,
+      "loss": 1.168,
+      "step": 4200
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00018983213605828646,
+      "loss": 1.1682,
+      "step": 4205
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00018979009440331924,
+      "loss": 1.2043,
+      "step": 4210
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00018974797068671854,
+      "loss": 1.2089,
+      "step": 4215
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00018970576494698228,
+      "loss": 1.2662,
+      "step": 4220
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001896634772226835,
+      "loss": 1.2474,
+      "step": 4225
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018962110755247014,
+      "loss": 1.2028,
+      "step": 4230
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00018957865597506496,
+      "loss": 1.2415,
+      "step": 4235
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001895361225292656,
+      "loss": 1.163,
+      "step": 4240
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018949350725394463,
+      "loss": 1.2265,
+      "step": 4245
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018945081018804921,
+      "loss": 1.2452,
+      "step": 4250
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00018940803137060147,
+      "loss": 1.1117,
+      "step": 4255
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001893651708406981,
+      "loss": 1.1876,
+      "step": 4260
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00018932222863751047,
+      "loss": 1.1784,
+      "step": 4265
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001892792048002847,
+      "loss": 1.1368,
+      "step": 4270
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018923609936834148,
+      "loss": 1.15,
+      "step": 4275
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018919291238107596,
+      "loss": 1.1065,
+      "step": 4280
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.000189149643877958,
+      "loss": 1.1503,
+      "step": 4285
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001891062938985318,
+      "loss": 1.2226,
+      "step": 4290
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018906286248241615,
+      "loss": 1.1709,
+      "step": 4295
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00018901934966930416,
+      "loss": 1.1334,
+      "step": 4300
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00018897575549896344,
+      "loss": 1.2418,
+      "step": 4305
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001889320800112359,
+      "loss": 1.2129,
+      "step": 4310
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00018888832324603772,
+      "loss": 1.1138,
+      "step": 4315
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00018884448524335943,
+      "loss": 1.1346,
+      "step": 4320
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00018880056604326582,
+      "loss": 1.1759,
+      "step": 4325
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001887565656858958,
+      "loss": 1.2258,
+      "step": 4330
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00018871248421146254,
+      "loss": 1.1465,
+      "step": 4335
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001886683216602533,
+      "loss": 1.1522,
+      "step": 4340
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00018862407807262947,
+      "loss": 1.0901,
+      "step": 4345
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018857975348902643,
+      "loss": 1.1998,
+      "step": 4350
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001885353479499537,
+      "loss": 1.2311,
+      "step": 4355
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00018849086149599464,
+      "loss": 1.1242,
+      "step": 4360
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001884462941678067,
+      "loss": 1.1467,
+      "step": 4365
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018840164600612117,
+      "loss": 1.211,
+      "step": 4370
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00018835691705174328,
+      "loss": 1.1764,
+      "step": 4375
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00018831210734555196,
+      "loss": 1.0915,
+      "step": 4380
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00018826721692850006,
+      "loss": 1.1972,
+      "step": 4385
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001882222458416142,
+      "loss": 1.2591,
+      "step": 4390
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001881771941259947,
+      "loss": 1.1628,
+      "step": 4395
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001881320618228155,
+      "loss": 1.1768,
+      "step": 4400
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001880868489733243,
+      "loss": 1.1984,
+      "step": 4405
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018804155561884234,
+      "loss": 1.2011,
+      "step": 4410
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018799618180076447,
+      "loss": 1.2258,
+      "step": 4415
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00018795072756055907,
+      "loss": 1.2561,
+      "step": 4420
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00018790519293976801,
+      "loss": 1.199,
+      "step": 4425
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00018785957798000667,
+      "loss": 1.2548,
+      "step": 4430
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001878138827229637,
+      "loss": 1.1135,
+      "step": 4435
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018776810721040144,
+      "loss": 1.1236,
+      "step": 4440
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001877222514841552,
+      "loss": 1.1822,
+      "step": 4445
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001876763155861339,
+      "loss": 1.1692,
+      "step": 4450
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.0001876302995583196,
+      "loss": 1.1848,
+      "step": 4455
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001875842034427676,
+      "loss": 1.248,
+      "step": 4460
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018753802728160637,
+      "loss": 1.2719,
+      "step": 4465
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018749177111703765,
+      "loss": 1.229,
+      "step": 4470
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00018744543499133614,
+      "loss": 1.1709,
+      "step": 4475
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018739901894684974,
+      "loss": 1.1611,
+      "step": 4480
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00018735252302599936,
+      "loss": 1.1862,
+      "step": 4485
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00018730594727127886,
+      "loss": 1.1517,
+      "step": 4490
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00018725929172525515,
+      "loss": 1.1763,
+      "step": 4495
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00018721255643056796,
+      "loss": 1.1797,
+      "step": 4500
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00018716574142993003,
+      "loss": 1.1808,
+      "step": 4505
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001871188467661268,
+      "loss": 1.2529,
+      "step": 4510
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018707187248201662,
+      "loss": 1.1896,
+      "step": 4515
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001870248186205306,
+      "loss": 1.1237,
+      "step": 4520
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001869776852246725,
+      "loss": 1.1672,
+      "step": 4525
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.150390625,
+      "learning_rate": 0.0001869304723375189,
+      "loss": 1.1375,
+      "step": 4530
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001868831800022189,
+      "loss": 1.2534,
+      "step": 4535
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001868358082619943,
+      "loss": 1.1793,
+      "step": 4540
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00018678835716013942,
+      "loss": 1.1798,
+      "step": 4545
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001867408267400211,
+      "loss": 1.1684,
+      "step": 4550
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00018669321704507875,
+      "loss": 1.1877,
+      "step": 4555
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00018664552811882413,
+      "loss": 1.2018,
+      "step": 4560
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001865977600048415,
+      "loss": 1.1956,
+      "step": 4565
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001865499127467874,
+      "loss": 1.1512,
+      "step": 4570
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018650198638839078,
+      "loss": 1.1813,
+      "step": 4575
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018645398097345286,
+      "loss": 1.1737,
+      "step": 4580
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001864058965458471,
+      "loss": 1.1519,
+      "step": 4585
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001863577331495192,
+      "loss": 1.1106,
+      "step": 4590
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.000186309490828487,
+      "loss": 1.1294,
+      "step": 4595
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00018626116962684047,
+      "loss": 1.1882,
+      "step": 4600
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00018621276958874166,
+      "loss": 1.1629,
+      "step": 4605
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00018616429075842477,
+      "loss": 1.1658,
+      "step": 4610
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00018611573318019585,
+      "loss": 1.0996,
+      "step": 4615
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 0.00018606709689843304,
+      "loss": 1.1976,
+      "step": 4620
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001860183819575864,
+      "loss": 1.2164,
+      "step": 4625
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001859695884021778,
+      "loss": 1.2458,
+      "step": 4630
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00018592071627680106,
+      "loss": 1.1858,
+      "step": 4635
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018587176562612174,
+      "loss": 1.1194,
+      "step": 4640
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 0.0001858227364948772,
+      "loss": 1.218,
+      "step": 4645
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00018577362892787646,
+      "loss": 1.2114,
+      "step": 4650
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018572444297000032,
+      "loss": 1.166,
+      "step": 4655
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00018567517866620116,
+      "loss": 1.1793,
+      "step": 4660
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018562583606150296,
+      "loss": 1.1417,
+      "step": 4665
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018557641520100133,
+      "loss": 1.1991,
+      "step": 4670
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018552691612986334,
+      "loss": 1.1739,
+      "step": 4675
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00018547733889332752,
+      "loss": 1.0675,
+      "step": 4680
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00018542768353670388,
+      "loss": 1.2482,
+      "step": 4685
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00018537795010537385,
+      "loss": 1.1433,
+      "step": 4690
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00018532813864479013,
+      "loss": 1.18,
+      "step": 4695
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00018527824920047678,
+      "loss": 1.1855,
+      "step": 4700
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.146484375,
+      "learning_rate": 0.00018522828181802918,
+      "loss": 1.1978,
+      "step": 4705
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00018517823654311385,
+      "loss": 1.283,
+      "step": 4710
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001851281134214686,
+      "loss": 1.1947,
+      "step": 4715
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00018507791249890225,
+      "loss": 1.215,
+      "step": 4720
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1484375,
+      "learning_rate": 0.00018502763382129484,
+      "loss": 1.2102,
+      "step": 4725
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00018497727743459742,
+      "loss": 1.1311,
+      "step": 4730
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00018492684338483213,
+      "loss": 1.2421,
+      "step": 4735
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00018487633171809198,
+      "loss": 1.1825,
+      "step": 4740
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00018482574248054098,
+      "loss": 1.1921,
+      "step": 4745
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00018477507571841402,
+      "loss": 1.2143,
+      "step": 4750
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00018472433147801684,
+      "loss": 1.1863,
+      "step": 4755
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.000184673509805726,
+      "loss": 1.2314,
+      "step": 4760
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018462261074798887,
+      "loss": 1.1281,
+      "step": 4765
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00018457163435132344,
+      "loss": 1.2231,
+      "step": 4770
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018452058066231844,
+      "loss": 1.2033,
+      "step": 4775
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018446944972763324,
+      "loss": 1.2101,
+      "step": 4780
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018441824159399784,
+      "loss": 1.1672,
+      "step": 4785
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00018436695630821273,
+      "loss": 1.1587,
+      "step": 4790
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018431559391714894,
+      "loss": 1.1598,
+      "step": 4795
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.000184264154467748,
+      "loss": 1.1382,
+      "step": 4800
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001842126380070218,
+      "loss": 1.0878,
+      "step": 4805
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00018416104458205266,
+      "loss": 1.2027,
+      "step": 4810
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00018410937423999325,
+      "loss": 1.1148,
+      "step": 4815
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018405762702806652,
+      "loss": 1.2245,
+      "step": 4820
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018400580299356557,
+      "loss": 1.241,
+      "step": 4825
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018395390218385393,
+      "loss": 1.1644,
+      "step": 4830
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018390192464636513,
+      "loss": 1.2225,
+      "step": 4835
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018384987042860287,
+      "loss": 1.1297,
+      "step": 4840
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001837977395781409,
+      "loss": 1.2017,
+      "step": 4845
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001837455321426231,
+      "loss": 1.1781,
+      "step": 4850
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018369324816976324,
+      "loss": 1.2031,
+      "step": 4855
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00018364088770734507,
+      "loss": 1.275,
+      "step": 4860
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 0.00018358845080322227,
+      "loss": 1.1411,
+      "step": 4865
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018353593750531836,
+      "loss": 1.2066,
+      "step": 4870
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018348334786162668,
+      "loss": 1.1122,
+      "step": 4875
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00018343068192021037,
+      "loss": 1.2278,
+      "step": 4880
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00018337793972920228,
+      "loss": 1.1732,
+      "step": 4885
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00018332512133680494,
+      "loss": 1.1678,
+      "step": 4890
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018327222679129048,
+      "loss": 1.1443,
+      "step": 4895
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018321925614100075,
+      "loss": 1.149,
+      "step": 4900
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018316620943434705,
+      "loss": 1.1907,
+      "step": 4905
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00018311308671981025,
+      "loss": 1.2136,
+      "step": 4910
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018305988804594064,
+      "loss": 1.1774,
+      "step": 4915
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018300661346135793,
+      "loss": 1.1947,
+      "step": 4920
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00018295326301475125,
+      "loss": 1.2368,
+      "step": 4925
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.14453125,
+      "learning_rate": 0.000182899836754879,
+      "loss": 1.1278,
+      "step": 4930
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00018284633473056896,
+      "loss": 1.2375,
+      "step": 4935
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018279275699071808,
+      "loss": 1.1362,
+      "step": 4940
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018273910358429252,
+      "loss": 1.1237,
+      "step": 4945
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00018268537456032758,
+      "loss": 1.1155,
+      "step": 4950
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018263156996792777,
+      "loss": 1.1875,
+      "step": 4955
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018257768985626652,
+      "loss": 1.2704,
+      "step": 4960
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018252373427458634,
+      "loss": 1.2003,
+      "step": 4965
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001824697032721987,
+      "loss": 1.2654,
+      "step": 4970
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001824155968984841,
+      "loss": 1.1861,
+      "step": 4975
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018236141520289176,
+      "loss": 1.1147,
+      "step": 4980
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00018230715823493987,
+      "loss": 1.2168,
+      "step": 4985
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018225282604421533,
+      "loss": 1.1606,
+      "step": 4990
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018219841868037385,
+      "loss": 1.2405,
+      "step": 4995
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001821439361931398,
+      "loss": 1.2015,
+      "step": 5000
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00018208937863230625,
+      "loss": 1.1589,
+      "step": 5005
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001820347460477348,
+      "loss": 1.1751,
+      "step": 5010
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001819800384893557,
+      "loss": 1.1279,
+      "step": 5015
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00018192525600716769,
+      "loss": 1.1365,
+      "step": 5020
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018187039865123793,
+      "loss": 1.1537,
+      "step": 5025
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00018181546647170216,
+      "loss": 1.2473,
+      "step": 5030
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00018176045951876434,
+      "loss": 1.2583,
+      "step": 5035
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00018170537784269688,
+      "loss": 1.2115,
+      "step": 5040
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00018165022149384044,
+      "loss": 1.227,
+      "step": 5045
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00018159499052260384,
+      "loss": 1.1894,
+      "step": 5050
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018153968497946425,
+      "loss": 1.2358,
+      "step": 5055
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018148430491496696,
+      "loss": 1.1847,
+      "step": 5060
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00018142885037972524,
+      "loss": 1.1305,
+      "step": 5065
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00018137332142442055,
+      "loss": 1.156,
+      "step": 5070
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018131771809980237,
+      "loss": 1.191,
+      "step": 5075
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018126204045668804,
+      "loss": 1.1395,
+      "step": 5080
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00018120628854596295,
+      "loss": 1.1608,
+      "step": 5085
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00018115046241858025,
+      "loss": 1.1623,
+      "step": 5090
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.000181094562125561,
+      "loss": 1.2233,
+      "step": 5095
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.000181038587717994,
+      "loss": 1.1292,
+      "step": 5100
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00018098253924703583,
+      "loss": 1.2846,
+      "step": 5105
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001809264167639107,
+      "loss": 1.1793,
+      "step": 5110
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00018087022031991045,
+      "loss": 1.2,
+      "step": 5115
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001808139499663946,
+      "loss": 1.2053,
+      "step": 5120
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001807576057547902,
+      "loss": 1.1382,
+      "step": 5125
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001807011877365917,
+      "loss": 1.2428,
+      "step": 5130
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00018064469596336115,
+      "loss": 1.2063,
+      "step": 5135
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001805881304867279,
+      "loss": 1.1638,
+      "step": 5140
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00018053149135838865,
+      "loss": 1.2003,
+      "step": 5145
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00018047477863010754,
+      "loss": 1.2658,
+      "step": 5150
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001804179923537158,
+      "loss": 1.1726,
+      "step": 5155
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00018036113258111205,
+      "loss": 1.1496,
+      "step": 5160
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00018030419936426192,
+      "loss": 1.163,
+      "step": 5165
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001802471927551983,
+      "loss": 1.2435,
+      "step": 5170
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00018019011280602108,
+      "loss": 1.1734,
+      "step": 5175
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00018013295956889722,
+      "loss": 1.1583,
+      "step": 5180
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00018007573309606058,
+      "loss": 1.1795,
+      "step": 5185
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00018001843343981203,
+      "loss": 1.1806,
+      "step": 5190
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00017996106065251932,
+      "loss": 1.1773,
+      "step": 5195
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00017990361478661696,
+      "loss": 1.1526,
+      "step": 5200
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00017984609589460636,
+      "loss": 1.1713,
+      "step": 5205
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001797885040290556,
+      "loss": 1.227,
+      "step": 5210
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017973083924259945,
+      "loss": 1.1794,
+      "step": 5215
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00017967310158793938,
+      "loss": 1.2075,
+      "step": 5220
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00017961529111784337,
+      "loss": 1.2101,
+      "step": 5225
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00017955740788514604,
+      "loss": 1.1414,
+      "step": 5230
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00017949945194274843,
+      "loss": 1.23,
+      "step": 5235
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00017944142334361807,
+      "loss": 1.1522,
+      "step": 5240
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00017938332214078893,
+      "loss": 1.1904,
+      "step": 5245
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00017932514838736124,
+      "loss": 1.194,
+      "step": 5250
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00017926690213650165,
+      "loss": 1.2166,
+      "step": 5255
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017920858344144298,
+      "loss": 1.2755,
+      "step": 5260
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001791501923554843,
+      "loss": 1.2374,
+      "step": 5265
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001790917289319908,
+      "loss": 1.1829,
+      "step": 5270
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001790331932243938,
+      "loss": 1.1124,
+      "step": 5275
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00017897458528619073,
+      "loss": 1.1659,
+      "step": 5280
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00017891590517094498,
+      "loss": 1.2141,
+      "step": 5285
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00017885715293228593,
+      "loss": 1.1627,
+      "step": 5290
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00017879832862390885,
+      "loss": 1.1381,
+      "step": 5295
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017873943229957488,
+      "loss": 1.2275,
+      "step": 5300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.000178680464013111,
+      "loss": 1.1676,
+      "step": 5305
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017862142381840996,
+      "loss": 1.1738,
+      "step": 5310
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001785623117694302,
+      "loss": 1.1766,
+      "step": 5315
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00017850312792019586,
+      "loss": 1.1603,
+      "step": 5320
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017844387232479665,
+      "loss": 1.1815,
+      "step": 5325
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001783845450373879,
+      "loss": 1.1621,
+      "step": 5330
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001783251461121905,
+      "loss": 1.1822,
+      "step": 5335
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001782656756034907,
+      "loss": 1.1437,
+      "step": 5340
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017820613356564025,
+      "loss": 1.1738,
+      "step": 5345
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00017814652005305624,
+      "loss": 1.1824,
+      "step": 5350
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017808683512022112,
+      "loss": 1.2294,
+      "step": 5355
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001780270788216826,
+      "loss": 1.1425,
+      "step": 5360
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00017796725121205356,
+      "loss": 1.2155,
+      "step": 5365
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001779073523460121,
+      "loss": 1.181,
+      "step": 5370
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00017784738227830146,
+      "loss": 1.1912,
+      "step": 5375
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00017778734106372996,
+      "loss": 1.1595,
+      "step": 5380
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017772722875717087,
+      "loss": 1.1602,
+      "step": 5385
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017766704541356248,
+      "loss": 1.1615,
+      "step": 5390
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.000177606791087908,
+      "loss": 1.1613,
+      "step": 5395
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00017754646583527554,
+      "loss": 1.1412,
+      "step": 5400
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.000177486069710798,
+      "loss": 1.2031,
+      "step": 5405
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00017742560276967306,
+      "loss": 1.0884,
+      "step": 5410
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00017736506506716308,
+      "loss": 1.1752,
+      "step": 5415
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00017730445665859517,
+      "loss": 1.2408,
+      "step": 5420
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017724377759936098,
+      "loss": 1.1372,
+      "step": 5425
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017718302794491683,
+      "loss": 1.1929,
+      "step": 5430
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00017712220775078344,
+      "loss": 1.2121,
+      "step": 5435
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00017706131707254607,
+      "loss": 1.2263,
+      "step": 5440
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00017700035596585441,
+      "loss": 1.1945,
+      "step": 5445
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00017693932448642244,
+      "loss": 1.1754,
+      "step": 5450
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017687822269002855,
+      "loss": 1.1458,
+      "step": 5455
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017681705063251534,
+      "loss": 1.1236,
+      "step": 5460
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017675580836978966,
+      "loss": 1.1502,
+      "step": 5465
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00017669449595782242,
+      "loss": 1.2136,
+      "step": 5470
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001766331134526488,
+      "loss": 1.1455,
+      "step": 5475
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00017657166091036787,
+      "loss": 1.2175,
+      "step": 5480
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00017651013838714287,
+      "loss": 1.1593,
+      "step": 5485
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001764485459392009,
+      "loss": 1.1017,
+      "step": 5490
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.000176386883622833,
+      "loss": 1.1422,
+      "step": 5495
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017632515149439403,
+      "loss": 1.1333,
+      "step": 5500
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00017626334961030272,
+      "loss": 1.1235,
+      "step": 5505
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00017620147802704147,
+      "loss": 1.1761,
+      "step": 5510
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.0001761395368011564,
+      "loss": 1.2288,
+      "step": 5515
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00017607752598925737,
+      "loss": 1.1904,
+      "step": 5520
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017601544564801772,
+      "loss": 1.1194,
+      "step": 5525
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00017595329583417434,
+      "loss": 1.2093,
+      "step": 5530
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00017589107660452772,
+      "loss": 1.1759,
+      "step": 5535
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017582878801594172,
+      "loss": 1.1685,
+      "step": 5540
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00017576643012534352,
+      "loss": 1.106,
+      "step": 5545
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001757040029897238,
+      "loss": 1.2092,
+      "step": 5550
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00017564150666613638,
+      "loss": 1.2147,
+      "step": 5555
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001755789412116984,
+      "loss": 1.1028,
+      "step": 5560
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00017551630668359008,
+      "loss": 1.2131,
+      "step": 5565
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00017545360313905492,
+      "loss": 1.1669,
+      "step": 5570
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00017539083063539934,
+      "loss": 1.2072,
+      "step": 5575
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017532798922999292,
+      "loss": 1.1919,
+      "step": 5580
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001752650789802681,
+      "loss": 1.2274,
+      "step": 5585
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001752020999437203,
+      "loss": 1.2395,
+      "step": 5590
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001751390521779078,
+      "loss": 1.1656,
+      "step": 5595
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00017507593574045166,
+      "loss": 1.148,
+      "step": 5600
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00017501275068903575,
+      "loss": 1.2134,
+      "step": 5605
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017494949708140662,
+      "loss": 1.2303,
+      "step": 5610
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017488617497537348,
+      "loss": 1.1557,
+      "step": 5615
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00017482278442880814,
+      "loss": 1.2181,
+      "step": 5620
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00017475932549964492,
+      "loss": 1.1929,
+      "step": 5625
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00017469579824588077,
+      "loss": 1.2005,
+      "step": 5630
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001746322027255749,
+      "loss": 1.1688,
+      "step": 5635
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00017456853899684905,
+      "loss": 1.1488,
+      "step": 5640
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00017450480711788728,
+      "loss": 1.1308,
+      "step": 5645
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017444100714693583,
+      "loss": 1.1607,
+      "step": 5650
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00017437713914230328,
+      "loss": 1.195,
+      "step": 5655
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00017431320316236038,
+      "loss": 1.2241,
+      "step": 5660
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00017424919926553993,
+      "loss": 1.1902,
+      "step": 5665
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001741851275103369,
+      "loss": 1.1647,
+      "step": 5670
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001741209879553082,
+      "loss": 1.1406,
+      "step": 5675
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017405678065907272,
+      "loss": 1.1755,
+      "step": 5680
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001739925056803113,
+      "loss": 1.1539,
+      "step": 5685
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00017392816307776655,
+      "loss": 1.2184,
+      "step": 5690
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.000173863752910243,
+      "loss": 1.1078,
+      "step": 5695
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001737992752366069,
+      "loss": 1.1312,
+      "step": 5700
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00017373473011578607,
+      "loss": 1.2622,
+      "step": 5705
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00017367011760677014,
+      "loss": 1.1566,
+      "step": 5710
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00017360543776861028,
+      "loss": 1.1736,
+      "step": 5715
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017354069066041907,
+      "loss": 1.1516,
+      "step": 5720
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017347587634137076,
+      "loss": 1.0988,
+      "step": 5725
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00017341099487070088,
+      "loss": 1.1291,
+      "step": 5730
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017334604630770643,
+      "loss": 1.1378,
+      "step": 5735
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00017328103071174566,
+      "loss": 1.1181,
+      "step": 5740
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00017321594814223814,
+      "loss": 1.2002,
+      "step": 5745
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00017315079865866457,
+      "loss": 1.2042,
+      "step": 5750
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001730855823205669,
+      "loss": 1.1689,
+      "step": 5755
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00017302029918754807,
+      "loss": 1.2722,
+      "step": 5760
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00017295494931927224,
+      "loss": 1.2428,
+      "step": 5765
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001728895327754643,
+      "loss": 1.1916,
+      "step": 5770
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00017282404961591036,
+      "loss": 1.1339,
+      "step": 5775
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00017275849990045723,
+      "loss": 1.1971,
+      "step": 5780
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00017269288368901257,
+      "loss": 1.1746,
+      "step": 5785
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00017262720104154483,
+      "loss": 1.1285,
+      "step": 5790
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00017256145201808326,
+      "loss": 1.2119,
+      "step": 5795
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017249563667871764,
+      "loss": 1.1001,
+      "step": 5800
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001724297550835984,
+      "loss": 1.1578,
+      "step": 5805
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001723638072929366,
+      "loss": 1.2169,
+      "step": 5810
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00017229779336700368,
+      "loss": 1.2047,
+      "step": 5815
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001722317133661316,
+      "loss": 1.2149,
+      "step": 5820
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00017216556735071272,
+      "loss": 1.1894,
+      "step": 5825
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00017209935538119967,
+      "loss": 1.1614,
+      "step": 5830
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001720330775181054,
+      "loss": 1.1752,
+      "step": 5835
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.000171966733822003,
+      "loss": 1.2206,
+      "step": 5840
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001719003243535259,
+      "loss": 1.1617,
+      "step": 5845
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00017183384917336748,
+      "loss": 1.1625,
+      "step": 5850
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00017176730834228123,
+      "loss": 1.052,
+      "step": 5855
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001717007019210807,
+      "loss": 1.1625,
+      "step": 5860
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00017163402997063927,
+      "loss": 1.1776,
+      "step": 5865
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00017156729255189028,
+      "loss": 1.1873,
+      "step": 5870
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001715004897258269,
+      "loss": 1.1842,
+      "step": 5875
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00017143362155350205,
+      "loss": 1.1114,
+      "step": 5880
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001713666880960284,
+      "loss": 1.1181,
+      "step": 5885
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00017129968941457827,
+      "loss": 1.1281,
+      "step": 5890
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001712326255703836,
+      "loss": 1.2166,
+      "step": 5895
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00017116549662473594,
+      "loss": 1.1349,
+      "step": 5900
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.00017109830263898617,
+      "loss": 1.1106,
+      "step": 5905
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00017103104367454476,
+      "loss": 1.1913,
+      "step": 5910
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00017096371979288154,
+      "loss": 1.2241,
+      "step": 5915
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00017089633105552566,
+      "loss": 1.1808,
+      "step": 5920
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001708288775240655,
+      "loss": 1.1804,
+      "step": 5925
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00017076135926014874,
+      "loss": 1.1897,
+      "step": 5930
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00017069377632548212,
+      "loss": 1.1233,
+      "step": 5935
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001706261287818316,
+      "loss": 1.2036,
+      "step": 5940
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00017055841669102212,
+      "loss": 1.2187,
+      "step": 5945
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.154296875,
+      "learning_rate": 0.0001704906401149376,
+      "loss": 1.1052,
+      "step": 5950
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001704227991155209,
+      "loss": 1.103,
+      "step": 5955
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00017035489375477383,
+      "loss": 1.224,
+      "step": 5960
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001702869240947569,
+      "loss": 1.17,
+      "step": 5965
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00017021889019758945,
+      "loss": 1.1957,
+      "step": 5970
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00017015079212544956,
+      "loss": 1.1352,
+      "step": 5975
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00017008262994057394,
+      "loss": 1.1536,
+      "step": 5980
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00017001440370525782,
+      "loss": 1.157,
+      "step": 5985
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00016994611348185513,
+      "loss": 1.1929,
+      "step": 5990
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001698777593327781,
+      "loss": 1.1516,
+      "step": 5995
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00016980934132049751,
+      "loss": 1.1501,
+      "step": 6000
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00016974085950754238,
+      "loss": 1.1923,
+      "step": 6005
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00016967231395650028,
+      "loss": 1.2744,
+      "step": 6010
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001696037047300167,
+      "loss": 1.1086,
+      "step": 6015
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00016953503189079557,
+      "loss": 1.2266,
+      "step": 6020
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001694662955015989,
+      "loss": 1.1767,
+      "step": 6025
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00016939749562524674,
+      "loss": 1.2378,
+      "step": 6030
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00016932863232461718,
+      "loss": 1.1885,
+      "step": 6035
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00016925970566264632,
+      "loss": 1.1633,
+      "step": 6040
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00016919071570232808,
+      "loss": 1.1965,
+      "step": 6045
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001691216625067143,
+      "loss": 1.1941,
+      "step": 6050
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001690525461389146,
+      "loss": 1.1419,
+      "step": 6055
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00016898336666209635,
+      "loss": 1.1146,
+      "step": 6060
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00016891412413948454,
+      "loss": 1.2513,
+      "step": 6065
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00016884481863436182,
+      "loss": 1.1577,
+      "step": 6070
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00016877545021006839,
+      "loss": 1.1731,
+      "step": 6075
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00016870601893000193,
+      "loss": 1.1596,
+      "step": 6080
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00016863652485761766,
+      "loss": 1.2516,
+      "step": 6085
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00016856696805642808,
+      "loss": 1.1978,
+      "step": 6090
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001684973485900031,
+      "loss": 1.1111,
+      "step": 6095
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00016842766652196977,
+      "loss": 1.1747,
+      "step": 6100
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00016835792191601255,
+      "loss": 1.1575,
+      "step": 6105
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001682881148358729,
+      "loss": 1.1989,
+      "step": 6110
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00016821824534534946,
+      "loss": 1.1933,
+      "step": 6115
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.00016814831350829782,
+      "loss": 1.1758,
+      "step": 6120
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00016807831938863067,
+      "loss": 1.2157,
+      "step": 6125
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001680082630503175,
+      "loss": 1.2277,
+      "step": 6130
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001679381445573848,
+      "loss": 1.1707,
+      "step": 6135
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00016786796397391568,
+      "loss": 1.1909,
+      "step": 6140
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001677977213640502,
+      "loss": 1.182,
+      "step": 6145
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00016772741679198493,
+      "loss": 1.1349,
+      "step": 6150
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00016765705032197324,
+      "loss": 1.1642,
+      "step": 6155
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00016758662201832487,
+      "loss": 1.1889,
+      "step": 6160
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001675161319454063,
+      "loss": 1.2314,
+      "step": 6165
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00016744558016764028,
+      "loss": 1.1359,
+      "step": 6170
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00016737496674950597,
+      "loss": 1.0452,
+      "step": 6175
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.000167304291755539,
+      "loss": 1.2646,
+      "step": 6180
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00016723355525033115,
+      "loss": 1.2203,
+      "step": 6185
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00016716275729853047,
+      "loss": 1.1943,
+      "step": 6190
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00016709189796484112,
+      "loss": 1.1427,
+      "step": 6195
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00016702097731402345,
+      "loss": 1.2121,
+      "step": 6200
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001669499954108937,
+      "loss": 1.1631,
+      "step": 6205
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00016687895232032424,
+      "loss": 1.1884,
+      "step": 6210
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00016680784810724333,
+      "loss": 1.1243,
+      "step": 6215
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.000166736682836635,
+      "loss": 1.2107,
+      "step": 6220
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.0001666654565735392,
+      "loss": 1.1739,
+      "step": 6225
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00016659416938305155,
+      "loss": 1.147,
+      "step": 6230
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00016652282133032334,
+      "loss": 1.144,
+      "step": 6235
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00016645141248056153,
+      "loss": 1.171,
+      "step": 6240
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00016637994289902866,
+      "loss": 1.1352,
+      "step": 6245
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001663084126510427,
+      "loss": 1.1661,
+      "step": 6250
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00016623682180197716,
+      "loss": 1.1711,
+      "step": 6255
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00016616517041726083,
+      "loss": 1.1735,
+      "step": 6260
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00016609345856237793,
+      "loss": 1.1531,
+      "step": 6265
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00016602168630286784,
+      "loss": 1.1722,
+      "step": 6270
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00016594985370432523,
+      "loss": 1.1805,
+      "step": 6275
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00016587796083239987,
+      "loss": 1.2066,
+      "step": 6280
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00016580600775279667,
+      "loss": 1.1717,
+      "step": 6285
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00016573399453127547,
+      "loss": 1.1868,
+      "step": 6290
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00016566192123365116,
+      "loss": 1.1424,
+      "step": 6295
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00016558978792579352,
+      "loss": 1.1829,
+      "step": 6300
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00016551759467362714,
+      "loss": 1.1477,
+      "step": 6305
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001654453415431314,
+      "loss": 1.1573,
+      "step": 6310
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00016537302860034048,
+      "loss": 1.1836,
+      "step": 6315
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001653006559113431,
+      "loss": 1.2053,
+      "step": 6320
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00016522822354228268,
+      "loss": 1.2679,
+      "step": 6325
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00016515573155935712,
+      "loss": 1.1288,
+      "step": 6330
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00016508318002881892,
+      "loss": 1.1524,
+      "step": 6335
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001650105690169748,
+      "loss": 1.1965,
+      "step": 6340
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00016493789859018605,
+      "loss": 1.1372,
+      "step": 6345
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00016486516881486813,
+      "loss": 1.1815,
+      "step": 6350
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00016479237975749078,
+      "loss": 1.2207,
+      "step": 6355
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001647195314845779,
+      "loss": 1.0963,
+      "step": 6360
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00016464662406270758,
+      "loss": 1.0869,
+      "step": 6365
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001645736575585119,
+      "loss": 1.1557,
+      "step": 6370
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001645006320386769,
+      "loss": 1.248,
+      "step": 6375
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00016442754756994265,
+      "loss": 1.1437,
+      "step": 6380
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00016435440421910306,
+      "loss": 1.1352,
+      "step": 6385
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00016428120205300585,
+      "loss": 1.1734,
+      "step": 6390
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00016420794113855243,
+      "loss": 1.2053,
+      "step": 6395
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00016413462154269808,
+      "loss": 1.1419,
+      "step": 6400
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00016406124333245145,
+      "loss": 1.1326,
+      "step": 6405
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00016398780657487505,
+      "loss": 1.1899,
+      "step": 6410
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00016391431133708462,
+      "loss": 1.1499,
+      "step": 6415
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00016384075768624955,
+      "loss": 1.1841,
+      "step": 6420
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001637671456895925,
+      "loss": 1.0846,
+      "step": 6425
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00016369347541438954,
+      "loss": 1.1874,
+      "step": 6430
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001636197469279699,
+      "loss": 1.1702,
+      "step": 6435
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00016354596029771615,
+      "loss": 1.2064,
+      "step": 6440
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00016347211559106383,
+      "loss": 1.1969,
+      "step": 6445
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00016339821287550171,
+      "loss": 1.1114,
+      "step": 6450
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00016332425221857148,
+      "loss": 1.1589,
+      "step": 6455
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00016325023368786784,
+      "loss": 1.1911,
+      "step": 6460
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00016317615735103838,
+      "loss": 1.1119,
+      "step": 6465
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00016310202327578338,
+      "loss": 1.2206,
+      "step": 6470
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00016302783152985614,
+      "loss": 1.22,
+      "step": 6475
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00016295358218106247,
+      "loss": 1.1452,
+      "step": 6480
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001628792752972609,
+      "loss": 1.2454,
+      "step": 6485
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00016280491094636257,
+      "loss": 1.1665,
+      "step": 6490
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00016273048919633105,
+      "loss": 1.1854,
+      "step": 6495
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001626560101151824,
+      "loss": 1.1945,
+      "step": 6500
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001625814737709852,
+      "loss": 1.172,
+      "step": 6505
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00016250688023186017,
+      "loss": 1.2228,
+      "step": 6510
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001624322295659804,
+      "loss": 1.2405,
+      "step": 6515
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00016235752184157125,
+      "loss": 1.1174,
+      "step": 6520
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001622827571269101,
+      "loss": 1.1757,
+      "step": 6525
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001622079354903265,
+      "loss": 1.1761,
+      "step": 6530
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.000162133057000202,
+      "loss": 1.1007,
+      "step": 6535
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001620581217249701,
+      "loss": 1.1801,
+      "step": 6540
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001619831297331162,
+      "loss": 1.1957,
+      "step": 6545
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00016190808109317755,
+      "loss": 1.1597,
+      "step": 6550
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001618329758737432,
+      "loss": 1.1873,
+      "step": 6555
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001617578141434538,
+      "loss": 1.1642,
+      "step": 6560
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00016168259597100178,
+      "loss": 1.1846,
+      "step": 6565
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00016160732142513106,
+      "loss": 1.1687,
+      "step": 6570
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00016153199057463712,
+      "loss": 1.123,
+      "step": 6575
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00016145660348836693,
+      "loss": 1.1444,
+      "step": 6580
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00016138116023521875,
+      "loss": 1.1467,
+      "step": 6585
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001613056608841423,
+      "loss": 1.1102,
+      "step": 6590
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001612301055041385,
+      "loss": 1.1791,
+      "step": 6595
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00016115449416425945,
+      "loss": 1.2453,
+      "step": 6600
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00016107882693360844,
+      "loss": 1.1349,
+      "step": 6605
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001610031038813399,
+      "loss": 1.0905,
+      "step": 6610
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00016092732507665914,
+      "loss": 1.1994,
+      "step": 6615
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001608514905888225,
+      "loss": 1.1689,
+      "step": 6620
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001607756004871372,
+      "loss": 1.1518,
+      "step": 6625
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001606996548409613,
+      "loss": 1.1862,
+      "step": 6630
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00016062365371970362,
+      "loss": 1.2025,
+      "step": 6635
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00016054759719282365,
+      "loss": 1.228,
+      "step": 6640
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00016047148532983156,
+      "loss": 1.1946,
+      "step": 6645
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00016039531820028806,
+      "loss": 1.1171,
+      "step": 6650
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001603190958738044,
+      "loss": 1.1555,
+      "step": 6655
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00016024281842004221,
+      "loss": 1.1259,
+      "step": 6660
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.15234375,
+      "learning_rate": 0.0001601664859087136,
+      "loss": 1.1639,
+      "step": 6665
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001600900984095809,
+      "loss": 1.1768,
+      "step": 6670
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.205078125,
+      "learning_rate": 0.0001600136559924568,
+      "loss": 1.1152,
+      "step": 6675
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001599371587272041,
+      "loss": 1.2518,
+      "step": 6680
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00015986060668373575,
+      "loss": 1.2053,
+      "step": 6685
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00015978399993201475,
+      "loss": 1.1711,
+      "step": 6690
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00015970733854205413,
+      "loss": 1.2553,
+      "step": 6695
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00015963062258391686,
+      "loss": 1.1595,
+      "step": 6700
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00015955385212771572,
+      "loss": 1.2224,
+      "step": 6705
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00015947702724361336,
+      "loss": 1.1493,
+      "step": 6710
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00015940014800182217,
+      "loss": 1.1706,
+      "step": 6715
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00015932321447260418,
+      "loss": 1.1524,
+      "step": 6720
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00015924622672627104,
+      "loss": 1.1213,
+      "step": 6725
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.000159169184833184,
+      "loss": 1.1586,
+      "step": 6730
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00015909208886375376,
+      "loss": 1.0952,
+      "step": 6735
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00015901493888844037,
+      "loss": 1.1649,
+      "step": 6740
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00015893773497775344,
+      "loss": 1.0904,
+      "step": 6745
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00015886047720225163,
+      "loss": 1.1368,
+      "step": 6750
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00015878316563254301,
+      "loss": 1.1242,
+      "step": 6755
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001587058003392847,
+      "loss": 1.137,
+      "step": 6760
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00015862838139318303,
+      "loss": 1.2024,
+      "step": 6765
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00015855090886499325,
+      "loss": 1.2072,
+      "step": 6770
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00015847338282551964,
+      "loss": 1.2086,
+      "step": 6775
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001583958033456154,
+      "loss": 1.0903,
+      "step": 6780
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001583181704961825,
+      "loss": 1.2337,
+      "step": 6785
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001582404843481718,
+      "loss": 1.2071,
+      "step": 6790
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001581627449725827,
+      "loss": 1.1655,
+      "step": 6795
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00015808495244046345,
+      "loss": 1.1426,
+      "step": 6800
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00015800710682291078,
+      "loss": 1.1939,
+      "step": 6805
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001579292081910699,
+      "loss": 1.1837,
+      "step": 6810
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001578512566161345,
+      "loss": 1.2064,
+      "step": 6815
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00015777325216934668,
+      "loss": 1.1863,
+      "step": 6820
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00015769519492199688,
+      "loss": 1.099,
+      "step": 6825
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00015761708494542372,
+      "loss": 1.1181,
+      "step": 6830
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00015753892231101403,
+      "loss": 1.1693,
+      "step": 6835
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00015746070709020285,
+      "loss": 1.1695,
+      "step": 6840
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00015738243935447318,
+      "loss": 1.0942,
+      "step": 6845
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001573041191753561,
+      "loss": 1.1937,
+      "step": 6850
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001572257466244305,
+      "loss": 1.1738,
+      "step": 6855
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00015714732177332324,
+      "loss": 1.189,
+      "step": 6860
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00015706884469370898,
+      "loss": 1.1427,
+      "step": 6865
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00015699031545731004,
+      "loss": 1.1307,
+      "step": 6870
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00015691173413589647,
+      "loss": 1.1661,
+      "step": 6875
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00015683310080128585,
+      "loss": 1.1778,
+      "step": 6880
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001567544155253434,
+      "loss": 1.2428,
+      "step": 6885
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 0.00015667567837998175,
+      "loss": 1.1149,
+      "step": 6890
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00015659688943716086,
+      "loss": 1.2712,
+      "step": 6895
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001565180487688882,
+      "loss": 1.1571,
+      "step": 6900
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00015643915644721843,
+      "loss": 1.2464,
+      "step": 6905
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00015636021254425333,
+      "loss": 1.1519,
+      "step": 6910
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.000156281217132142,
+      "loss": 1.1258,
+      "step": 6915
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001562021702830804,
+      "loss": 1.154,
+      "step": 6920
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00015612307206931168,
+      "loss": 1.1526,
+      "step": 6925
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001560439225631259,
+      "loss": 1.1815,
+      "step": 6930
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00015596472183685992,
+      "loss": 1.1635,
+      "step": 6935
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00015588546996289749,
+      "loss": 1.237,
+      "step": 6940
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00015580616701366904,
+      "loss": 1.1708,
+      "step": 6945
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00015572681306165174,
+      "loss": 1.1548,
+      "step": 6950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00015564740817936934,
+      "loss": 1.191,
+      "step": 6955
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00015556795243939212,
+      "loss": 1.1753,
+      "step": 6960
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001554884459143369,
+      "loss": 1.1277,
+      "step": 6965
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001554088886768669,
+      "loss": 1.1678,
+      "step": 6970
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00015532928079969158,
+      "loss": 1.2238,
+      "step": 6975
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00015524962235556684,
+      "loss": 1.2124,
+      "step": 6980
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001551699134172947,
+      "loss": 1.1457,
+      "step": 6985
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001550901540577233,
+      "loss": 1.1325,
+      "step": 6990
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00015501034434974695,
+      "loss": 1.1362,
+      "step": 6995
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00015493048436630597,
+      "loss": 1.1841,
+      "step": 7000
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001548505741803865,
+      "loss": 1.1923,
+      "step": 7005
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00015477061386502078,
+      "loss": 1.2032,
+      "step": 7010
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00015469060349328665,
+      "loss": 1.1615,
+      "step": 7015
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00015461054313830778,
+      "loss": 1.1173,
+      "step": 7020
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00015453043287325357,
+      "loss": 1.1551,
+      "step": 7025
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00015445027277133898,
+      "loss": 1.1771,
+      "step": 7030
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00015437006290582448,
+      "loss": 1.1434,
+      "step": 7035
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00015428980335001618,
+      "loss": 1.1717,
+      "step": 7040
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00015420949417726537,
+      "loss": 1.2549,
+      "step": 7045
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00015412913546096889,
+      "loss": 1.2945,
+      "step": 7050
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00015404872727456875,
+      "loss": 1.1672,
+      "step": 7055
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00015396826969155217,
+      "loss": 1.2175,
+      "step": 7060
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00015388776278545162,
+      "loss": 1.1626,
+      "step": 7065
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001538072066298445,
+      "loss": 1.1752,
+      "step": 7070
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00015372660129835337,
+      "loss": 1.1765,
+      "step": 7075
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00015364594686464558,
+      "loss": 1.1739,
+      "step": 7080
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00015356524340243345,
+      "loss": 1.1457,
+      "step": 7085
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001534844909854741,
+      "loss": 1.1647,
+      "step": 7090
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00015340368968756938,
+      "loss": 1.1566,
+      "step": 7095
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00015332283958256583,
+      "loss": 1.233,
+      "step": 7100
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00015324194074435457,
+      "loss": 1.1282,
+      "step": 7105
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001531609932468713,
+      "loss": 1.1809,
+      "step": 7110
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00015307999716409612,
+      "loss": 1.1903,
+      "step": 7115
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00015299895257005357,
+      "loss": 1.2071,
+      "step": 7120
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00015291785953881255,
+      "loss": 1.0904,
+      "step": 7125
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.0001528367181444862,
+      "loss": 1.183,
+      "step": 7130
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001527555284612319,
+      "loss": 1.1049,
+      "step": 7135
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00015267429056325108,
+      "loss": 1.0839,
+      "step": 7140
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00015259300452478934,
+      "loss": 1.1634,
+      "step": 7145
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00015251167042013619,
+      "loss": 1.1184,
+      "step": 7150
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001524302883236251,
+      "loss": 1.1734,
+      "step": 7155
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00015234885830963342,
+      "loss": 1.1759,
+      "step": 7160
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00015226738045258226,
+      "loss": 1.1699,
+      "step": 7165
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001521858548269365,
+      "loss": 1.1255,
+      "step": 7170
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001521042815072046,
+      "loss": 1.2116,
+      "step": 7175
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00015202266056793873,
+      "loss": 1.2304,
+      "step": 7180
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00015194099208373443,
+      "loss": 1.1679,
+      "step": 7185
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001518592761292308,
+      "loss": 1.1427,
+      "step": 7190
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001517775127791103,
+      "loss": 1.1441,
+      "step": 7195
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00015169570210809874,
+      "loss": 1.1659,
+      "step": 7200
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00015161384419096506,
+      "loss": 1.2178,
+      "step": 7205
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00015153193910252152,
+      "loss": 1.1661,
+      "step": 7210
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001514499869176234,
+      "loss": 1.1505,
+      "step": 7215
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00015136798771116908,
+      "loss": 1.1764,
+      "step": 7220
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00015128594155809988,
+      "loss": 1.1145,
+      "step": 7225
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00015120384853340002,
+      "loss": 1.1814,
+      "step": 7230
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00015112170871209654,
+      "loss": 1.1457,
+      "step": 7235
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00015103952216925933,
+      "loss": 1.2408,
+      "step": 7240
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001509572889800009,
+      "loss": 1.11,
+      "step": 7245
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00015087500921947647,
+      "loss": 1.1904,
+      "step": 7250
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001507926829628837,
+      "loss": 1.1597,
+      "step": 7255
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00015071031028546288,
+      "loss": 1.1375,
+      "step": 7260
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00015062789126249658,
+      "loss": 1.2153,
+      "step": 7265
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001505454259693099,
+      "loss": 1.2355,
+      "step": 7270
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001504629144812701,
+      "loss": 1.1851,
+      "step": 7275
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001503803568737867,
+      "loss": 1.1688,
+      "step": 7280
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00015029775322231135,
+      "loss": 1.0928,
+      "step": 7285
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001502151036023378,
+      "loss": 1.2275,
+      "step": 7290
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00015013240808940182,
+      "loss": 1.1867,
+      "step": 7295
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001500496667590811,
+      "loss": 1.2231,
+      "step": 7300
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00014996687968699522,
+      "loss": 1.18,
+      "step": 7305
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00014988404694880555,
+      "loss": 1.1973,
+      "step": 7310
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00014980116862021525,
+      "loss": 1.1616,
+      "step": 7315
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014971824477696903,
+      "loss": 1.2154,
+      "step": 7320
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00014963527549485332,
+      "loss": 1.2213,
+      "step": 7325
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.000149552260849696,
+      "loss": 1.0926,
+      "step": 7330
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00014946920091736643,
+      "loss": 1.1684,
+      "step": 7335
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00014938609577377538,
+      "loss": 1.1545,
+      "step": 7340
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00014930294549487494,
+      "loss": 1.2216,
+      "step": 7345
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014921975015665836,
+      "loss": 1.2078,
+      "step": 7350
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00014913650983516024,
+      "loss": 1.2067,
+      "step": 7355
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00014905322460645614,
+      "loss": 1.123,
+      "step": 7360
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014896989454666275,
+      "loss": 1.1822,
+      "step": 7365
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014888651973193765,
+      "loss": 1.1361,
+      "step": 7370
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00014880310023847941,
+      "loss": 1.224,
+      "step": 7375
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00014871963614252742,
+      "loss": 1.1574,
+      "step": 7380
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00014863612752036176,
+      "loss": 1.1223,
+      "step": 7385
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014855257444830328,
+      "loss": 1.1701,
+      "step": 7390
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001484689770027134,
+      "loss": 1.1517,
+      "step": 7395
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00014838533525999417,
+      "loss": 1.1162,
+      "step": 7400
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00014830164929658806,
+      "loss": 1.1335,
+      "step": 7405
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014821791918897794,
+      "loss": 1.1782,
+      "step": 7410
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00014813414501368708,
+      "loss": 1.1286,
+      "step": 7415
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.000148050326847279,
+      "loss": 1.1696,
+      "step": 7420
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001479664647663574,
+      "loss": 1.0777,
+      "step": 7425
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00014788255884756618,
+      "loss": 1.1785,
+      "step": 7430
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00014779860916758924,
+      "loss": 1.2023,
+      "step": 7435
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00014771461580315049,
+      "loss": 1.1364,
+      "step": 7440
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001476305788310138,
+      "loss": 1.0836,
+      "step": 7445
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00014754649832798283,
+      "loss": 1.1177,
+      "step": 7450
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001474623743709011,
+      "loss": 1.1556,
+      "step": 7455
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00014737820703665178,
+      "loss": 1.1634,
+      "step": 7460
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001472939964021577,
+      "loss": 1.1871,
+      "step": 7465
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00014720974254438132,
+      "loss": 1.209,
+      "step": 7470
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.0001471254455403245,
+      "loss": 1.0775,
+      "step": 7475
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001470411054670286,
+      "loss": 1.2501,
+      "step": 7480
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00014695672240157439,
+      "loss": 1.1184,
+      "step": 7485
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014687229642108176,
+      "loss": 1.149,
+      "step": 7490
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00014678782760271003,
+      "loss": 1.1579,
+      "step": 7495
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00014670331602365753,
+      "loss": 1.1178,
+      "step": 7500
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014661876176116172,
+      "loss": 1.1766,
+      "step": 7505
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00014653416489249905,
+      "loss": 1.2319,
+      "step": 7510
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00014644952549498495,
+      "loss": 1.2514,
+      "step": 7515
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014636484364597367,
+      "loss": 1.1396,
+      "step": 7520
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00014628011942285828,
+      "loss": 1.153,
+      "step": 7525
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00014619535290307057,
+      "loss": 1.276,
+      "step": 7530
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00014611054416408103,
+      "loss": 1.2446,
+      "step": 7535
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00014602569328339864,
+      "loss": 1.2219,
+      "step": 7540
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00014594080033857097,
+      "loss": 1.1463,
+      "step": 7545
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.000145855865407184,
+      "loss": 1.1184,
+      "step": 7550
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014577088856686213,
+      "loss": 1.1214,
+      "step": 7555
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.000145685869895268,
+      "loss": 1.2583,
+      "step": 7560
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00014560080947010256,
+      "loss": 1.2364,
+      "step": 7565
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00014551570736910484,
+      "loss": 1.2237,
+      "step": 7570
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00014543056367005195,
+      "loss": 1.1011,
+      "step": 7575
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001453453784507591,
+      "loss": 1.2022,
+      "step": 7580
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00014526015178907935,
+      "loss": 1.1686,
+      "step": 7585
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00014517488376290376,
+      "loss": 1.1226,
+      "step": 7590
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00014508957445016107,
+      "loss": 1.1691,
+      "step": 7595
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00014500422392881783,
+      "loss": 1.1732,
+      "step": 7600
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001449188322768782,
+      "loss": 1.1826,
+      "step": 7605
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00014483339957238397,
+      "loss": 1.2111,
+      "step": 7610
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00014474792589341443,
+      "loss": 1.1708,
+      "step": 7615
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014466241131808635,
+      "loss": 1.1544,
+      "step": 7620
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001445768559245538,
+      "loss": 1.1789,
+      "step": 7625
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001444912597910083,
+      "loss": 1.1848,
+      "step": 7630
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00014440562299567836,
+      "loss": 1.1384,
+      "step": 7635
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001443199456168299,
+      "loss": 1.1512,
+      "step": 7640
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001442342277327658,
+      "loss": 1.1939,
+      "step": 7645
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.000144148469421826,
+      "loss": 1.1397,
+      "step": 7650
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00014406267076238737,
+      "loss": 1.2094,
+      "step": 7655
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001439768318328637,
+      "loss": 1.1209,
+      "step": 7660
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00014389095271170549,
+      "loss": 1.2254,
+      "step": 7665
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00014380503347740005,
+      "loss": 1.1778,
+      "step": 7670
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00014371907420847132,
+      "loss": 1.1528,
+      "step": 7675
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014363307498347985,
+      "loss": 1.1539,
+      "step": 7680
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00014354703588102268,
+      "loss": 1.1372,
+      "step": 7685
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.0001434609569797333,
+      "loss": 1.1385,
+      "step": 7690
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00014337483835828163,
+      "loss": 1.2349,
+      "step": 7695
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014328868009537378,
+      "loss": 1.1939,
+      "step": 7700
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00014320248226975223,
+      "loss": 1.1573,
+      "step": 7705
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00014311624496019546,
+      "loss": 1.283,
+      "step": 7710
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00014302996824551813,
+      "loss": 1.2353,
+      "step": 7715
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00014294365220457093,
+      "loss": 1.1532,
+      "step": 7720
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001428572969162405,
+      "loss": 1.1465,
+      "step": 7725
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001427709024594492,
+      "loss": 1.1481,
+      "step": 7730
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00014268446891315537,
+      "loss": 1.1494,
+      "step": 7735
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014259799635635298,
+      "loss": 1.1859,
+      "step": 7740
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00014251148486807172,
+      "loss": 1.1282,
+      "step": 7745
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014242493452737675,
+      "loss": 1.2092,
+      "step": 7750
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00014233834541336888,
+      "loss": 1.1788,
+      "step": 7755
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00014225171760518415,
+      "loss": 1.1056,
+      "step": 7760
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00014216505118199425,
+      "loss": 1.176,
+      "step": 7765
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001420783462230059,
+      "loss": 1.1368,
+      "step": 7770
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001419916028074612,
+      "loss": 1.1946,
+      "step": 7775
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001419048210146373,
+      "loss": 1.1036,
+      "step": 7780
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014181800092384647,
+      "loss": 1.2529,
+      "step": 7785
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.000141731142614436,
+      "loss": 1.1413,
+      "step": 7790
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00014164424616578805,
+      "loss": 1.1956,
+      "step": 7795
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014155731165731972,
+      "loss": 1.1081,
+      "step": 7800
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00014147033916848275,
+      "loss": 1.2275,
+      "step": 7805
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00014138332877876381,
+      "loss": 1.2397,
+      "step": 7810
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.000141296280567684,
+      "loss": 1.229,
+      "step": 7815
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001412091946147991,
+      "loss": 1.1537,
+      "step": 7820
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014112207099969937,
+      "loss": 1.1069,
+      "step": 7825
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00014103490980200945,
+      "loss": 1.2241,
+      "step": 7830
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00014094771110138835,
+      "loss": 1.2112,
+      "step": 7835
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00014086047497752937,
+      "loss": 1.1526,
+      "step": 7840
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00014077320151015997,
+      "loss": 1.1632,
+      "step": 7845
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00014068589077904185,
+      "loss": 1.1236,
+      "step": 7850
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00014059854286397057,
+      "loss": 1.1998,
+      "step": 7855
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00014051115784477594,
+      "loss": 1.2108,
+      "step": 7860
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00014042373580132136,
+      "loss": 1.1749,
+      "step": 7865
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00014033627681350435,
+      "loss": 1.173,
+      "step": 7870
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 0.00014024878096125605,
+      "loss": 1.1113,
+      "step": 7875
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001401612483245413,
+      "loss": 1.214,
+      "step": 7880
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001400736789833586,
+      "loss": 1.1316,
+      "step": 7885
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00013998607301773998,
+      "loss": 1.1805,
+      "step": 7890
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.00013989843050775096,
+      "loss": 1.096,
+      "step": 7895
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00013981075153349037,
+      "loss": 1.2747,
+      "step": 7900
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001397230361750905,
+      "loss": 1.1922,
+      "step": 7905
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00013963528451271679,
+      "loss": 1.2204,
+      "step": 7910
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001395474966265679,
+      "loss": 1.2171,
+      "step": 7915
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00013945967259687558,
+      "loss": 1.1588,
+      "step": 7920
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00013937181250390468,
+      "loss": 1.2271,
+      "step": 7925
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001392839164279529,
+      "loss": 1.2605,
+      "step": 7930
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00013919598444935087,
+      "loss": 1.1003,
+      "step": 7935
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00013910801664846208,
+      "loss": 1.221,
+      "step": 7940
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001390200131056827,
+      "loss": 1.1455,
+      "step": 7945
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00013893197390144155,
+      "loss": 1.1203,
+      "step": 7950
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013884389911620012,
+      "loss": 1.1516,
+      "step": 7955
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00013875578883045238,
+      "loss": 1.1548,
+      "step": 7960
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00013866764312472473,
+      "loss": 1.1942,
+      "step": 7965
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00013857946207957594,
+      "loss": 1.1953,
+      "step": 7970
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001384912457755971,
+      "loss": 1.0613,
+      "step": 7975
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00013840299429341153,
+      "loss": 1.1485,
+      "step": 7980
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00013831470771367464,
+      "loss": 1.2536,
+      "step": 7985
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00013822638611707397,
+      "loss": 1.1073,
+      "step": 7990
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001381380295843291,
+      "loss": 1.1628,
+      "step": 7995
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001380496381961914,
+      "loss": 1.1307,
+      "step": 8000
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001379612120334443,
+      "loss": 1.2052,
+      "step": 8005
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001378727511769028,
+      "loss": 1.1994,
+      "step": 8010
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00013778425570741377,
+      "loss": 1.1662,
+      "step": 8015
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00013769572570585567,
+      "loss": 1.1763,
+      "step": 8020
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001376071612531384,
+      "loss": 1.1969,
+      "step": 8025
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00013751856243020359,
+      "loss": 1.1872,
+      "step": 8030
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00013742992931802398,
+      "loss": 1.1696,
+      "step": 8035
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00013734126199760392,
+      "loss": 1.1599,
+      "step": 8040
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00013725256054997886,
+      "loss": 1.1414,
+      "step": 8045
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001371638250562155,
+      "loss": 1.1845,
+      "step": 8050
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00013707505559741164,
+      "loss": 1.2117,
+      "step": 8055
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00013698625225469615,
+      "loss": 1.154,
+      "step": 8060
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001368974151092288,
+      "loss": 1.1619,
+      "step": 8065
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00013680854424220036,
+      "loss": 1.1804,
+      "step": 8070
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001367196397348323,
+      "loss": 1.2434,
+      "step": 8075
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001366307016683769,
+      "loss": 1.1144,
+      "step": 8080
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00013654173012411712,
+      "loss": 1.1867,
+      "step": 8085
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001364527251833665,
+      "loss": 1.1751,
+      "step": 8090
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00013636368692746906,
+      "loss": 1.1101,
+      "step": 8095
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00013627461543779936,
+      "loss": 1.1906,
+      "step": 8100
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00013618551079576228,
+      "loss": 1.1638,
+      "step": 8105
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013609637308279292,
+      "loss": 1.2165,
+      "step": 8110
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00013600720238035683,
+      "loss": 1.1542,
+      "step": 8115
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.0001359179987699494,
+      "loss": 1.1143,
+      "step": 8120
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001358287623330964,
+      "loss": 1.1966,
+      "step": 8125
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00013573949315135334,
+      "loss": 1.1892,
+      "step": 8130
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00013565019130630588,
+      "loss": 1.2578,
+      "step": 8135
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00013556085687956934,
+      "loss": 1.0711,
+      "step": 8140
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.000135471489952789,
+      "loss": 1.1693,
+      "step": 8145
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00013538209060763966,
+      "loss": 1.1988,
+      "step": 8150
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001352926589258259,
+      "loss": 1.1489,
+      "step": 8155
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00013520319498908177,
+      "loss": 1.1932,
+      "step": 8160
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001351136988791708,
+      "loss": 1.2381,
+      "step": 8165
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00013502417067788594,
+      "loss": 1.0664,
+      "step": 8170
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 0.00013493461046704954,
+      "loss": 1.1675,
+      "step": 8175
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00013484501832851305,
+      "loss": 1.2074,
+      "step": 8180
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00013475539434415726,
+      "loss": 1.1126,
+      "step": 8185
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001346657385958919,
+      "loss": 1.2108,
+      "step": 8190
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00013457605116565593,
+      "loss": 1.0783,
+      "step": 8195
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00013448633213541708,
+      "loss": 1.2537,
+      "step": 8200
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00013439658158717206,
+      "loss": 1.2103,
+      "step": 8205
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013430679960294635,
+      "loss": 1.2296,
+      "step": 8210
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001342169862647942,
+      "loss": 1.1953,
+      "step": 8215
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00013412714165479846,
+      "loss": 1.1904,
+      "step": 8220
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001340372658550706,
+      "loss": 1.154,
+      "step": 8225
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00013394735894775057,
+      "loss": 1.2235,
+      "step": 8230
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013385742101500677,
+      "loss": 1.2637,
+      "step": 8235
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00013376745213903592,
+      "loss": 1.1311,
+      "step": 8240
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.0001336774524020631,
+      "loss": 1.1393,
+      "step": 8245
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001335874218863415,
+      "loss": 1.1448,
+      "step": 8250
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00013349736067415252,
+      "loss": 1.2296,
+      "step": 8255
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00013340726884780553,
+      "loss": 1.178,
+      "step": 8260
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013331714648963792,
+      "loss": 1.1577,
+      "step": 8265
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00013322699368201508,
+      "loss": 1.1838,
+      "step": 8270
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00013313681050733001,
+      "loss": 1.1692,
+      "step": 8275
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00013304659704800367,
+      "loss": 1.2479,
+      "step": 8280
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00013295635338648462,
+      "loss": 1.2184,
+      "step": 8285
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00013286607960524897,
+      "loss": 1.2417,
+      "step": 8290
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00013277577578680048,
+      "loss": 1.1967,
+      "step": 8295
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001326854420136702,
+      "loss": 1.1231,
+      "step": 8300
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001325950783684167,
+      "loss": 1.185,
+      "step": 8305
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00013250468493362583,
+      "loss": 1.1716,
+      "step": 8310
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013241426179191057,
+      "loss": 1.1983,
+      "step": 8315
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00013232380902591115,
+      "loss": 1.1448,
+      "step": 8320
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00013223332671829483,
+      "loss": 1.271,
+      "step": 8325
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00013214281495175584,
+      "loss": 1.2046,
+      "step": 8330
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00013205227380901543,
+      "loss": 1.1249,
+      "step": 8335
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00013196170337282162,
+      "loss": 1.1973,
+      "step": 8340
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.0001318711037259492,
+      "loss": 1.0632,
+      "step": 8345
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00013178047495119964,
+      "loss": 1.1831,
+      "step": 8350
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.185546875,
+      "learning_rate": 0.00013168981713140116,
+      "loss": 1.1754,
+      "step": 8355
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00013159913034940836,
+      "loss": 1.142,
+      "step": 8360
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00013150841468810242,
+      "loss": 1.2121,
+      "step": 8365
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00013141767023039085,
+      "loss": 1.1684,
+      "step": 8370
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00013132689705920748,
+      "loss": 1.1947,
+      "step": 8375
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00013123609525751248,
+      "loss": 1.175,
+      "step": 8380
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00013114526490829204,
+      "loss": 1.1841,
+      "step": 8385
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00013105440609455856,
+      "loss": 1.2425,
+      "step": 8390
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013096351889935033,
+      "loss": 1.2554,
+      "step": 8395
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00013087260340573174,
+      "loss": 1.1532,
+      "step": 8400
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00013078165969679291,
+      "loss": 1.1028,
+      "step": 8405
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00013069068785564984,
+      "loss": 1.1521,
+      "step": 8410
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00013059968796544412,
+      "loss": 1.1432,
+      "step": 8415
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00013050866010934313,
+      "loss": 1.2014,
+      "step": 8420
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00013041760437053964,
+      "loss": 1.1032,
+      "step": 8425
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00013032652083225208,
+      "loss": 1.208,
+      "step": 8430
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00013023540957772414,
+      "loss": 1.1257,
+      "step": 8435
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001301442706902249,
+      "loss": 1.1944,
+      "step": 8440
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001300531042530487,
+      "loss": 1.228,
+      "step": 8445
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00012996191034951512,
+      "loss": 1.1473,
+      "step": 8450
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00012987068906296863,
+      "loss": 1.2271,
+      "step": 8455
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00012977944047677897,
+      "loss": 1.238,
+      "step": 8460
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00012968816467434074,
+      "loss": 1.1768,
+      "step": 8465
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00012959686173907333,
+      "loss": 1.1546,
+      "step": 8470
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00012950553175442107,
+      "loss": 1.2455,
+      "step": 8475
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00012941417480385293,
+      "loss": 1.2709,
+      "step": 8480
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00012932279097086248,
+      "loss": 1.1166,
+      "step": 8485
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00012923138033896797,
+      "loss": 1.1345,
+      "step": 8490
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00012913994299171207,
+      "loss": 1.1113,
+      "step": 8495
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00012904847901266188,
+      "loss": 1.2091,
+      "step": 8500
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.0001289569884854089,
+      "loss": 1.214,
+      "step": 8505
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00012886547149356872,
+      "loss": 1.1845,
+      "step": 8510
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00012877392812078134,
+      "loss": 1.232,
+      "step": 8515
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00012868235845071068,
+      "loss": 1.1696,
+      "step": 8520
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.00012859076256704484,
+      "loss": 1.1541,
+      "step": 8525
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00012849914055349578,
+      "loss": 1.1485,
+      "step": 8530
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00012840749249379938,
+      "loss": 1.116,
+      "step": 8535
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00012831581847171534,
+      "loss": 1.1878,
+      "step": 8540
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00012822411857102702,
+      "loss": 1.083,
+      "step": 8545
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00012813239287554155,
+      "loss": 1.2502,
+      "step": 8550
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00012804064146908946,
+      "loss": 1.1709,
+      "step": 8555
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.000127948864435525,
+      "loss": 1.1941,
+      "step": 8560
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00012785706185872565,
+      "loss": 1.2039,
+      "step": 8565
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00012776523382259232,
+      "loss": 1.2117,
+      "step": 8570
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00012767338041104917,
+      "loss": 1.1662,
+      "step": 8575
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00012758150170804358,
+      "loss": 1.0912,
+      "step": 8580
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00012748959779754597,
+      "loss": 1.2631,
+      "step": 8585
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001273976687635499,
+      "loss": 1.154,
+      "step": 8590
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001273057146900718,
+      "loss": 1.1527,
+      "step": 8595
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00012721373566115105,
+      "loss": 1.1861,
+      "step": 8600
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00012712173176084978,
+      "loss": 1.1469,
+      "step": 8605
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00012702970307325293,
+      "loss": 1.2622,
+      "step": 8610
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.000126937649682468,
+      "loss": 1.198,
+      "step": 8615
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00012684557167262517,
+      "loss": 1.1417,
+      "step": 8620
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.000126753469127877,
+      "loss": 1.2068,
+      "step": 8625
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00012666134213239853,
+      "loss": 1.2055,
+      "step": 8630
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001265691907703872,
+      "loss": 1.0553,
+      "step": 8635
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00012647701512606262,
+      "loss": 1.082,
+      "step": 8640
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00012638481528366675,
+      "loss": 1.186,
+      "step": 8645
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00012629259132746344,
+      "loss": 1.199,
+      "step": 8650
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00012620034334173873,
+      "loss": 1.1182,
+      "step": 8655
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00012610807141080058,
+      "loss": 1.1438,
+      "step": 8660
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001260157756189789,
+      "loss": 1.1269,
+      "step": 8665
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001259234560506253,
+      "loss": 1.202,
+      "step": 8670
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00012583111279011318,
+      "loss": 1.1333,
+      "step": 8675
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00012573874592183756,
+      "loss": 1.213,
+      "step": 8680
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00012564635553021511,
+      "loss": 1.1436,
+      "step": 8685
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001255539416996839,
+      "loss": 1.1571,
+      "step": 8690
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00012546150451470348,
+      "loss": 1.2428,
+      "step": 8695
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00012536904405975472,
+      "loss": 1.2165,
+      "step": 8700
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012527656041933983,
+      "loss": 1.1764,
+      "step": 8705
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00012518405367798206,
+      "loss": 1.1994,
+      "step": 8710
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00012509152392022591,
+      "loss": 1.2052,
+      "step": 8715
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00012499897123063687,
+      "loss": 1.1536,
+      "step": 8720
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001249063956938014,
+      "loss": 1.1935,
+      "step": 8725
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00012481379739432674,
+      "loss": 1.1697,
+      "step": 8730
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00012472117641684112,
+      "loss": 1.1128,
+      "step": 8735
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00012462853284599335,
+      "loss": 1.1837,
+      "step": 8740
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00012453586676645296,
+      "loss": 1.1359,
+      "step": 8745
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00012444317826290998,
+      "loss": 1.1335,
+      "step": 8750
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012435046742007504,
+      "loss": 1.1713,
+      "step": 8755
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001242577343226791,
+      "loss": 1.0868,
+      "step": 8760
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00012416497905547345,
+      "loss": 1.2478,
+      "step": 8765
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001240722017032297,
+      "loss": 1.1887,
+      "step": 8770
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00012397940235073964,
+      "loss": 1.1038,
+      "step": 8775
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001238865810828151,
+      "loss": 1.1008,
+      "step": 8780
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00012379373798428807,
+      "loss": 1.1604,
+      "step": 8785
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00012370087314001035,
+      "loss": 1.1985,
+      "step": 8790
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00012360798663485365,
+      "loss": 1.1248,
+      "step": 8795
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012351507855370953,
+      "loss": 1.1265,
+      "step": 8800
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00012342214898148924,
+      "loss": 1.1833,
+      "step": 8805
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00012332919800312365,
+      "loss": 1.2389,
+      "step": 8810
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012323622570356324,
+      "loss": 1.1159,
+      "step": 8815
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001231432321677779,
+      "loss": 1.1641,
+      "step": 8820
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00012305021748075705,
+      "loss": 1.1587,
+      "step": 8825
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.15625,
+      "learning_rate": 0.0001229571817275093,
+      "loss": 1.1954,
+      "step": 8830
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001228641249930626,
+      "loss": 1.1489,
+      "step": 8835
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00012277104736246408,
+      "loss": 1.095,
+      "step": 8840
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00012267794892077992,
+      "loss": 1.1478,
+      "step": 8845
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00012258482975309535,
+      "loss": 1.1428,
+      "step": 8850
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00012249168994451451,
+      "loss": 1.1707,
+      "step": 8855
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.0001223985295801605,
+      "loss": 1.1852,
+      "step": 8860
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00012230534874517506,
+      "loss": 1.1895,
+      "step": 8865
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00012221214752471875,
+      "loss": 1.2153,
+      "step": 8870
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001221189260039707,
+      "loss": 1.1326,
+      "step": 8875
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00012202568426812866,
+      "loss": 1.1269,
+      "step": 8880
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00012193242240240878,
+      "loss": 1.2114,
+      "step": 8885
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00012183914049204563,
+      "loss": 1.2026,
+      "step": 8890
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00012174583862229217,
+      "loss": 1.0864,
+      "step": 8895
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00012165251687841948,
+      "loss": 1.1961,
+      "step": 8900
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012155917534571688,
+      "loss": 1.231,
+      "step": 8905
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00012146581410949176,
+      "loss": 1.1196,
+      "step": 8910
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.181640625,
+      "learning_rate": 0.00012137243325506955,
+      "loss": 1.1796,
+      "step": 8915
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00012127903286779352,
+      "loss": 1.1415,
+      "step": 8920
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00012118561303302491,
+      "loss": 1.1687,
+      "step": 8925
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00012109217383614264,
+      "loss": 1.1745,
+      "step": 8930
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00012099871536254337,
+      "loss": 1.1503,
+      "step": 8935
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00012090523769764135,
+      "loss": 1.1252,
+      "step": 8940
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.158203125,
+      "learning_rate": 0.00012081174092686842,
+      "loss": 1.0908,
+      "step": 8945
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001207182251356738,
+      "loss": 1.1352,
+      "step": 8950
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00012062469040952415,
+      "loss": 1.1196,
+      "step": 8955
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00012053113683390346,
+      "loss": 1.2844,
+      "step": 8960
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00012043756449431289,
+      "loss": 1.2494,
+      "step": 8965
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00012034397347627075,
+      "loss": 1.2029,
+      "step": 8970
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001202503638653125,
+      "loss": 1.1702,
+      "step": 8975
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00012015673574699047,
+      "loss": 1.1436,
+      "step": 8980
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00012006308920687402,
+      "loss": 1.1993,
+      "step": 8985
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00011996942433054927,
+      "loss": 1.0998,
+      "step": 8990
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00011987574120361911,
+      "loss": 1.112,
+      "step": 8995
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00011978203991170313,
+      "loss": 1.1976,
+      "step": 9000
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00011968832054043755,
+      "loss": 1.1803,
+      "step": 9005
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00011959458317547503,
+      "loss": 1.1088,
+      "step": 9010
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00011950082790248473,
+      "loss": 1.1638,
+      "step": 9015
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00011940705480715222,
+      "loss": 1.1852,
+      "step": 9020
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011931326397517924,
+      "loss": 1.1152,
+      "step": 9025
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00011921945549228386,
+      "loss": 1.1023,
+      "step": 9030
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00011912562944420018,
+      "loss": 1.191,
+      "step": 9035
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011903178591667845,
+      "loss": 1.1539,
+      "step": 9040
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00011893792499548482,
+      "loss": 1.1156,
+      "step": 9045
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00011884404676640137,
+      "loss": 1.2176,
+      "step": 9050
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 0.000118750151315226,
+      "loss": 1.2067,
+      "step": 9055
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011865623872777232,
+      "loss": 1.1671,
+      "step": 9060
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00011856230908986965,
+      "loss": 1.1284,
+      "step": 9065
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00011846836248736286,
+      "loss": 1.1997,
+      "step": 9070
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00011837439900611229,
+      "loss": 1.1906,
+      "step": 9075
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011828041873199382,
+      "loss": 1.1339,
+      "step": 9080
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00011818642175089855,
+      "loss": 1.2474,
+      "step": 9085
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.0001180924081487329,
+      "loss": 1.1568,
+      "step": 9090
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001179983780114185,
+      "loss": 1.1465,
+      "step": 9095
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00011790433142489208,
+      "loss": 1.2031,
+      "step": 9100
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00011781026847510541,
+      "loss": 1.1925,
+      "step": 9105
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00011771618924802516,
+      "loss": 1.158,
+      "step": 9110
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00011762209382963298,
+      "loss": 1.2164,
+      "step": 9115
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00011752798230592521,
+      "loss": 1.1389,
+      "step": 9120
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00011743385476291298,
+      "loss": 1.137,
+      "step": 9125
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00011733971128662202,
+      "loss": 1.1422,
+      "step": 9130
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00011724555196309266,
+      "loss": 1.172,
+      "step": 9135
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00011715137687837965,
+      "loss": 1.1292,
+      "step": 9140
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00011705718611855223,
+      "loss": 1.164,
+      "step": 9145
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001169629797696939,
+      "loss": 1.1617,
+      "step": 9150
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011686875791790246,
+      "loss": 1.1447,
+      "step": 9155
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.0001167745206492898,
+      "loss": 1.1206,
+      "step": 9160
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.000116680268049982,
+      "loss": 1.0839,
+      "step": 9165
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00011658600020611906,
+      "loss": 1.1799,
+      "step": 9170
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00011649171720385496,
+      "loss": 1.1548,
+      "step": 9175
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00011639741912935753,
+      "loss": 1.1429,
+      "step": 9180
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00011630310606880839,
+      "loss": 1.1355,
+      "step": 9185
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001162087781084028,
+      "loss": 1.1415,
+      "step": 9190
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001161144353343497,
+      "loss": 1.1226,
+      "step": 9195
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011602007783287153,
+      "loss": 1.1547,
+      "step": 9200
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001159257056902042,
+      "loss": 1.277,
+      "step": 9205
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 0.00011583131899259702,
+      "loss": 1.1556,
+      "step": 9210
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00011573691782631256,
+      "loss": 1.1532,
+      "step": 9215
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00011564250227762668,
+      "loss": 1.1602,
+      "step": 9220
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00011554807243282828,
+      "loss": 1.1712,
+      "step": 9225
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011545362837821947,
+      "loss": 1.1379,
+      "step": 9230
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.0001153591702001152,
+      "loss": 1.2032,
+      "step": 9235
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00011526469798484341,
+      "loss": 1.1966,
+      "step": 9240
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00011517021181874486,
+      "loss": 1.1637,
+      "step": 9245
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00011507571178817307,
+      "loss": 1.1833,
+      "step": 9250
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00011498119797949416,
+      "loss": 1.1807,
+      "step": 9255
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00011488667047908695,
+      "loss": 1.0899,
+      "step": 9260
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00011479212937334267,
+      "loss": 1.1128,
+      "step": 9265
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00011469757474866511,
+      "loss": 1.2238,
+      "step": 9270
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00011460300669147024,
+      "loss": 1.1462,
+      "step": 9275
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00011450842528818649,
+      "loss": 1.2737,
+      "step": 9280
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00011441383062525435,
+      "loss": 1.1189,
+      "step": 9285
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.0001143192227891265,
+      "loss": 1.1053,
+      "step": 9290
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00011422460186626765,
+      "loss": 1.2605,
+      "step": 9295
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011412996794315445,
+      "loss": 1.1205,
+      "step": 9300
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011403532110627541,
+      "loss": 1.1824,
+      "step": 9305
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00011394066144213094,
+      "loss": 1.1099,
+      "step": 9310
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011384598903723307,
+      "loss": 1.0561,
+      "step": 9315
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00011375130397810547,
+      "loss": 1.1908,
+      "step": 9320
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001136566063512835,
+      "loss": 1.1312,
+      "step": 9325
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00011356189624331386,
+      "loss": 1.1794,
+      "step": 9330
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011346717374075478,
+      "loss": 1.1816,
+      "step": 9335
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011337243893017566,
+      "loss": 1.229,
+      "step": 9340
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011327769189815733,
+      "loss": 1.1432,
+      "step": 9345
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00011318293273129166,
+      "loss": 1.1915,
+      "step": 9350
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00011308816151618171,
+      "loss": 1.1272,
+      "step": 9355
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00011299337833944141,
+      "loss": 1.2189,
+      "step": 9360
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.0001128985832876958,
+      "loss": 1.1982,
+      "step": 9365
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00011280377644758063,
+      "loss": 1.0604,
+      "step": 9370
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00011270895790574246,
+      "loss": 1.2464,
+      "step": 9375
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00011261412774883857,
+      "loss": 1.149,
+      "step": 9380
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001125192860635369,
+      "loss": 1.1785,
+      "step": 9385
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00011242443293651583,
+      "loss": 1.1137,
+      "step": 9390
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001123295684544642,
+      "loss": 1.1899,
+      "step": 9395
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011223469270408136,
+      "loss": 1.1728,
+      "step": 9400
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00011213980577207677,
+      "loss": 1.1643,
+      "step": 9405
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011204490774517026,
+      "loss": 1.1044,
+      "step": 9410
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00011194999871009171,
+      "loss": 1.199,
+      "step": 9415
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00011185507875358112,
+      "loss": 1.1379,
+      "step": 9420
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011176014796238842,
+      "loss": 1.2192,
+      "step": 9425
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00011166520642327352,
+      "loss": 1.1892,
+      "step": 9430
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011157025422300603,
+      "loss": 1.1677,
+      "step": 9435
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00011147529144836541,
+      "loss": 1.2184,
+      "step": 9440
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00011138031818614072,
+      "loss": 1.1363,
+      "step": 9445
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001112853345231307,
+      "loss": 1.2203,
+      "step": 9450
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00011119034054614346,
+      "loss": 1.1162,
+      "step": 9455
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00011109533634199662,
+      "loss": 1.0821,
+      "step": 9460
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00011100032199751718,
+      "loss": 1.2297,
+      "step": 9465
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00011090529759954129,
+      "loss": 1.1898,
+      "step": 9470
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.0001108102632349144,
+      "loss": 1.2135,
+      "step": 9475
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00011071521899049102,
+      "loss": 1.2227,
+      "step": 9480
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.0001106201649531347,
+      "loss": 1.1815,
+      "step": 9485
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00011052510120971791,
+      "loss": 1.2157,
+      "step": 9490
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00011043002784712204,
+      "loss": 1.1603,
+      "step": 9495
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00011033494495223724,
+      "loss": 1.2134,
+      "step": 9500
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001102398526119624,
+      "loss": 1.1588,
+      "step": 9505
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00011014475091320496,
+      "loss": 1.1676,
+      "step": 9510
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00011004963994288102,
+      "loss": 1.2596,
+      "step": 9515
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00010995451978791511,
+      "loss": 1.2051,
+      "step": 9520
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010985939053524017,
+      "loss": 1.1736,
+      "step": 9525
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00010976425227179738,
+      "loss": 1.2479,
+      "step": 9530
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010966910508453627,
+      "loss": 1.1076,
+      "step": 9535
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00010957394906041442,
+      "loss": 1.1039,
+      "step": 9540
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00010947878428639757,
+      "loss": 1.2402,
+      "step": 9545
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001093836108494594,
+      "loss": 1.1539,
+      "step": 9550
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00010928842883658152,
+      "loss": 1.2417,
+      "step": 9555
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010919323833475343,
+      "loss": 1.1806,
+      "step": 9560
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00010909803943097233,
+      "loss": 1.1364,
+      "step": 9565
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00010900283221224306,
+      "loss": 1.2511,
+      "step": 9570
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00010890761676557816,
+      "loss": 1.1083,
+      "step": 9575
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.17578125,
+      "learning_rate": 0.00010881239317799765,
+      "loss": 1.0718,
+      "step": 9580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00010871716153652895,
+      "loss": 1.1785,
+      "step": 9585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010862192192820693,
+      "loss": 1.1352,
+      "step": 9590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00010852667444007358,
+      "loss": 1.1345,
+      "step": 9595
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00010843141915917829,
+      "loss": 1.1186,
+      "step": 9600
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00010833615617257743,
+      "loss": 1.1213,
+      "step": 9605
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00010824088556733452,
+      "loss": 1.1469,
+      "step": 9610
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001081456074305199,
+      "loss": 1.1345,
+      "step": 9615
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010805032184921091,
+      "loss": 1.1589,
+      "step": 9620
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00010795502891049167,
+      "loss": 1.1469,
+      "step": 9625
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00010785972870145302,
+      "loss": 1.1811,
+      "step": 9630
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00010776442130919241,
+      "loss": 1.1993,
+      "step": 9635
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.0001076691068208139,
+      "loss": 1.1724,
+      "step": 9640
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00010757378532342803,
+      "loss": 1.1881,
+      "step": 9645
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00010747845690415172,
+      "loss": 1.1064,
+      "step": 9650
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00010738312165010824,
+      "loss": 1.1816,
+      "step": 9655
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.0001072877796484271,
+      "loss": 1.1331,
+      "step": 9660
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00010719243098624394,
+      "loss": 1.215,
+      "step": 9665
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 0.00010709707575070059,
+      "loss": 1.1443,
+      "step": 9670
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00010700171402894472,
+      "loss": 1.0988,
+      "step": 9675
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00010690634590813014,
+      "loss": 1.157,
+      "step": 9680
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.0001068109714754163,
+      "loss": 1.1724,
+      "step": 9685
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00010671559081796857,
+      "loss": 1.1435,
+      "step": 9690
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00010662020402295796,
+      "loss": 1.1928,
+      "step": 9695
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.000106524811177561,
+      "loss": 1.1886,
+      "step": 9700
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 0.00010642941236895994,
+      "loss": 1.1617,
+      "step": 9705
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001063340076843423,
+      "loss": 1.1931,
+      "step": 9710
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 0.00010623859721090106,
+      "loss": 1.2535,
+      "step": 9715
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00010614318103583445,
+      "loss": 1.1818,
+      "step": 9720
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.000106047759246346,
+      "loss": 1.1468,
+      "step": 9725
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00010595233192964422,
+      "loss": 1.194,
+      "step": 9730
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00010585689917294282,
+      "loss": 1.1362,
+      "step": 9735
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00010576146106346037,
+      "loss": 1.178,
+      "step": 9740
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00010566601768842043,
+      "loss": 1.1273,
+      "step": 9745
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.0001055705691350513,
+      "loss": 1.1585,
+      "step": 9750
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010547511549058604,
+      "loss": 1.2623,
+      "step": 9755
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00010537965684226233,
+      "loss": 1.1143,
+      "step": 9760
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00010528419327732247,
+      "loss": 1.1588,
+      "step": 9765
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00010518872488301323,
+      "loss": 1.0327,
+      "step": 9770
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 0.00010509325174658582,
+      "loss": 1.0866,
+      "step": 9775
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00010499777395529571,
+      "loss": 1.1541,
+      "step": 9780
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00010490229159640275,
+      "loss": 1.1544,
+      "step": 9785
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00010480680475717084,
+      "loss": 1.1667,
+      "step": 9790
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00010471131352486801,
+      "loss": 1.137,
+      "step": 9795
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00010461581798676636,
+      "loss": 1.171,
+      "step": 9800
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00010452031823014186,
+      "loss": 1.1982,
+      "step": 9805
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00010442481434227437,
+      "loss": 1.1579,
+      "step": 9810
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010432930641044754,
+      "loss": 1.1684,
+      "step": 9815
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00010423379452194863,
+      "loss": 1.1205,
+      "step": 9820
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00010413827876406866,
+      "loss": 1.2203,
+      "step": 9825
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00010404275922410202,
+      "loss": 1.2415,
+      "step": 9830
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00010394723598934668,
+      "loss": 1.1107,
+      "step": 9835
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.000103851709147104,
+      "loss": 1.1226,
+      "step": 9840
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.0001037561787846785,
+      "loss": 1.1235,
+      "step": 9845
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 0.00010366064498937806,
+      "loss": 1.1031,
+      "step": 9850
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.00010356510784851358,
+      "loss": 1.1934,
+      "step": 9855
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.16796875,
+      "learning_rate": 0.00010346956744939918,
+      "loss": 1.1664,
+      "step": 9860
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010337402387935173,
+      "loss": 1.2113,
+      "step": 9865
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00010327847722569121,
+      "loss": 1.1374,
+      "step": 9870
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00010318292757574029,
+      "loss": 1.1662,
+      "step": 9875
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00010308737501682444,
+      "loss": 1.2542,
+      "step": 9880
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00010299181963627175,
+      "loss": 1.1446,
+      "step": 9885
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010289626152141294,
+      "loss": 1.1309,
+      "step": 9890
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 0.00010280070075958119,
+      "loss": 1.2414,
+      "step": 9895
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010270513743811205,
+      "loss": 1.1768,
+      "step": 9900
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 0.00010260957164434352,
+      "loss": 1.1825,
+      "step": 9905
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 0.00010251400346561578,
+      "loss": 1.1436,
+      "step": 9910
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00010241843298927121,
+      "loss": 1.0725,
+      "step": 9915
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.0001023228603026543,
+      "loss": 1.136,
+      "step": 9920
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.0001022272854931116,
+      "loss": 1.1164,
+      "step": 9925
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1640625,
+      "learning_rate": 0.00010213170864799146,
+      "loss": 1.1769,
+      "step": 9930
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00010203612985464429,
+      "loss": 1.2069,
+      "step": 9935
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.162109375,
+      "learning_rate": 0.00010194054920042208,
+      "loss": 1.1555,
+      "step": 9940
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010184496677267867,
+      "loss": 1.1883,
+      "step": 9945
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.00010174938265876945,
+      "loss": 1.1809,
+      "step": 9950
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010165379694605137,
+      "loss": 1.2525,
+      "step": 9955
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010155820972188284,
+      "loss": 1.1263,
+      "step": 9960
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00010146262107362363,
+      "loss": 1.1874,
+      "step": 9965
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 0.00010136703108863486,
+      "loss": 1.2121,
+      "step": 9970
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00010127143985427883,
+      "loss": 1.1992,
+      "step": 9975
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2109375,
+      "learning_rate": 0.00010117584745791898,
+      "loss": 1.166,
+      "step": 9980
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.166015625,
+      "learning_rate": 0.00010108025398691988,
+      "loss": 1.1574,
+      "step": 9985
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.173828125,
+      "learning_rate": 0.00010098465952864693,
+      "loss": 1.1448,
+      "step": 9990
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.171875,
+      "learning_rate": 0.00010088906417046665,
+      "loss": 1.1529,
+      "step": 9995
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00010079346799974616,
+      "loss": 1.1718,
+      "step": 10000
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 0.0001006978711038535,
+      "loss": 1.093,
+      "step": 10005
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.177734375,
+      "learning_rate": 0.00010060227357015723,
+      "loss": 1.1666,
+      "step": 10010
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.16015625,
+      "learning_rate": 0.00010050667548602664,
+      "loss": 1.1145,
+      "step": 10015
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 0.0001004110769388314,
+      "loss": 1.137,
+      "step": 10020
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 0.00010031547801594164,
+      "loss": 1.1741,
+      "step": 10025
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 0.00010021987880472788,
+      "loss": 1.161,
+      "step": 10030
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.169921875,
+      "learning_rate": 0.00010012427939256081,
+      "loss": 1.1705,
+      "step": 10035
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1796875,
+      "learning_rate": 0.00010002867986681142,
+      "loss": 1.2379,
+      "step": 10040
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.166015625,
+      "learning_rate": 9.993308031485071e-05,
+      "loss": 1.1709,
+      "step": 10045
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.98374808240497e-05,
+      "loss": 1.2326,
+      "step": 10050
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.974188148177946e-05,
+      "loss": 1.2494,
+      "step": 10055
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1640625,
+      "learning_rate": 9.964628237541084e-05,
+      "loss": 1.0937,
+      "step": 10060
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.955068359231446e-05,
+      "loss": 1.166,
+      "step": 10065
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.169921875,
+      "learning_rate": 9.945508521986066e-05,
+      "loss": 1.1593,
+      "step": 10070
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.935948734541947e-05,
+      "loss": 1.1114,
+      "step": 10075
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 9.926389005636039e-05,
+      "loss": 1.2054,
+      "step": 10080
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.91682934400524e-05,
+      "loss": 1.1198,
+      "step": 10085
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.907269758386386e-05,
+      "loss": 1.1255,
+      "step": 10090
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.89771025751625e-05,
+      "loss": 1.1594,
+      "step": 10095
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.888150850131516e-05,
+      "loss": 1.1785,
+      "step": 10100
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.878591544968795e-05,
+      "loss": 1.1016,
+      "step": 10105
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 9.869032350764597e-05,
+      "loss": 1.1373,
+      "step": 10110
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.859473276255324e-05,
+      "loss": 1.1844,
+      "step": 10115
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.185546875,
+      "learning_rate": 9.849914330177291e-05,
+      "loss": 1.1619,
+      "step": 10120
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.162109375,
+      "learning_rate": 9.840355521266672e-05,
+      "loss": 1.1239,
+      "step": 10125
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1640625,
+      "learning_rate": 9.830796858259528e-05,
+      "loss": 1.0935,
+      "step": 10130
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.821238349891778e-05,
+      "loss": 1.1757,
+      "step": 10135
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.811680004899214e-05,
+      "loss": 1.081,
+      "step": 10140
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 9.802121832017466e-05,
+      "loss": 1.1609,
+      "step": 10145
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.792563839982012e-05,
+      "loss": 1.2551,
+      "step": 10150
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.783006037528157e-05,
+      "loss": 1.2193,
+      "step": 10155
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 9.77344843339105e-05,
+      "loss": 1.1624,
+      "step": 10160
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.76389103630564e-05,
+      "loss": 1.1218,
+      "step": 10165
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.754333855006699e-05,
+      "loss": 1.1217,
+      "step": 10170
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.74477689822879e-05,
+      "loss": 1.1508,
+      "step": 10175
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.73522017470629e-05,
+      "loss": 1.1326,
+      "step": 10180
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.725663693173343e-05,
+      "loss": 1.1927,
+      "step": 10185
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 9.716107462363881e-05,
+      "loss": 1.1657,
+      "step": 10190
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.70655149101161e-05,
+      "loss": 1.2035,
+      "step": 10195
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 9.696995787849985e-05,
+      "loss": 1.1384,
+      "step": 10200
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.687440361612238e-05,
+      "loss": 1.1734,
+      "step": 10205
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1640625,
+      "learning_rate": 9.677885221031329e-05,
+      "loss": 1.2824,
+      "step": 10210
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.668330374839965e-05,
+      "loss": 1.0996,
+      "step": 10215
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 9.658775831770575e-05,
+      "loss": 1.1382,
+      "step": 10220
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 9.64922160055533e-05,
+      "loss": 1.1684,
+      "step": 10225
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.6396676899261e-05,
+      "loss": 1.1403,
+      "step": 10230
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 9.630114108614464e-05,
+      "loss": 1.1393,
+      "step": 10235
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 9.6205608653517e-05,
+      "loss": 1.1194,
+      "step": 10240
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.162109375,
+      "learning_rate": 9.611007968868787e-05,
+      "loss": 1.1387,
+      "step": 10245
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.169921875,
+      "learning_rate": 9.601455427896377e-05,
+      "loss": 1.1942,
+      "step": 10250
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.591903251164797e-05,
+      "loss": 1.1733,
+      "step": 10255
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.582351447404042e-05,
+      "loss": 1.1244,
+      "step": 10260
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.169921875,
+      "learning_rate": 9.572800025343774e-05,
+      "loss": 1.2431,
+      "step": 10265
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.563248993713296e-05,
+      "loss": 1.177,
+      "step": 10270
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.553698361241557e-05,
+      "loss": 1.1976,
+      "step": 10275
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.544148136657145e-05,
+      "loss": 1.1354,
+      "step": 10280
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.534598328688273e-05,
+      "loss": 1.1658,
+      "step": 10285
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.52504894606277e-05,
+      "loss": 1.1143,
+      "step": 10290
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.515499997508083e-05,
+      "loss": 1.1975,
+      "step": 10295
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.505951491751256e-05,
+      "loss": 1.1698,
+      "step": 10300
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.169921875,
+      "learning_rate": 9.496403437518925e-05,
+      "loss": 1.1516,
+      "step": 10305
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.486855843537332e-05,
+      "loss": 1.1572,
+      "step": 10310
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.477308718532275e-05,
+      "loss": 1.182,
+      "step": 10315
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.467762071229138e-05,
+      "loss": 1.0752,
+      "step": 10320
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.45821591035286e-05,
+      "loss": 1.2191,
+      "step": 10325
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.169921875,
+      "learning_rate": 9.448670244627945e-05,
+      "loss": 1.1375,
+      "step": 10330
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 9.439125082778438e-05,
+      "loss": 1.1327,
+      "step": 10335
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.169921875,
+      "learning_rate": 9.429580433527923e-05,
+      "loss": 1.1228,
+      "step": 10340
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.420036305599512e-05,
+      "loss": 1.1423,
+      "step": 10345
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.410492707715856e-05,
+      "loss": 1.1457,
+      "step": 10350
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.400949648599107e-05,
+      "loss": 1.1912,
+      "step": 10355
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.391407136970926e-05,
+      "loss": 1.1711,
+      "step": 10360
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.166015625,
+      "learning_rate": 9.381865181552479e-05,
+      "loss": 1.1608,
+      "step": 10365
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.162109375,
+      "learning_rate": 9.372323791064423e-05,
+      "loss": 1.1022,
+      "step": 10370
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.362782974226891e-05,
+      "loss": 1.1634,
+      "step": 10375
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.353242739759506e-05,
+      "loss": 1.1867,
+      "step": 10380
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.343703096381342e-05,
+      "loss": 1.1356,
+      "step": 10385
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.334164052810945e-05,
+      "loss": 1.1598,
+      "step": 10390
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1796875,
+      "learning_rate": 9.324625617766309e-05,
+      "loss": 1.2104,
+      "step": 10395
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 9.315087799964873e-05,
+      "loss": 1.1648,
+      "step": 10400
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.305550608123509e-05,
+      "loss": 1.2151,
+      "step": 10405
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.296014050958513e-05,
+      "loss": 1.1961,
+      "step": 10410
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.185546875,
+      "learning_rate": 9.286478137185619e-05,
+      "loss": 1.1608,
+      "step": 10415
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.169921875,
+      "learning_rate": 9.276942875519952e-05,
+      "loss": 1.2612,
+      "step": 10420
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.267408274676051e-05,
+      "loss": 1.1399,
+      "step": 10425
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.257874343367846e-05,
+      "loss": 1.2151,
+      "step": 10430
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.248341090308666e-05,
+      "loss": 1.1745,
+      "step": 10435
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 9.238808524211211e-05,
+      "loss": 1.1334,
+      "step": 10440
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.229276653787552e-05,
+      "loss": 1.1511,
+      "step": 10445
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 9.219745487749125e-05,
+      "loss": 1.1586,
+      "step": 10450
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.210215034806732e-05,
+      "loss": 1.1586,
+      "step": 10455
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 9.200685303670509e-05,
+      "loss": 1.1163,
+      "step": 10460
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.191156303049944e-05,
+      "loss": 1.0988,
+      "step": 10465
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.181628041653848e-05,
+      "loss": 1.1843,
+      "step": 10470
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.172100528190363e-05,
+      "loss": 1.2074,
+      "step": 10475
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.162573771366947e-05,
+      "loss": 1.1822,
+      "step": 10480
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.153047779890363e-05,
+      "loss": 1.1655,
+      "step": 10485
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.143522562466672e-05,
+      "loss": 1.2011,
+      "step": 10490
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.166015625,
+      "learning_rate": 9.133998127801239e-05,
+      "loss": 1.1183,
+      "step": 10495
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.124474484598706e-05,
+      "loss": 1.1215,
+      "step": 10500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.114951641562988e-05,
+      "loss": 1.1731,
+      "step": 10505
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1640625,
+      "learning_rate": 9.105429607397274e-05,
+      "loss": 1.1641,
+      "step": 10510
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.09590839080401e-05,
+      "loss": 1.2396,
+      "step": 10515
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.086388000484903e-05,
+      "loss": 1.1676,
+      "step": 10520
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.076868445140897e-05,
+      "loss": 1.1063,
+      "step": 10525
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.067349733472175e-05,
+      "loss": 1.237,
+      "step": 10530
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 9.057831874178145e-05,
+      "loss": 1.2237,
+      "step": 10535
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.048314875957445e-05,
+      "loss": 1.1732,
+      "step": 10540
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.038798747507923e-05,
+      "loss": 1.2999,
+      "step": 10545
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.029283497526622e-05,
+      "loss": 1.1659,
+      "step": 10550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.019769134709797e-05,
+      "loss": 1.0748,
+      "step": 10555
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 9.010255667752883e-05,
+      "loss": 1.2317,
+      "step": 10560
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 9.000743105350502e-05,
+      "loss": 1.2144,
+      "step": 10565
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 8.991231456196444e-05,
+      "loss": 1.1363,
+      "step": 10570
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.981720728983662e-05,
+      "loss": 1.092,
+      "step": 10575
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 8.972210932404277e-05,
+      "loss": 1.1072,
+      "step": 10580
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.962702075149552e-05,
+      "loss": 1.1435,
+      "step": 10585
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.953194165909893e-05,
+      "loss": 1.1958,
+      "step": 10590
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.943687213374833e-05,
+      "loss": 1.1728,
+      "step": 10595
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.934181226233046e-05,
+      "loss": 1.1884,
+      "step": 10600
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.924676213172312e-05,
+      "loss": 1.1324,
+      "step": 10605
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 8.915172182879524e-05,
+      "loss": 1.149,
+      "step": 10610
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.905669144040673e-05,
+      "loss": 1.1547,
+      "step": 10615
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.896167105340844e-05,
+      "loss": 1.2046,
+      "step": 10620
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.166015625,
+      "learning_rate": 8.886666075464223e-05,
+      "loss": 1.1899,
+      "step": 10625
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.877166063094053e-05,
+      "loss": 1.1937,
+      "step": 10630
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.171875,
+      "learning_rate": 8.867667076912659e-05,
+      "loss": 1.1743,
+      "step": 10635
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 8.858169125601423e-05,
+      "loss": 1.2328,
+      "step": 10640
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.84867221784079e-05,
+      "loss": 1.2201,
+      "step": 10645
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.839176362310237e-05,
+      "loss": 1.1055,
+      "step": 10650
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.829681567688293e-05,
+      "loss": 1.1566,
+      "step": 10655
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 8.820187842652505e-05,
+      "loss": 1.141,
+      "step": 10660
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.810695195879458e-05,
+      "loss": 1.1052,
+      "step": 10665
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.801203636044738e-05,
+      "loss": 1.0451,
+      "step": 10670
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 8.791713171822943e-05,
+      "loss": 1.1743,
+      "step": 10675
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.782223811887663e-05,
+      "loss": 1.0914,
+      "step": 10680
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1640625,
+      "learning_rate": 8.772735564911494e-05,
+      "loss": 1.1892,
+      "step": 10685
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 8.763248439566002e-05,
+      "loss": 1.2377,
+      "step": 10690
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.753762444521731e-05,
+      "loss": 1.16,
+      "step": 10695
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.744277588448191e-05,
+      "loss": 1.2214,
+      "step": 10700
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.73479388001385e-05,
+      "loss": 1.1693,
+      "step": 10705
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 8.725311327886139e-05,
+      "loss": 1.1424,
+      "step": 10710
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 8.715829940731415e-05,
+      "loss": 1.2334,
+      "step": 10715
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.70634972721498e-05,
+      "loss": 1.1767,
+      "step": 10720
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.69687069600106e-05,
+      "loss": 1.1229,
+      "step": 10725
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.687392855752808e-05,
+      "loss": 1.2189,
+      "step": 10730
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.677916215132275e-05,
+      "loss": 1.1704,
+      "step": 10735
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.66844078280043e-05,
+      "loss": 1.122,
+      "step": 10740
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.658966567417124e-05,
+      "loss": 1.1558,
+      "step": 10745
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1796875,
+      "learning_rate": 8.649493577641107e-05,
+      "loss": 1.2495,
+      "step": 10750
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.166015625,
+      "learning_rate": 8.640021822130004e-05,
+      "loss": 1.1238,
+      "step": 10755
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 8.630551309540314e-05,
+      "loss": 1.1335,
+      "step": 10760
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1796875,
+      "learning_rate": 8.621082048527392e-05,
+      "loss": 1.2267,
+      "step": 10765
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.611614047745466e-05,
+      "loss": 1.1267,
+      "step": 10770
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.602147315847595e-05,
+      "loss": 1.1962,
+      "step": 10775
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.592681861485687e-05,
+      "loss": 1.1619,
+      "step": 10780
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 8.583217693310475e-05,
+      "loss": 1.1219,
+      "step": 10785
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.573754819971533e-05,
+      "loss": 1.1878,
+      "step": 10790
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.564293250117237e-05,
+      "loss": 1.1166,
+      "step": 10795
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1640625,
+      "learning_rate": 8.554832992394773e-05,
+      "loss": 1.1859,
+      "step": 10800
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 8.54537405545013e-05,
+      "loss": 1.1452,
+      "step": 10805
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.53591644792809e-05,
+      "loss": 1.1658,
+      "step": 10810
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.526460178472225e-05,
+      "loss": 1.1467,
+      "step": 10815
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.17578125,
+      "learning_rate": 8.517005255724874e-05,
+      "loss": 1.2163,
+      "step": 10820
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.507551688327157e-05,
+      "loss": 1.1833,
+      "step": 10825
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.17578125,
+      "learning_rate": 8.498099484918943e-05,
+      "loss": 1.1367,
+      "step": 10830
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.162109375,
+      "learning_rate": 8.48864865413886e-05,
+      "loss": 1.1606,
+      "step": 10835
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.479199204624288e-05,
+      "loss": 1.1066,
+      "step": 10840
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.469751145011333e-05,
+      "loss": 1.1006,
+      "step": 10845
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.171875,
+      "learning_rate": 8.460304483934834e-05,
+      "loss": 1.189,
+      "step": 10850
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1640625,
+      "learning_rate": 8.450859230028363e-05,
+      "loss": 1.153,
+      "step": 10855
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.441415391924192e-05,
+      "loss": 1.159,
+      "step": 10860
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 8.431972978253303e-05,
+      "loss": 1.1256,
+      "step": 10865
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.162109375,
+      "learning_rate": 8.422531997645374e-05,
+      "loss": 1.2179,
+      "step": 10870
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.413092458728786e-05,
+      "loss": 1.175,
+      "step": 10875
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.403654370130585e-05,
+      "loss": 1.1468,
+      "step": 10880
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.17578125,
+      "learning_rate": 8.394217740476505e-05,
+      "loss": 1.1275,
+      "step": 10885
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.384782578390931e-05,
+      "loss": 1.1314,
+      "step": 10890
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.375348892496929e-05,
+      "loss": 1.2109,
+      "step": 10895
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 8.365916691416197e-05,
+      "loss": 1.1558,
+      "step": 10900
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.166015625,
+      "learning_rate": 8.356485983769085e-05,
+      "loss": 1.1935,
+      "step": 10905
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.347056778174572e-05,
+      "loss": 1.1658,
+      "step": 10910
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.337629083250272e-05,
+      "loss": 1.1647,
+      "step": 10915
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.328202907612412e-05,
+      "loss": 1.1896,
+      "step": 10920
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.318778259875836e-05,
+      "loss": 1.1991,
+      "step": 10925
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.309355148653986e-05,
+      "loss": 1.1244,
+      "step": 10930
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.171875,
+      "learning_rate": 8.299933582558899e-05,
+      "loss": 1.2094,
+      "step": 10935
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.290513570201211e-05,
+      "loss": 1.1765,
+      "step": 10940
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.281095120190125e-05,
+      "loss": 1.102,
+      "step": 10945
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.27167824113342e-05,
+      "loss": 1.2151,
+      "step": 10950
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 8.26226294163744e-05,
+      "loss": 1.1786,
+      "step": 10955
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.252849230307092e-05,
+      "loss": 1.1476,
+      "step": 10960
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.24343711574582e-05,
+      "loss": 1.1859,
+      "step": 10965
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.234026606555617e-05,
+      "loss": 1.2477,
+      "step": 10970
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1640625,
+      "learning_rate": 8.224617711336999e-05,
+      "loss": 1.141,
+      "step": 10975
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 8.215210438689023e-05,
+      "loss": 1.112,
+      "step": 10980
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1796875,
+      "learning_rate": 8.20580479720925e-05,
+      "loss": 1.1282,
+      "step": 10985
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.196400795493753e-05,
+      "loss": 1.1125,
+      "step": 10990
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 8.186998442137106e-05,
+      "loss": 1.1781,
+      "step": 10995
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 8.177597745732384e-05,
+      "loss": 1.1558,
+      "step": 11000
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.168198714871137e-05,
+      "loss": 1.1321,
+      "step": 11005
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.158801358143397e-05,
+      "loss": 1.2147,
+      "step": 11010
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.149405684137669e-05,
+      "loss": 1.2106,
+      "step": 11015
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.140011701440912e-05,
+      "loss": 1.1424,
+      "step": 11020
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.130619418638554e-05,
+      "loss": 1.1261,
+      "step": 11025
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.121228844314457e-05,
+      "loss": 1.1197,
+      "step": 11030
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.166015625,
+      "learning_rate": 8.111839987050923e-05,
+      "loss": 1.2139,
+      "step": 11035
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.10245285542868e-05,
+      "loss": 1.2004,
+      "step": 11040
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.093067458026898e-05,
+      "loss": 1.1045,
+      "step": 11045
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.083683803423143e-05,
+      "loss": 1.1802,
+      "step": 11050
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.074301900193396e-05,
+      "loss": 1.1863,
+      "step": 11055
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.171875,
+      "learning_rate": 8.064921756912032e-05,
+      "loss": 1.1537,
+      "step": 11060
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.055543382151828e-05,
+      "loss": 1.2008,
+      "step": 11065
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.046166784483935e-05,
+      "loss": 1.15,
+      "step": 11070
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.036791972477884e-05,
+      "loss": 1.2001,
+      "step": 11075
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 8.02741895470157e-05,
+      "loss": 1.2361,
+      "step": 11080
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.17578125,
+      "learning_rate": 8.018047739721255e-05,
+      "loss": 1.2323,
+      "step": 11085
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.00867833610155e-05,
+      "loss": 1.2584,
+      "step": 11090
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.999310752405408e-05,
+      "loss": 1.159,
+      "step": 11095
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.989944997194123e-05,
+      "loss": 1.1296,
+      "step": 11100
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.980581079027314e-05,
+      "loss": 1.1494,
+      "step": 11105
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.971219006462925e-05,
+      "loss": 1.1472,
+      "step": 11110
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.96185878805721e-05,
+      "loss": 1.0716,
+      "step": 11115
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.952500432364731e-05,
+      "loss": 1.1093,
+      "step": 11120
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.943143947938342e-05,
+      "loss": 1.1564,
+      "step": 11125
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.933789343329199e-05,
+      "loss": 1.1545,
+      "step": 11130
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.924436627086726e-05,
+      "loss": 1.093,
+      "step": 11135
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.91508580775863e-05,
+      "loss": 1.1415,
+      "step": 11140
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.905736893890875e-05,
+      "loss": 1.1978,
+      "step": 11145
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.8963898940277e-05,
+      "loss": 1.1466,
+      "step": 11150
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.887044816711581e-05,
+      "loss": 1.1666,
+      "step": 11155
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.877701670483242e-05,
+      "loss": 1.0945,
+      "step": 11160
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.868360463881633e-05,
+      "loss": 1.1524,
+      "step": 11165
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.859021205443951e-05,
+      "loss": 1.1879,
+      "step": 11170
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.849683903705595e-05,
+      "loss": 1.1988,
+      "step": 11175
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.84034856720018e-05,
+      "loss": 1.1913,
+      "step": 11180
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.831015204459527e-05,
+      "loss": 1.148,
+      "step": 11185
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.821683824013658e-05,
+      "loss": 1.1996,
+      "step": 11190
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.812354434390771e-05,
+      "loss": 1.1253,
+      "step": 11195
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.803027044117257e-05,
+      "loss": 1.1595,
+      "step": 11200
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.79370166171767e-05,
+      "loss": 1.2007,
+      "step": 11205
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.784378295714731e-05,
+      "loss": 1.216,
+      "step": 11210
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.77505695462933e-05,
+      "loss": 1.1905,
+      "step": 11215
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.76573764698049e-05,
+      "loss": 1.1776,
+      "step": 11220
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.756420381285379e-05,
+      "loss": 1.222,
+      "step": 11225
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.747105166059306e-05,
+      "loss": 1.2169,
+      "step": 11230
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.737792009815704e-05,
+      "loss": 1.1233,
+      "step": 11235
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.72848092106612e-05,
+      "loss": 1.2118,
+      "step": 11240
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.719171908320214e-05,
+      "loss": 1.2159,
+      "step": 11245
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.709864980085744e-05,
+      "loss": 1.1048,
+      "step": 11250
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.700560144868576e-05,
+      "loss": 1.2757,
+      "step": 11255
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.69125741117265e-05,
+      "loss": 1.1914,
+      "step": 11260
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.681956787499991e-05,
+      "loss": 1.1629,
+      "step": 11265
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.672658282350687e-05,
+      "loss": 1.1249,
+      "step": 11270
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.66336190422291e-05,
+      "loss": 1.1994,
+      "step": 11275
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.654067661612865e-05,
+      "loss": 1.181,
+      "step": 11280
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.644775563014824e-05,
+      "loss": 1.0829,
+      "step": 11285
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.635485616921083e-05,
+      "loss": 1.1677,
+      "step": 11290
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.626197831821985e-05,
+      "loss": 1.0531,
+      "step": 11295
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.616912216205888e-05,
+      "loss": 1.254,
+      "step": 11300
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.607628778559175e-05,
+      "loss": 1.2373,
+      "step": 11305
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 7.598347527366233e-05,
+      "loss": 1.1418,
+      "step": 11310
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.589068471109446e-05,
+      "loss": 1.2046,
+      "step": 11315
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.579791618269207e-05,
+      "loss": 1.1768,
+      "step": 11320
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.570516977323886e-05,
+      "loss": 1.1717,
+      "step": 11325
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.56124455674983e-05,
+      "loss": 1.2207,
+      "step": 11330
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.551974365021353e-05,
+      "loss": 1.1449,
+      "step": 11335
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.542706410610747e-05,
+      "loss": 1.2054,
+      "step": 11340
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.533440701988249e-05,
+      "loss": 1.1281,
+      "step": 11345
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.524177247622042e-05,
+      "loss": 1.0928,
+      "step": 11350
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.514916055978246e-05,
+      "loss": 1.19,
+      "step": 11355
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.505657135520929e-05,
+      "loss": 1.1097,
+      "step": 11360
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.496400494712068e-05,
+      "loss": 1.1422,
+      "step": 11365
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.487146142011559e-05,
+      "loss": 1.2908,
+      "step": 11370
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.477894085877211e-05,
+      "loss": 1.2214,
+      "step": 11375
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.468644334764731e-05,
+      "loss": 1.137,
+      "step": 11380
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.459396897127726e-05,
+      "loss": 1.1791,
+      "step": 11385
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.450151781417677e-05,
+      "loss": 1.1886,
+      "step": 11390
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.44090899608395e-05,
+      "loss": 1.1103,
+      "step": 11395
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.431668549573785e-05,
+      "loss": 1.1789,
+      "step": 11400
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.422430450332276e-05,
+      "loss": 1.1381,
+      "step": 11405
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.413194706802378e-05,
+      "loss": 1.1534,
+      "step": 11410
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.403961327424888e-05,
+      "loss": 1.1222,
+      "step": 11415
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.394730320638441e-05,
+      "loss": 1.1583,
+      "step": 11420
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.385501694879518e-05,
+      "loss": 1.1558,
+      "step": 11425
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.376275458582409e-05,
+      "loss": 1.2163,
+      "step": 11430
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.367051620179221e-05,
+      "loss": 1.134,
+      "step": 11435
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.357830188099875e-05,
+      "loss": 1.1114,
+      "step": 11440
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.348611170772095e-05,
+      "loss": 1.1102,
+      "step": 11445
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.339394576621393e-05,
+      "loss": 1.188,
+      "step": 11450
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.330180414071063e-05,
+      "loss": 1.1481,
+      "step": 11455
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.320968691542185e-05,
+      "loss": 1.1679,
+      "step": 11460
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.311759417453605e-05,
+      "loss": 1.1306,
+      "step": 11465
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.30255260022193e-05,
+      "loss": 1.1473,
+      "step": 11470
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 7.293348248261526e-05,
+      "loss": 1.1137,
+      "step": 11475
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.284146369984498e-05,
+      "loss": 1.2396,
+      "step": 11480
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.274946973800703e-05,
+      "loss": 1.1465,
+      "step": 11485
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.265750068117716e-05,
+      "loss": 1.1559,
+      "step": 11490
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.256555661340843e-05,
+      "loss": 1.201,
+      "step": 11495
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.247363761873103e-05,
+      "loss": 1.101,
+      "step": 11500
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.23817437811523e-05,
+      "loss": 1.1014,
+      "step": 11505
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.228987518465652e-05,
+      "loss": 1.1576,
+      "step": 11510
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.219803191320492e-05,
+      "loss": 1.1553,
+      "step": 11515
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.210621405073558e-05,
+      "loss": 1.1841,
+      "step": 11520
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.201442168116335e-05,
+      "loss": 1.1712,
+      "step": 11525
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.192265488837986e-05,
+      "loss": 1.0964,
+      "step": 11530
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.183091375625327e-05,
+      "loss": 1.1773,
+      "step": 11535
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.173919836862829e-05,
+      "loss": 1.1961,
+      "step": 11540
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.164750880932616e-05,
+      "loss": 1.1477,
+      "step": 11545
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.155584516214451e-05,
+      "loss": 1.2039,
+      "step": 11550
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.14642075108572e-05,
+      "loss": 1.2151,
+      "step": 11555
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.137259593921447e-05,
+      "loss": 1.1193,
+      "step": 11560
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.128101053094261e-05,
+      "loss": 1.1474,
+      "step": 11565
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.118945136974402e-05,
+      "loss": 1.1494,
+      "step": 11570
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.10979185392972e-05,
+      "loss": 1.1307,
+      "step": 11575
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 7.100641212325648e-05,
+      "loss": 1.102,
+      "step": 11580
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.185546875,
+      "learning_rate": 7.091493220525203e-05,
+      "loss": 1.1107,
+      "step": 11585
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.082347886888996e-05,
+      "loss": 1.1833,
+      "step": 11590
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.193359375,
+      "learning_rate": 7.073205219775194e-05,
+      "loss": 1.0751,
+      "step": 11595
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.064065227539532e-05,
+      "loss": 1.14,
+      "step": 11600
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.0549279185353e-05,
+      "loss": 1.1199,
+      "step": 11605
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.04579330111333e-05,
+      "loss": 1.177,
+      "step": 11610
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.036661383622012e-05,
+      "loss": 1.175,
+      "step": 11615
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.027532174407248e-05,
+      "loss": 1.1929,
+      "step": 11620
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.018405681812473e-05,
+      "loss": 1.1093,
+      "step": 11625
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.009281914178636e-05,
+      "loss": 1.1812,
+      "step": 11630
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.000160879844209e-05,
+      "loss": 1.2149,
+      "step": 11635
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.99104258714515e-05,
+      "loss": 1.1322,
+      "step": 11640
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.166015625,
+      "learning_rate": 6.981927044414915e-05,
+      "loss": 1.1873,
+      "step": 11645
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.972814259984452e-05,
+      "loss": 1.1271,
+      "step": 11650
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.963704242182184e-05,
+      "loss": 1.1403,
+      "step": 11655
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.16796875,
+      "learning_rate": 6.954596999334008e-05,
+      "loss": 1.1593,
+      "step": 11660
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.945492539763285e-05,
+      "loss": 1.1941,
+      "step": 11665
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.93639087179082e-05,
+      "loss": 1.1829,
+      "step": 11670
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.927292003734892e-05,
+      "loss": 1.1681,
+      "step": 11675
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.9181959439112e-05,
+      "loss": 1.325,
+      "step": 11680
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.166015625,
+      "learning_rate": 6.90910270063288e-05,
+      "loss": 1.1048,
+      "step": 11685
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.900012282210495e-05,
+      "loss": 1.1311,
+      "step": 11690
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.16796875,
+      "learning_rate": 6.890924696952038e-05,
+      "loss": 1.1954,
+      "step": 11695
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.881839953162896e-05,
+      "loss": 1.2225,
+      "step": 11700
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.87275805914587e-05,
+      "loss": 1.1493,
+      "step": 11705
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.863679023201146e-05,
+      "loss": 1.1678,
+      "step": 11710
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.854602853626305e-05,
+      "loss": 1.1627,
+      "step": 11715
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.845529558716315e-05,
+      "loss": 1.1878,
+      "step": 11720
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.836459146763504e-05,
+      "loss": 1.1604,
+      "step": 11725
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.827391626057571e-05,
+      "loss": 1.1919,
+      "step": 11730
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.818327004885571e-05,
+      "loss": 1.1219,
+      "step": 11735
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.809265291531913e-05,
+      "loss": 1.1519,
+      "step": 11740
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.800206494278345e-05,
+      "loss": 1.1143,
+      "step": 11745
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.16796875,
+      "learning_rate": 6.791150621403953e-05,
+      "loss": 1.2425,
+      "step": 11750
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.16796875,
+      "learning_rate": 6.78209768118514e-05,
+      "loss": 1.0954,
+      "step": 11755
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.773047681895646e-05,
+      "loss": 1.1626,
+      "step": 11760
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.764000631806511e-05,
+      "loss": 1.1976,
+      "step": 11765
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.75495653918608e-05,
+      "loss": 1.1826,
+      "step": 11770
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 6.745915412299994e-05,
+      "loss": 1.1087,
+      "step": 11775
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.7368772594112e-05,
+      "loss": 1.1276,
+      "step": 11780
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.727842088779905e-05,
+      "loss": 1.2056,
+      "step": 11785
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.16796875,
+      "learning_rate": 6.7188099086636e-05,
+      "loss": 1.198,
+      "step": 11790
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.709780727317041e-05,
+      "loss": 1.1536,
+      "step": 11795
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.162109375,
+      "learning_rate": 6.70075455299225e-05,
+      "loss": 1.1171,
+      "step": 11800
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.691731393938494e-05,
+      "loss": 1.1072,
+      "step": 11805
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.682711258402285e-05,
+      "loss": 1.0588,
+      "step": 11810
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.673694154627372e-05,
+      "loss": 1.1593,
+      "step": 11815
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 6.664680090854734e-05,
+      "loss": 1.155,
+      "step": 11820
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.655669075322573e-05,
+      "loss": 1.1108,
+      "step": 11825
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.6466611162663e-05,
+      "loss": 1.2183,
+      "step": 11830
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.637656221918541e-05,
+      "loss": 1.2246,
+      "step": 11835
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1640625,
+      "learning_rate": 6.62865440050911e-05,
+      "loss": 1.1802,
+      "step": 11840
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.619655660265028e-05,
+      "loss": 1.1221,
+      "step": 11845
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.610660009410485e-05,
+      "loss": 1.1762,
+      "step": 11850
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.601667456166854e-05,
+      "loss": 1.0455,
+      "step": 11855
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.592678008752673e-05,
+      "loss": 1.1638,
+      "step": 11860
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.583691675383652e-05,
+      "loss": 1.1619,
+      "step": 11865
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 6.574708464272645e-05,
+      "loss": 1.1493,
+      "step": 11870
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.565728383629655e-05,
+      "loss": 1.2081,
+      "step": 11875
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.556751441661818e-05,
+      "loss": 1.1543,
+      "step": 11880
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.54777764657342e-05,
+      "loss": 1.2102,
+      "step": 11885
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.538807006565851e-05,
+      "loss": 1.2417,
+      "step": 11890
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.529839529837629e-05,
+      "loss": 1.1262,
+      "step": 11895
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.520875224584369e-05,
+      "loss": 1.1561,
+      "step": 11900
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.511914098998809e-05,
+      "loss": 1.2026,
+      "step": 11905
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 6.502956161270758e-05,
+      "loss": 1.1308,
+      "step": 11910
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.494001419587123e-05,
+      "loss": 1.1639,
+      "step": 11915
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.485049882131893e-05,
+      "loss": 1.1152,
+      "step": 11920
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.476101557086117e-05,
+      "loss": 1.164,
+      "step": 11925
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.467156452627919e-05,
+      "loss": 1.1221,
+      "step": 11930
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.458214576932478e-05,
+      "loss": 1.1723,
+      "step": 11935
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.449275938172016e-05,
+      "loss": 1.1294,
+      "step": 11940
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.440340544515798e-05,
+      "loss": 1.1893,
+      "step": 11945
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.431408404130134e-05,
+      "loss": 1.2588,
+      "step": 11950
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.422479525178347e-05,
+      "loss": 1.1218,
+      "step": 11955
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.413553915820785e-05,
+      "loss": 1.1684,
+      "step": 11960
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.404631584214804e-05,
+      "loss": 1.1561,
+      "step": 11965
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.395712538514778e-05,
+      "loss": 1.2067,
+      "step": 11970
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.386796786872058e-05,
+      "loss": 1.188,
+      "step": 11975
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.377884337435001e-05,
+      "loss": 1.2199,
+      "step": 11980
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.368975198348932e-05,
+      "loss": 1.2381,
+      "step": 11985
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.16796875,
+      "learning_rate": 6.360069377756166e-05,
+      "loss": 1.1386,
+      "step": 11990
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.166015625,
+      "learning_rate": 6.351166883795973e-05,
+      "loss": 1.1694,
+      "step": 11995
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.342267724604584e-05,
+      "loss": 1.1833,
+      "step": 12000
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.33337190831519e-05,
+      "loss": 1.1712,
+      "step": 12005
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.324479443057922e-05,
+      "loss": 1.1911,
+      "step": 12010
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.315590336959842e-05,
+      "loss": 1.1977,
+      "step": 12015
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.306704598144956e-05,
+      "loss": 1.2255,
+      "step": 12020
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.297822234734182e-05,
+      "loss": 1.0571,
+      "step": 12025
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.28894325484535e-05,
+      "loss": 1.1637,
+      "step": 12030
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.280067666593213e-05,
+      "loss": 1.2216,
+      "step": 12035
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.271195478089411e-05,
+      "loss": 1.1278,
+      "step": 12040
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.26232669744248e-05,
+      "loss": 1.21,
+      "step": 12045
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.16796875,
+      "learning_rate": 6.253461332757838e-05,
+      "loss": 1.1584,
+      "step": 12050
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.244599392137794e-05,
+      "loss": 1.2096,
+      "step": 12055
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.235740883681515e-05,
+      "loss": 1.1499,
+      "step": 12060
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.226885815485036e-05,
+      "loss": 1.2646,
+      "step": 12065
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.21803419564124e-05,
+      "loss": 1.1436,
+      "step": 12070
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.209186032239877e-05,
+      "loss": 1.0668,
+      "step": 12075
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.200341333367521e-05,
+      "loss": 1.1911,
+      "step": 12080
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.191500107107586e-05,
+      "loss": 1.1395,
+      "step": 12085
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.182662361540305e-05,
+      "loss": 1.2165,
+      "step": 12090
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.173828104742749e-05,
+      "loss": 1.2104,
+      "step": 12095
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.164997344788776e-05,
+      "loss": 1.2337,
+      "step": 12100
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.15617008974907e-05,
+      "loss": 1.1937,
+      "step": 12105
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.147346347691097e-05,
+      "loss": 1.1332,
+      "step": 12110
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.166015625,
+      "learning_rate": 6.138526126679112e-05,
+      "loss": 1.1138,
+      "step": 12115
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.129709434774166e-05,
+      "loss": 1.1369,
+      "step": 12120
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.120896280034074e-05,
+      "loss": 1.1501,
+      "step": 12125
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.112086670513419e-05,
+      "loss": 1.1227,
+      "step": 12130
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.103280614263539e-05,
+      "loss": 1.0718,
+      "step": 12135
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.094478119332542e-05,
+      "loss": 1.2111,
+      "step": 12140
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.085679193765264e-05,
+      "loss": 1.12,
+      "step": 12145
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.076883845603285e-05,
+      "loss": 1.1479,
+      "step": 12150
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.068092082884912e-05,
+      "loss": 1.1859,
+      "step": 12155
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.0593039136451845e-05,
+      "loss": 1.2051,
+      "step": 12160
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.050519345915851e-05,
+      "loss": 1.1908,
+      "step": 12165
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.041738387725366e-05,
+      "loss": 1.137,
+      "step": 12170
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.16796875,
+      "learning_rate": 6.032961047098886e-05,
+      "loss": 1.1643,
+      "step": 12175
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.024187332058271e-05,
+      "loss": 1.1522,
+      "step": 12180
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.015417250622057e-05,
+      "loss": 1.1674,
+      "step": 12185
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.006650810805461e-05,
+      "loss": 1.146,
+      "step": 12190
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.997888020620373e-05,
+      "loss": 1.1807,
+      "step": 12195
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.9891288880753486e-05,
+      "loss": 1.1484,
+      "step": 12200
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.980373421175601e-05,
+      "loss": 1.1014,
+      "step": 12205
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.971621627922992e-05,
+      "loss": 1.1192,
+      "step": 12210
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.962873516316023e-05,
+      "loss": 1.1931,
+      "step": 12215
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.9541290943498316e-05,
+      "loss": 1.1835,
+      "step": 12220
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.945388370016192e-05,
+      "loss": 1.1538,
+      "step": 12225
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.9366513513034883e-05,
+      "loss": 1.1747,
+      "step": 12230
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.185546875,
+      "learning_rate": 5.9279180461967235e-05,
+      "loss": 1.1316,
+      "step": 12235
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.9191884626774993e-05,
+      "loss": 1.2002,
+      "step": 12240
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.91046260872403e-05,
+      "loss": 1.1811,
+      "step": 12245
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 5.901740492311111e-05,
+      "loss": 1.1516,
+      "step": 12250
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.893022121410121e-05,
+      "loss": 1.1061,
+      "step": 12255
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 5.8843075039890174e-05,
+      "loss": 1.0782,
+      "step": 12260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.8755966480123345e-05,
+      "loss": 1.2757,
+      "step": 12265
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 5.86688956144116e-05,
+      "loss": 1.1038,
+      "step": 12270
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.8581862522331376e-05,
+      "loss": 1.1952,
+      "step": 12275
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.8494867283424615e-05,
+      "loss": 1.2524,
+      "step": 12280
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.840790997719868e-05,
+      "loss": 1.1178,
+      "step": 12285
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.832099068312623e-05,
+      "loss": 1.1701,
+      "step": 12290
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.823410948064516e-05,
+      "loss": 1.0914,
+      "step": 12295
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.814726644915862e-05,
+      "loss": 1.1672,
+      "step": 12300
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.806046166803485e-05,
+      "loss": 1.1787,
+      "step": 12305
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.797369521660714e-05,
+      "loss": 1.1533,
+      "step": 12310
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.166015625,
+      "learning_rate": 5.788696717417366e-05,
+      "loss": 1.2159,
+      "step": 12315
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.78002776199976e-05,
+      "loss": 1.205,
+      "step": 12320
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.7713626633306924e-05,
+      "loss": 1.1475,
+      "step": 12325
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.7627014293294334e-05,
+      "loss": 1.1225,
+      "step": 12330
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 5.754044067911728e-05,
+      "loss": 1.2068,
+      "step": 12335
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.166015625,
+      "learning_rate": 5.7453905869897696e-05,
+      "loss": 1.1058,
+      "step": 12340
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.736740994472214e-05,
+      "loss": 1.2131,
+      "step": 12345
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.728095298264161e-05,
+      "loss": 1.0874,
+      "step": 12350
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.719453506267156e-05,
+      "loss": 1.1569,
+      "step": 12355
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.181640625,
+      "learning_rate": 5.710815626379161e-05,
+      "loss": 1.1321,
+      "step": 12360
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.7021816664945747e-05,
+      "loss": 1.0889,
+      "step": 12365
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 5.6935516345042125e-05,
+      "loss": 1.1424,
+      "step": 12370
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.6849255382953e-05,
+      "loss": 1.1572,
+      "step": 12375
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.6763033857514555e-05,
+      "loss": 1.1449,
+      "step": 12380
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 5.667685184752707e-05,
+      "loss": 1.1553,
+      "step": 12385
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.659070943175462e-05,
+      "loss": 1.1084,
+      "step": 12390
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.6504606688925185e-05,
+      "loss": 1.1901,
+      "step": 12395
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.641854369773034e-05,
+      "loss": 1.1956,
+      "step": 12400
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.633252053682545e-05,
+      "loss": 1.1692,
+      "step": 12405
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 5.624653728482947e-05,
+      "loss": 1.1563,
+      "step": 12410
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.616059402032485e-05,
+      "loss": 1.1523,
+      "step": 12415
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.607469082185748e-05,
+      "loss": 1.1506,
+      "step": 12420
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1640625,
+      "learning_rate": 5.5988827767936633e-05,
+      "loss": 1.1812,
+      "step": 12425
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.590300493703496e-05,
+      "loss": 1.0443,
+      "step": 12430
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.581722240758829e-05,
+      "loss": 1.2583,
+      "step": 12435
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.5731480257995636e-05,
+      "loss": 1.1145,
+      "step": 12440
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.564577856661908e-05,
+      "loss": 1.1461,
+      "step": 12445
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 5.556011741178376e-05,
+      "loss": 1.2273,
+      "step": 12450
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.181640625,
+      "learning_rate": 5.547449687177777e-05,
+      "loss": 1.1375,
+      "step": 12455
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.5388917024852096e-05,
+      "loss": 1.1263,
+      "step": 12460
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.166015625,
+      "learning_rate": 5.5303377949220446e-05,
+      "loss": 1.1205,
+      "step": 12465
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.521787972305936e-05,
+      "loss": 1.1976,
+      "step": 12470
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 5.5132422424508025e-05,
+      "loss": 1.1398,
+      "step": 12475
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 5.504700613166827e-05,
+      "loss": 1.1556,
+      "step": 12480
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.4961630922604266e-05,
+      "loss": 1.2524,
+      "step": 12485
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1953125,
+      "learning_rate": 5.487629687534284e-05,
+      "loss": 1.2138,
+      "step": 12490
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.47910040678731e-05,
+      "loss": 1.1351,
+      "step": 12495
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.470575257814654e-05,
+      "loss": 1.2218,
+      "step": 12500
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.462054248407675e-05,
+      "loss": 1.19,
+      "step": 12505
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.453537386353962e-05,
+      "loss": 1.1435,
+      "step": 12510
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.4450246794373094e-05,
+      "loss": 1.2136,
+      "step": 12515
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.4365161354377124e-05,
+      "loss": 1.2098,
+      "step": 12520
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.4280117621313685e-05,
+      "loss": 1.1028,
+      "step": 12525
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.4195115672906494e-05,
+      "loss": 1.1161,
+      "step": 12530
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.41101555868412e-05,
+      "loss": 1.2346,
+      "step": 12535
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.4025237440765195e-05,
+      "loss": 1.1362,
+      "step": 12540
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.3940361312287505e-05,
+      "loss": 1.1449,
+      "step": 12545
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.166015625,
+      "learning_rate": 5.385552727897868e-05,
+      "loss": 1.0648,
+      "step": 12550
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.3770735418370924e-05,
+      "loss": 1.0629,
+      "step": 12555
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.368598580795785e-05,
+      "loss": 1.1927,
+      "step": 12560
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 5.3601278525194476e-05,
+      "loss": 1.1831,
+      "step": 12565
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.3516613647497075e-05,
+      "loss": 1.1503,
+      "step": 12570
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.343199125224319e-05,
+      "loss": 1.1571,
+      "step": 12575
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 5.33474114167716e-05,
+      "loss": 1.209,
+      "step": 12580
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.326287421838214e-05,
+      "loss": 1.1659,
+      "step": 12585
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.317837973433563e-05,
+      "loss": 1.1821,
+      "step": 12590
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.309392804185391e-05,
+      "loss": 1.1669,
+      "step": 12595
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 5.300951921811974e-05,
+      "loss": 1.1642,
+      "step": 12600
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.2925153340276626e-05,
+      "loss": 1.1552,
+      "step": 12605
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.2840830485428917e-05,
+      "loss": 1.2297,
+      "step": 12610
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.2756550730641505e-05,
+      "loss": 1.1281,
+      "step": 12615
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.181640625,
+      "learning_rate": 5.267231415294002e-05,
+      "loss": 1.1497,
+      "step": 12620
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.258812082931057e-05,
+      "loss": 1.1669,
+      "step": 12625
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 5.250397083669979e-05,
+      "loss": 1.1949,
+      "step": 12630
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 5.241986425201459e-05,
+      "loss": 1.1882,
+      "step": 12635
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.233580115212233e-05,
+      "loss": 1.1623,
+      "step": 12640
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.2251781613850606e-05,
+      "loss": 1.1773,
+      "step": 12645
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.2167805713987184e-05,
+      "loss": 1.16,
+      "step": 12650
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.208387352927988e-05,
+      "loss": 1.1388,
+      "step": 12655
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1796875,
+      "learning_rate": 5.199998513643667e-05,
+      "loss": 1.1086,
+      "step": 12660
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.191614061212546e-05,
+      "loss": 1.1496,
+      "step": 12665
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.18323400329741e-05,
+      "loss": 1.1491,
+      "step": 12670
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.174858347557015e-05,
+      "loss": 1.1115,
+      "step": 12675
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.166487101646109e-05,
+      "loss": 1.2119,
+      "step": 12680
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 5.1581202732154035e-05,
+      "loss": 1.2117,
+      "step": 12685
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.14975786991157e-05,
+      "loss": 1.2826,
+      "step": 12690
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.141399899377245e-05,
+      "loss": 1.1498,
+      "step": 12695
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.133046369250998e-05,
+      "loss": 1.1839,
+      "step": 12700
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.124697287167355e-05,
+      "loss": 1.2456,
+      "step": 12705
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.116352660756771e-05,
+      "loss": 1.1911,
+      "step": 12710
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.1080124976456334e-05,
+      "loss": 1.2231,
+      "step": 12715
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 5.099676805456237e-05,
+      "loss": 1.2012,
+      "step": 12720
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1640625,
+      "learning_rate": 5.091345591806807e-05,
+      "loss": 1.1584,
+      "step": 12725
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 5.083018864311464e-05,
+      "loss": 1.1039,
+      "step": 12730
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.07469663058024e-05,
+      "loss": 1.205,
+      "step": 12735
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.066378898219044e-05,
+      "loss": 1.1802,
+      "step": 12740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.058065674829685e-05,
+      "loss": 1.1798,
+      "step": 12745
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 5.0497569680098445e-05,
+      "loss": 1.1172,
+      "step": 12750
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 5.0414527853530846e-05,
+      "loss": 1.0927,
+      "step": 12755
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.033153134448814e-05,
+      "loss": 1.1358,
+      "step": 12760
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.02485802288232e-05,
+      "loss": 1.1605,
+      "step": 12765
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.016567458234729e-05,
+      "loss": 1.225,
+      "step": 12770
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 5.008281448083021e-05,
+      "loss": 1.2231,
+      "step": 12775
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.000000000000002e-05,
+      "loss": 1.143,
+      "step": 12780
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.991723121554318e-05,
+      "loss": 1.1943,
+      "step": 12785
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.983450820310433e-05,
+      "loss": 1.1777,
+      "step": 12790
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1640625,
+      "learning_rate": 4.975183103828635e-05,
+      "loss": 1.1359,
+      "step": 12795
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.966919979665018e-05,
+      "loss": 1.1229,
+      "step": 12800
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.958661455371466e-05,
+      "loss": 1.1743,
+      "step": 12805
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.950407538495688e-05,
+      "loss": 1.1673,
+      "step": 12810
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.942158236581153e-05,
+      "loss": 1.2264,
+      "step": 12815
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 4.9339135571671344e-05,
+      "loss": 1.1127,
+      "step": 12820
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.9256735077886605e-05,
+      "loss": 1.1564,
+      "step": 12825
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.917438095976544e-05,
+      "loss": 1.1626,
+      "step": 12830
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.9092073292573545e-05,
+      "loss": 1.2197,
+      "step": 12835
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.900981215153419e-05,
+      "loss": 1.1686,
+      "step": 12840
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.892759761182801e-05,
+      "loss": 1.1162,
+      "step": 12845
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.884542974859319e-05,
+      "loss": 1.1793,
+      "step": 12850
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.8763308636925154e-05,
+      "loss": 1.1517,
+      "step": 12855
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.868123435187673e-05,
+      "loss": 1.1833,
+      "step": 12860
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1640625,
+      "learning_rate": 4.8599206968457746e-05,
+      "loss": 1.09,
+      "step": 12865
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.166015625,
+      "learning_rate": 4.851722656163534e-05,
+      "loss": 1.2365,
+      "step": 12870
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.8435293206333656e-05,
+      "loss": 1.2035,
+      "step": 12875
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.835340697743383e-05,
+      "loss": 1.2277,
+      "step": 12880
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.827156794977399e-05,
+      "loss": 1.115,
+      "step": 12885
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.818977619814893e-05,
+      "loss": 1.1453,
+      "step": 12890
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.810803179731056e-05,
+      "loss": 1.1685,
+      "step": 12895
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.802633482196719e-05,
+      "loss": 1.1615,
+      "step": 12900
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.7944685346784024e-05,
+      "loss": 1.1596,
+      "step": 12905
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.786308344638265e-05,
+      "loss": 1.1528,
+      "step": 12910
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.778152919534143e-05,
+      "loss": 1.1603,
+      "step": 12915
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.7700022668194925e-05,
+      "loss": 1.1256,
+      "step": 12920
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.7618563939434244e-05,
+      "loss": 1.1293,
+      "step": 12925
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.7537153083506716e-05,
+      "loss": 1.1085,
+      "step": 12930
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.745579017481596e-05,
+      "loss": 1.2238,
+      "step": 12935
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.737447528772181e-05,
+      "loss": 1.1304,
+      "step": 12940
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.729320849654018e-05,
+      "loss": 1.2142,
+      "step": 12945
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 4.721198987554296e-05,
+      "loss": 1.2238,
+      "step": 12950
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.7130819498958125e-05,
+      "loss": 1.1509,
+      "step": 12955
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.7049697440969496e-05,
+      "loss": 1.2085,
+      "step": 12960
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.696862377571676e-05,
+      "loss": 1.2026,
+      "step": 12965
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.68875985772954e-05,
+      "loss": 1.1621,
+      "step": 12970
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.68066219197565e-05,
+      "loss": 1.128,
+      "step": 12975
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 4.672569387710689e-05,
+      "loss": 1.2856,
+      "step": 12980
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.6644814523308924e-05,
+      "loss": 1.1252,
+      "step": 12985
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.6563983932280495e-05,
+      "loss": 1.1572,
+      "step": 12990
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.64832021778948e-05,
+      "loss": 1.1686,
+      "step": 12995
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.6402469333980656e-05,
+      "loss": 1.2313,
+      "step": 13000
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.632178547432192e-05,
+      "loss": 1.1801,
+      "step": 13005
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.624115067265784e-05,
+      "loss": 1.2349,
+      "step": 13010
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.616056500268268e-05,
+      "loss": 1.1818,
+      "step": 13015
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.608002853804604e-05,
+      "loss": 1.1842,
+      "step": 13020
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.5999541352352294e-05,
+      "loss": 1.1208,
+      "step": 13025
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.591910351916097e-05,
+      "loss": 1.0402,
+      "step": 13030
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.583871511198634e-05,
+      "loss": 1.187,
+      "step": 13035
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.575837620429762e-05,
+      "loss": 1.2104,
+      "step": 13040
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 4.567808686951872e-05,
+      "loss": 1.1241,
+      "step": 13045
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.166015625,
+      "learning_rate": 4.5597847181028296e-05,
+      "loss": 1.1456,
+      "step": 13050
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.551765721215964e-05,
+      "loss": 1.1684,
+      "step": 13055
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.5437517036200474e-05,
+      "loss": 1.1856,
+      "step": 13060
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.5357426726393146e-05,
+      "loss": 1.1849,
+      "step": 13065
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 4.52773863559344e-05,
+      "loss": 1.1626,
+      "step": 13070
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1640625,
+      "learning_rate": 4.5197395997975365e-05,
+      "loss": 1.1048,
+      "step": 13075
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.5117455725621304e-05,
+      "loss": 1.267,
+      "step": 13080
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.5037565611931995e-05,
+      "loss": 1.2042,
+      "step": 13085
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.4957725729921074e-05,
+      "loss": 1.144,
+      "step": 13090
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.177734375,
+      "learning_rate": 4.48779361525565e-05,
+      "loss": 1.1252,
+      "step": 13095
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.479819695276003e-05,
+      "loss": 1.2143,
+      "step": 13100
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.166015625,
+      "learning_rate": 4.471850820340766e-05,
+      "loss": 1.0556,
+      "step": 13105
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.463886997732901e-05,
+      "loss": 1.1739,
+      "step": 13110
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.455928234730774e-05,
+      "loss": 1.0889,
+      "step": 13115
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.447974538608107e-05,
+      "loss": 1.1132,
+      "step": 13120
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.4400259166340076e-05,
+      "loss": 1.1859,
+      "step": 13125
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.432082376072937e-05,
+      "loss": 1.1718,
+      "step": 13130
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 4.4241439241847205e-05,
+      "loss": 1.1603,
+      "step": 13135
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.416210568224519e-05,
+      "loss": 1.179,
+      "step": 13140
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.40828231544285e-05,
+      "loss": 1.2242,
+      "step": 13145
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.40035917308556e-05,
+      "loss": 1.1504,
+      "step": 13150
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.3924411483938254e-05,
+      "loss": 1.17,
+      "step": 13155
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 4.384528248604153e-05,
+      "loss": 1.1896,
+      "step": 13160
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.376620480948345e-05,
+      "loss": 1.1076,
+      "step": 13165
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.3687178526535444e-05,
+      "loss": 1.2358,
+      "step": 13170
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18359375,
+      "learning_rate": 4.360820370942168e-05,
+      "loss": 1.133,
+      "step": 13175
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.35292804303195e-05,
+      "loss": 1.1723,
+      "step": 13180
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 4.345040876135894e-05,
+      "loss": 1.1845,
+      "step": 13185
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.337158877462312e-05,
+      "loss": 1.1915,
+      "step": 13190
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.177734375,
+      "learning_rate": 4.3292820542147695e-05,
+      "loss": 1.1755,
+      "step": 13195
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 4.321410413592118e-05,
+      "loss": 1.1164,
+      "step": 13200
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.166015625,
+      "learning_rate": 4.3135439627884565e-05,
+      "loss": 1.1593,
+      "step": 13205
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.305682708993164e-05,
+      "loss": 1.1905,
+      "step": 13210
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.297826659390844e-05,
+      "loss": 1.2257,
+      "step": 13215
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.289975821161366e-05,
+      "loss": 1.041,
+      "step": 13220
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.282130201479818e-05,
+      "loss": 1.1574,
+      "step": 13225
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 4.274289807516532e-05,
+      "loss": 1.1792,
+      "step": 13230
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.266454646437057e-05,
+      "loss": 1.2165,
+      "step": 13235
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.258624725402165e-05,
+      "loss": 1.1101,
+      "step": 13240
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.250800051567837e-05,
+      "loss": 1.1152,
+      "step": 13245
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.166015625,
+      "learning_rate": 4.242980632085247e-05,
+      "loss": 1.0985,
+      "step": 13250
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.235166474100793e-05,
+      "loss": 1.1884,
+      "step": 13255
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.2273575847560364e-05,
+      "loss": 1.1598,
+      "step": 13260
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.219553971187744e-05,
+      "loss": 1.1828,
+      "step": 13265
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.21175564052784e-05,
+      "loss": 1.142,
+      "step": 13270
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.2039625999034505e-05,
+      "loss": 1.2091,
+      "step": 13275
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.18359375,
+      "learning_rate": 4.1961748564368356e-05,
+      "loss": 1.0797,
+      "step": 13280
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.1883924172454356e-05,
+      "loss": 1.2071,
+      "step": 13285
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.180615289441826e-05,
+      "loss": 1.1632,
+      "step": 13290
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.17284348013375e-05,
+      "loss": 1.1654,
+      "step": 13295
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 4.16507699642407e-05,
+      "loss": 1.1511,
+      "step": 13300
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.157315845410792e-05,
+      "loss": 1.2286,
+      "step": 13305
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.1495600341870375e-05,
+      "loss": 1.2273,
+      "step": 13310
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.141809569841068e-05,
+      "loss": 1.1825,
+      "step": 13315
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.134064459456235e-05,
+      "loss": 1.1725,
+      "step": 13320
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.126324710111012e-05,
+      "loss": 1.0658,
+      "step": 13325
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.118590328878973e-05,
+      "loss": 1.2339,
+      "step": 13330
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.1108613228287716e-05,
+      "loss": 1.1675,
+      "step": 13335
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.1031376990241644e-05,
+      "loss": 1.1443,
+      "step": 13340
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.095419464523982e-05,
+      "loss": 1.1169,
+      "step": 13345
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 4.087706626382136e-05,
+      "loss": 1.1747,
+      "step": 13350
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.079999191647589e-05,
+      "loss": 1.1475,
+      "step": 13355
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.072297167364393e-05,
+      "loss": 1.108,
+      "step": 13360
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.0646005605716266e-05,
+      "loss": 1.1033,
+      "step": 13365
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.0569093783034396e-05,
+      "loss": 1.1738,
+      "step": 13370
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.049223627589002e-05,
+      "loss": 1.1878,
+      "step": 13375
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 4.0415433154525484e-05,
+      "loss": 1.1707,
+      "step": 13380
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.033868448913316e-05,
+      "loss": 1.1536,
+      "step": 13385
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.177734375,
+      "learning_rate": 4.0261990349855827e-05,
+      "loss": 1.1218,
+      "step": 13390
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.018535080678626e-05,
+      "loss": 1.1605,
+      "step": 13395
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.0108765929967585e-05,
+      "loss": 1.0793,
+      "step": 13400
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 4.0032235789392716e-05,
+      "loss": 1.2144,
+      "step": 13405
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.99557604550047e-05,
+      "loss": 1.1918,
+      "step": 13410
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.987933999669642e-05,
+      "loss": 1.1431,
+      "step": 13415
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.98029744843107e-05,
+      "loss": 1.1728,
+      "step": 13420
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.972666398763997e-05,
+      "loss": 1.1354,
+      "step": 13425
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.965040857642656e-05,
+      "loss": 1.1052,
+      "step": 13430
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.9574208320362396e-05,
+      "loss": 1.2923,
+      "step": 13435
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.949806328908888e-05,
+      "loss": 1.1323,
+      "step": 13440
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.181640625,
+      "learning_rate": 3.942197355219721e-05,
+      "loss": 1.116,
+      "step": 13445
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.177734375,
+      "learning_rate": 3.9345939179227766e-05,
+      "loss": 1.2529,
+      "step": 13450
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.173828125,
+      "learning_rate": 3.926996023967052e-05,
+      "loss": 1.2199,
+      "step": 13455
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.16796875,
+      "learning_rate": 3.9194036802964604e-05,
+      "loss": 1.0931,
+      "step": 13460
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.911816893849869e-05,
+      "loss": 1.1403,
+      "step": 13465
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 3.9042356715610375e-05,
+      "loss": 1.1292,
+      "step": 13470
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.89666002035866e-05,
+      "loss": 1.2141,
+      "step": 13475
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 3.8890899471663244e-05,
+      "loss": 1.2205,
+      "step": 13480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.8815254589025405e-05,
+      "loss": 1.1424,
+      "step": 13485
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.873966562480692e-05,
+      "loss": 1.1041,
+      "step": 13490
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.866413264809069e-05,
+      "loss": 1.1941,
+      "step": 13495
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.858865572790826e-05,
+      "loss": 1.1208,
+      "step": 13500
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.16796875,
+      "learning_rate": 3.8513234933240216e-05,
+      "loss": 1.2222,
+      "step": 13505
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.843787033301559e-05,
+      "loss": 1.1738,
+      "step": 13510
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.836256199611218e-05,
+      "loss": 1.1422,
+      "step": 13515
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 3.828730999135639e-05,
+      "loss": 1.2219,
+      "step": 13520
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.821211438752299e-05,
+      "loss": 1.1706,
+      "step": 13525
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.813697525333544e-05,
+      "loss": 1.0983,
+      "step": 13530
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 3.806189265746536e-05,
+      "loss": 1.2105,
+      "step": 13535
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 3.7986866668532864e-05,
+      "loss": 1.2181,
+      "step": 13540
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.791189735510616e-05,
+      "loss": 1.2017,
+      "step": 13545
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.78369847857019e-05,
+      "loss": 1.1555,
+      "step": 13550
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.173828125,
+      "learning_rate": 3.7762129028784634e-05,
+      "loss": 1.2115,
+      "step": 13555
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.173828125,
+      "learning_rate": 3.768733015276717e-05,
+      "loss": 1.2153,
+      "step": 13560
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 3.761258822601015e-05,
+      "loss": 1.1868,
+      "step": 13565
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.75379033168224e-05,
+      "loss": 1.2226,
+      "step": 13570
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.7463275493460425e-05,
+      "loss": 1.2322,
+      "step": 13575
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.7388704824128696e-05,
+      "loss": 1.1601,
+      "step": 13580
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.73141913769793e-05,
+      "loss": 1.2101,
+      "step": 13585
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.723973522011226e-05,
+      "loss": 1.1101,
+      "step": 13590
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 3.7165336421575006e-05,
+      "loss": 1.1107,
+      "step": 13595
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.709099504936266e-05,
+      "loss": 1.1833,
+      "step": 13600
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.701671117141786e-05,
+      "loss": 1.1344,
+      "step": 13605
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.177734375,
+      "learning_rate": 3.6942484855630674e-05,
+      "loss": 1.2202,
+      "step": 13610
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.6868316169838615e-05,
+      "loss": 1.1184,
+      "step": 13615
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.173828125,
+      "learning_rate": 3.6794205181826404e-05,
+      "loss": 1.0868,
+      "step": 13620
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.672015195932618e-05,
+      "loss": 1.1645,
+      "step": 13625
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.664615657001711e-05,
+      "loss": 1.1464,
+      "step": 13630
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.657221908152576e-05,
+      "loss": 1.2247,
+      "step": 13635
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.649833956142552e-05,
+      "loss": 1.1753,
+      "step": 13640
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.177734375,
+      "learning_rate": 3.642451807723699e-05,
+      "loss": 1.1876,
+      "step": 13645
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.173828125,
+      "learning_rate": 3.635075469642753e-05,
+      "loss": 1.1502,
+      "step": 13650
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.6277049486411685e-05,
+      "loss": 1.1478,
+      "step": 13655
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.6203402514550545e-05,
+      "loss": 1.1439,
+      "step": 13660
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 3.6129813848152186e-05,
+      "loss": 1.1978,
+      "step": 13665
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.6056283554471185e-05,
+      "loss": 1.1789,
+      "step": 13670
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.598281170070905e-05,
+      "loss": 1.1936,
+      "step": 13675
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.590939835401363e-05,
+      "loss": 1.0905,
+      "step": 13680
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.177734375,
+      "learning_rate": 3.583604358147943e-05,
+      "loss": 1.1509,
+      "step": 13685
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.5762747450147384e-05,
+      "loss": 1.1356,
+      "step": 13690
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.5689510027004866e-05,
+      "loss": 1.2413,
+      "step": 13695
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.5616331378985524e-05,
+      "loss": 1.201,
+      "step": 13700
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.554321157296936e-05,
+      "loss": 1.0684,
+      "step": 13705
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.547015067578255e-05,
+      "loss": 1.0633,
+      "step": 13710
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.53971487541975e-05,
+      "loss": 1.1793,
+      "step": 13715
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.532420587493266e-05,
+      "loss": 1.178,
+      "step": 13720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 3.5251322104652485e-05,
+      "loss": 1.1171,
+      "step": 13725
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.517849750996751e-05,
+      "loss": 1.2193,
+      "step": 13730
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.5105732157434056e-05,
+      "loss": 1.126,
+      "step": 13735
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.503302611355449e-05,
+      "loss": 1.1239,
+      "step": 13740
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.496037944477676e-05,
+      "loss": 1.0908,
+      "step": 13745
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.4887792217494743e-05,
+      "loss": 1.2219,
+      "step": 13750
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.481526449804777e-05,
+      "loss": 1.1626,
+      "step": 13755
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.173828125,
+      "learning_rate": 3.4742796352721065e-05,
+      "loss": 1.1822,
+      "step": 13760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.467038784774516e-05,
+      "loss": 1.1902,
+      "step": 13765
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.45980390492962e-05,
+      "loss": 1.1101,
+      "step": 13770
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.452575002349574e-05,
+      "loss": 1.137,
+      "step": 13775
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.445352083641076e-05,
+      "loss": 1.1581,
+      "step": 13780
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.177734375,
+      "learning_rate": 3.43813515540534e-05,
+      "loss": 1.1684,
+      "step": 13785
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.430924224238119e-05,
+      "loss": 1.1619,
+      "step": 13790
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.423719296729683e-05,
+      "loss": 1.1864,
+      "step": 13795
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.416520379464812e-05,
+      "loss": 1.1556,
+      "step": 13800
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 3.409327479022796e-05,
+      "loss": 1.173,
+      "step": 13805
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 3.40214060197742e-05,
+      "loss": 1.1378,
+      "step": 13810
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 3.39495975489697e-05,
+      "loss": 1.1369,
+      "step": 13815
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.387784944344221e-05,
+      "loss": 1.1497,
+      "step": 13820
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 3.380616176876432e-05,
+      "loss": 1.214,
+      "step": 13825
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.373453459045329e-05,
+      "loss": 1.1564,
+      "step": 13830
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.366296797397125e-05,
+      "loss": 1.1412,
+      "step": 13835
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.359146198472478e-05,
+      "loss": 1.2285,
+      "step": 13840
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18359375,
+      "learning_rate": 3.35200166880653e-05,
+      "loss": 1.1703,
+      "step": 13845
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.177734375,
+      "learning_rate": 3.344863214928855e-05,
+      "loss": 1.1571,
+      "step": 13850
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.3377308433634856e-05,
+      "loss": 1.1294,
+      "step": 13855
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.330604560628885e-05,
+      "loss": 1.085,
+      "step": 13860
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.3234843732379706e-05,
+      "loss": 1.2972,
+      "step": 13865
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.181640625,
+      "learning_rate": 3.316370287698067e-05,
+      "loss": 1.1964,
+      "step": 13870
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.177734375,
+      "learning_rate": 3.309262310510937e-05,
+      "loss": 1.1041,
+      "step": 13875
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.302160448172755e-05,
+      "loss": 1.237,
+      "step": 13880
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 3.295064707174108e-05,
+      "loss": 1.1725,
+      "step": 13885
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 3.287975093999993e-05,
+      "loss": 1.1727,
+      "step": 13890
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.2808916151297954e-05,
+      "loss": 1.1962,
+      "step": 13895
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.273814277037303e-05,
+      "loss": 1.0735,
+      "step": 13900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 3.26674308619069e-05,
+      "loss": 1.2206,
+      "step": 13905
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.177734375,
+      "learning_rate": 3.259678049052517e-05,
+      "loss": 1.1535,
+      "step": 13910
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 3.252619172079706e-05,
+      "loss": 1.1734,
+      "step": 13915
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.177734375,
+      "learning_rate": 3.245566461723565e-05,
+      "loss": 1.1578,
+      "step": 13920
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.238519924429757e-05,
+      "loss": 1.119,
+      "step": 13925
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.231479566638311e-05,
+      "loss": 1.157,
+      "step": 13930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 3.2244453947835964e-05,
+      "loss": 1.176,
+      "step": 13935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.2174174152943416e-05,
+      "loss": 1.1568,
+      "step": 13940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.16796875,
+      "learning_rate": 3.210395634593602e-05,
+      "loss": 1.1072,
+      "step": 13945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.203380059098786e-05,
+      "loss": 1.1266,
+      "step": 13950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.196370695221615e-05,
+      "loss": 1.1422,
+      "step": 13955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.18936754936814e-05,
+      "loss": 1.2068,
+      "step": 13960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.182370627938728e-05,
+      "loss": 1.131,
+      "step": 13965
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.175379937328059e-05,
+      "loss": 1.2384,
+      "step": 13970
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.168395483925119e-05,
+      "loss": 1.2203,
+      "step": 13975
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18359375,
+      "learning_rate": 3.161417274113188e-05,
+      "loss": 1.1862,
+      "step": 13980
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.154445314269843e-05,
+      "loss": 1.1876,
+      "step": 13985
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 3.1474796107669524e-05,
+      "loss": 1.2134,
+      "step": 13990
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.140520169970667e-05,
+      "loss": 1.1462,
+      "step": 13995
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.133566998241403e-05,
+      "loss": 1.1342,
+      "step": 14000
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.1266201019338594e-05,
+      "loss": 1.1929,
+      "step": 14005
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.173828125,
+      "learning_rate": 3.1196794873969946e-05,
+      "loss": 1.1846,
+      "step": 14010
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.112745160974031e-05,
+      "loss": 1.2214,
+      "step": 14015
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.1058171290024326e-05,
+      "loss": 1.2702,
+      "step": 14020
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.098895397813927e-05,
+      "loss": 1.2535,
+      "step": 14025
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.189453125,
+      "learning_rate": 3.091979973734461e-05,
+      "loss": 1.1729,
+      "step": 14030
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.085070863084247e-05,
+      "loss": 1.1687,
+      "step": 14035
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.0781680721777e-05,
+      "loss": 1.1672,
+      "step": 14040
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.071271607323472e-05,
+      "loss": 1.1815,
+      "step": 14045
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.064381474824434e-05,
+      "loss": 1.2,
+      "step": 14050
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.0574976809776676e-05,
+      "loss": 1.1494,
+      "step": 14055
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 3.0506202320744582e-05,
+      "loss": 1.1557,
+      "step": 14060
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.0437491344002967e-05,
+      "loss": 1.0796,
+      "step": 14065
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.0368843942348668e-05,
+      "loss": 1.2548,
+      "step": 14070
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.173828125,
+      "learning_rate": 3.0300260178520456e-05,
+      "loss": 1.1587,
+      "step": 14075
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.0231740115198927e-05,
+      "loss": 1.1738,
+      "step": 14080
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.0163283815006404e-05,
+      "loss": 1.1717,
+      "step": 14085
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.181640625,
+      "learning_rate": 3.009489134050699e-05,
+      "loss": 1.1571,
+      "step": 14090
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.173828125,
+      "learning_rate": 3.0026562754206465e-05,
+      "loss": 1.2267,
+      "step": 14095
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.995829811855223e-05,
+      "loss": 1.183,
+      "step": 14100
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.9890097495933134e-05,
+      "loss": 1.1724,
+      "step": 14105
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.982196094867965e-05,
+      "loss": 1.1948,
+      "step": 14110
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.9753888539063623e-05,
+      "loss": 1.212,
+      "step": 14115
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.968588032929833e-05,
+      "loss": 1.1881,
+      "step": 14120
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.961793638153829e-05,
+      "loss": 1.1855,
+      "step": 14125
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.9550056757879375e-05,
+      "loss": 1.1817,
+      "step": 14130
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.9482241520358624e-05,
+      "loss": 1.1128,
+      "step": 14135
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.9414490730954292e-05,
+      "loss": 1.2449,
+      "step": 14140
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.9346804451585607e-05,
+      "loss": 1.1301,
+      "step": 14145
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 2.9279182744112965e-05,
+      "loss": 1.1794,
+      "step": 14150
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.9211625670337695e-05,
+      "loss": 1.2097,
+      "step": 14155
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.9144133292002064e-05,
+      "loss": 1.2003,
+      "step": 14160
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.907670567078924e-05,
+      "loss": 1.1768,
+      "step": 14165
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.900934286832312e-05,
+      "loss": 1.1687,
+      "step": 14170
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.8942044946168445e-05,
+      "loss": 1.1567,
+      "step": 14175
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.887481196583063e-05,
+      "loss": 1.1866,
+      "step": 14180
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.8807643988755782e-05,
+      "loss": 1.11,
+      "step": 14185
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.874054107633051e-05,
+      "loss": 1.1675,
+      "step": 14190
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.8673503289882007e-05,
+      "loss": 1.1883,
+      "step": 14195
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.860653069067798e-05,
+      "loss": 1.1704,
+      "step": 14200
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.8539623339926546e-05,
+      "loss": 1.1541,
+      "step": 14205
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.8472781298776097e-05,
+      "loss": 1.1938,
+      "step": 14210
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.8406004628315464e-05,
+      "loss": 1.1602,
+      "step": 14215
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.8339293389573664e-05,
+      "loss": 1.0625,
+      "step": 14220
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.8272647643519956e-05,
+      "loss": 1.1587,
+      "step": 14225
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.8206067451063666e-05,
+      "loss": 1.109,
+      "step": 14230
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 2.8139552873054297e-05,
+      "loss": 1.1802,
+      "step": 14235
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.8073103970281346e-05,
+      "loss": 1.1886,
+      "step": 14240
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.8006720803474273e-05,
+      "loss": 1.1782,
+      "step": 14245
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.7940403433302532e-05,
+      "loss": 1.1051,
+      "step": 14250
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.7874151920375325e-05,
+      "loss": 1.1943,
+      "step": 14255
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.780796632524174e-05,
+      "loss": 1.1166,
+      "step": 14260
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.774184670839063e-05,
+      "loss": 1.1605,
+      "step": 14265
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.7675793130250548e-05,
+      "loss": 1.2724,
+      "step": 14270
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.7609805651189614e-05,
+      "loss": 1.1806,
+      "step": 14275
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.7543884331515636e-05,
+      "loss": 1.1059,
+      "step": 14280
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.747802923147591e-05,
+      "loss": 1.1892,
+      "step": 14285
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1640625,
+      "learning_rate": 2.7412240411257262e-05,
+      "loss": 1.1662,
+      "step": 14290
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.7346517930985838e-05,
+      "loss": 1.1538,
+      "step": 14295
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.7280861850727236e-05,
+      "loss": 1.1386,
+      "step": 14300
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.7215272230486356e-05,
+      "loss": 1.1457,
+      "step": 14305
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.714974913020739e-05,
+      "loss": 1.0917,
+      "step": 14310
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.708429260977363e-05,
+      "loss": 1.1846,
+      "step": 14315
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.7018902729007632e-05,
+      "loss": 1.265,
+      "step": 14320
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.6953579547671004e-05,
+      "loss": 1.1437,
+      "step": 14325
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.6888323125464378e-05,
+      "loss": 1.1078,
+      "step": 14330
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.6823133522027423e-05,
+      "loss": 1.209,
+      "step": 14335
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.675801079693867e-05,
+      "loss": 1.1419,
+      "step": 14340
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.669295500971557e-05,
+      "loss": 1.1218,
+      "step": 14345
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.6627966219814405e-05,
+      "loss": 1.0754,
+      "step": 14350
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.656304448663025e-05,
+      "loss": 1.1922,
+      "step": 14355
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.6498189869496803e-05,
+      "loss": 1.1682,
+      "step": 14360
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.6433402427686494e-05,
+      "loss": 1.1014,
+      "step": 14365
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.6368682220410358e-05,
+      "loss": 1.0514,
+      "step": 14370
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.630402930681799e-05,
+      "loss": 1.2006,
+      "step": 14375
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.6239443745997406e-05,
+      "loss": 1.1759,
+      "step": 14380
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.6174925596975164e-05,
+      "loss": 1.1712,
+      "step": 14385
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.6110474918716176e-05,
+      "loss": 1.0893,
+      "step": 14390
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.6046091770123694e-05,
+      "loss": 1.1495,
+      "step": 14395
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.598177621003921e-05,
+      "loss": 1.2217,
+      "step": 14400
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.591752829724249e-05,
+      "loss": 1.1603,
+      "step": 14405
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.585334809045148e-05,
+      "loss": 1.09,
+      "step": 14410
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.5789235648322263e-05,
+      "loss": 1.14,
+      "step": 14415
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.572519102944889e-05,
+      "loss": 1.2157,
+      "step": 14420
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.5661214292363545e-05,
+      "loss": 1.151,
+      "step": 14425
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.5597305495536317e-05,
+      "loss": 1.135,
+      "step": 14430
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.55334646973752e-05,
+      "loss": 1.1895,
+      "step": 14435
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 2.5469691956226095e-05,
+      "loss": 1.1144,
+      "step": 14440
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.5405987330372605e-05,
+      "loss": 1.1324,
+      "step": 14445
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.534235087803616e-05,
+      "loss": 1.2282,
+      "step": 14450
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.5278782657375853e-05,
+      "loss": 1.1355,
+      "step": 14455
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.5215282726488476e-05,
+      "loss": 1.1699,
+      "step": 14460
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1640625,
+      "learning_rate": 2.515185114340829e-05,
+      "loss": 1.2136,
+      "step": 14465
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 2.50884879661072e-05,
+      "loss": 1.2011,
+      "step": 14470
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.5025193252494538e-05,
+      "loss": 1.122,
+      "step": 14475
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.4961967060417114e-05,
+      "loss": 1.2412,
+      "step": 14480
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.4898809447659034e-05,
+      "loss": 1.1561,
+      "step": 14485
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.4835720471941803e-05,
+      "loss": 1.1655,
+      "step": 14490
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.4772700190924158e-05,
+      "loss": 1.1939,
+      "step": 14495
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.4709748662202094e-05,
+      "loss": 1.1419,
+      "step": 14500
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.4646865943308694e-05,
+      "loss": 1.1741,
+      "step": 14505
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.4584052091714237e-05,
+      "loss": 1.1634,
+      "step": 14510
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.4521307164826002e-05,
+      "loss": 1.167,
+      "step": 14515
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.4458631219988325e-05,
+      "loss": 1.1621,
+      "step": 14520
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.439602431448249e-05,
+      "loss": 1.1669,
+      "step": 14525
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 2.4333486505526615e-05,
+      "loss": 1.1708,
+      "step": 14530
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.4271017850275757e-05,
+      "loss": 1.1187,
+      "step": 14535
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.4208618405821715e-05,
+      "loss": 1.1206,
+      "step": 14540
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.4146288229193102e-05,
+      "loss": 1.1602,
+      "step": 14545
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.408402737735511e-05,
+      "loss": 1.2055,
+      "step": 14550
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.4021835907209688e-05,
+      "loss": 1.1385,
+      "step": 14555
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.39597138755953e-05,
+      "loss": 1.1764,
+      "step": 14560
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.3897661339287036e-05,
+      "loss": 1.1861,
+      "step": 14565
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.383567835499634e-05,
+      "loss": 1.1928,
+      "step": 14570
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.3773764979371194e-05,
+      "loss": 1.194,
+      "step": 14575
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.371192126899593e-05,
+      "loss": 1.1139,
+      "step": 14580
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.3650147280391244e-05,
+      "loss": 1.1799,
+      "step": 14585
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.3588443070014023e-05,
+      "loss": 1.1999,
+      "step": 14590
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.3526808694257475e-05,
+      "loss": 1.1243,
+      "step": 14595
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.3465244209450943e-05,
+      "loss": 1.181,
+      "step": 14600
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.3403749671859908e-05,
+      "loss": 1.143,
+      "step": 14605
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.3342325137685938e-05,
+      "loss": 1.1431,
+      "step": 14610
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.3280970663066536e-05,
+      "loss": 1.1875,
+      "step": 14615
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.3219686304075293e-05,
+      "loss": 1.0855,
+      "step": 14620
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.3158472116721654e-05,
+      "loss": 1.1312,
+      "step": 14625
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.309732815695096e-05,
+      "loss": 1.1876,
+      "step": 14630
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.3036254480644325e-05,
+      "loss": 1.0694,
+      "step": 14635
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.297525114361868e-05,
+      "loss": 1.1991,
+      "step": 14640
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.291431820162665e-05,
+      "loss": 1.1271,
+      "step": 14645
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.2853455710356542e-05,
+      "loss": 1.2456,
+      "step": 14650
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.2792663725432218e-05,
+      "loss": 1.2064,
+      "step": 14655
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.2731942302413158e-05,
+      "loss": 1.1134,
+      "step": 14660
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.2671291496794344e-05,
+      "loss": 1.145,
+      "step": 14665
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.2610711364006243e-05,
+      "loss": 1.1989,
+      "step": 14670
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.2550201959414642e-05,
+      "loss": 1.1864,
+      "step": 14675
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.248976333832078e-05,
+      "loss": 1.1518,
+      "step": 14680
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.2429395555961163e-05,
+      "loss": 1.1685,
+      "step": 14685
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.236909866750758e-05,
+      "loss": 1.2221,
+      "step": 14690
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.230887272806703e-05,
+      "loss": 1.1326,
+      "step": 14695
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.2248717792681607e-05,
+      "loss": 1.1789,
+      "step": 14700
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.218863391632857e-05,
+      "loss": 1.1708,
+      "step": 14705
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.212862115392025e-05,
+      "loss": 1.1773,
+      "step": 14710
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.2068679560303963e-05,
+      "loss": 1.1225,
+      "step": 14715
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.2008809190261916e-05,
+      "loss": 1.2139,
+      "step": 14720
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.194901009851137e-05,
+      "loss": 1.1995,
+      "step": 14725
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.1889282339704286e-05,
+      "loss": 1.1509,
+      "step": 14730
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.1829625968427568e-05,
+      "loss": 1.0717,
+      "step": 14735
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.177004103920275e-05,
+      "loss": 1.1712,
+      "step": 14740
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.1710527606486153e-05,
+      "loss": 1.1454,
+      "step": 14745
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.1651085724668764e-05,
+      "loss": 1.1789,
+      "step": 14750
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.1591715448076143e-05,
+      "loss": 1.1542,
+      "step": 14755
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.1532416830968393e-05,
+      "loss": 1.2123,
+      "step": 14760
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.147318992754016e-05,
+      "loss": 1.108,
+      "step": 14765
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.1414034791920547e-05,
+      "loss": 1.1693,
+      "step": 14770
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.1354951478173068e-05,
+      "loss": 1.1721,
+      "step": 14775
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.129594004029555e-05,
+      "loss": 1.1436,
+      "step": 14780
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.1237000532220186e-05,
+      "loss": 1.0732,
+      "step": 14785
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.1178133007813418e-05,
+      "loss": 1.1218,
+      "step": 14790
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.111933752087588e-05,
+      "loss": 1.174,
+      "step": 14795
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.106061412514242e-05,
+      "loss": 1.2131,
+      "step": 14800
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.1001962874281878e-05,
+      "loss": 1.1411,
+      "step": 14805
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.0943383821897357e-05,
+      "loss": 1.0683,
+      "step": 14810
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.088487702152577e-05,
+      "loss": 1.1771,
+      "step": 14815
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.0826442526638147e-05,
+      "loss": 1.1603,
+      "step": 14820
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.07680803906393e-05,
+      "loss": 1.1607,
+      "step": 14825
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.07097906668681e-05,
+      "loss": 1.076,
+      "step": 14830
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.0651573408597035e-05,
+      "loss": 1.1833,
+      "step": 14835
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.0593428669032522e-05,
+      "loss": 1.2107,
+      "step": 14840
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.0535356501314572e-05,
+      "loss": 1.1851,
+      "step": 14845
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.047735695851697e-05,
+      "loss": 1.1028,
+      "step": 14850
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.0419430093647073e-05,
+      "loss": 1.1625,
+      "step": 14855
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.0361575959645886e-05,
+      "loss": 1.144,
+      "step": 14860
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.0303794609387815e-05,
+      "loss": 1.148,
+      "step": 14865
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.024608609568086e-05,
+      "loss": 1.1555,
+      "step": 14870
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.018845047126642e-05,
+      "loss": 1.1315,
+      "step": 14875
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.0130887788819275e-05,
+      "loss": 1.1403,
+      "step": 14880
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.007339810094756e-05,
+      "loss": 1.1378,
+      "step": 14885
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.001598146019261e-05,
+      "loss": 1.1418,
+      "step": 14890
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.9958637919029166e-05,
+      "loss": 1.2768,
+      "step": 14895
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.990136752986499e-05,
+      "loss": 1.1747,
+      "step": 14900
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.9844170345041124e-05,
+      "loss": 1.2055,
+      "step": 14905
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.9787046416831555e-05,
+      "loss": 1.135,
+      "step": 14910
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.9729995797443524e-05,
+      "loss": 1.2799,
+      "step": 14915
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.9673018539017075e-05,
+      "loss": 1.1108,
+      "step": 14920
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.9616114693625355e-05,
+      "loss": 1.1695,
+      "step": 14925
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.9559284313274252e-05,
+      "loss": 1.1855,
+      "step": 14930
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.9502527449902718e-05,
+      "loss": 1.1635,
+      "step": 14935
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.9445844155382353e-05,
+      "loss": 1.1343,
+      "step": 14940
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.9389234481517604e-05,
+      "loss": 1.098,
+      "step": 14945
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.9332698480045575e-05,
+      "loss": 1.186,
+      "step": 14950
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.9276236202636077e-05,
+      "loss": 1.1687,
+      "step": 14955
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.9219847700891558e-05,
+      "loss": 1.1899,
+      "step": 14960
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.9163533026347e-05,
+      "loss": 1.2169,
+      "step": 14965
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.9107292230469975e-05,
+      "loss": 1.2982,
+      "step": 14970
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.9051125364660416e-05,
+      "loss": 1.1029,
+      "step": 14975
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.8995032480250806e-05,
+      "loss": 1.1488,
+      "step": 14980
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.8939013628505963e-05,
+      "loss": 1.1138,
+      "step": 14985
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.8883068860623098e-05,
+      "loss": 1.1779,
+      "step": 14990
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.882719822773158e-05,
+      "loss": 1.2502,
+      "step": 14995
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.8771401780893217e-05,
+      "loss": 1.1273,
+      "step": 15000
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.871567957110183e-05,
+      "loss": 1.1555,
+      "step": 15005
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.866003164928355e-05,
+      "loss": 1.1827,
+      "step": 15010
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.860445806629645e-05,
+      "loss": 1.1184,
+      "step": 15015
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.8548958872930868e-05,
+      "loss": 1.1143,
+      "step": 15020
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.849353411990894e-05,
+      "loss": 1.1244,
+      "step": 15025
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.843818385788495e-05,
+      "loss": 1.1501,
+      "step": 15030
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.8382908137444955e-05,
+      "loss": 1.1973,
+      "step": 15035
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.8327707009106976e-05,
+      "loss": 1.1341,
+      "step": 15040
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.827258052332085e-05,
+      "loss": 1.1807,
+      "step": 15045
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.8217528730468193e-05,
+      "loss": 1.1366,
+      "step": 15050
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.816255168086236e-05,
+      "loss": 1.1123,
+      "step": 15055
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.810764942474834e-05,
+      "loss": 1.1146,
+      "step": 15060
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.805282201230285e-05,
+      "loss": 1.2433,
+      "step": 15065
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.799806949363415e-05,
+      "loss": 1.1296,
+      "step": 15070
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.7943391918782117e-05,
+      "loss": 1.1524,
+      "step": 15075
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.7888789337717993e-05,
+      "loss": 1.2408,
+      "step": 15080
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.7834261800344687e-05,
+      "loss": 1.0967,
+      "step": 15085
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.7779809356496337e-05,
+      "loss": 1.1405,
+      "step": 15090
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.7725432055938573e-05,
+      "loss": 1.1318,
+      "step": 15095
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.7671129948368236e-05,
+      "loss": 1.2091,
+      "step": 15100
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.7616903083413615e-05,
+      "loss": 1.2158,
+      "step": 15105
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.7562751510634046e-05,
+      "loss": 1.138,
+      "step": 15110
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 1.7508675279520194e-05,
+      "loss": 1.1527,
+      "step": 15115
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.7454674439493734e-05,
+      "loss": 1.2086,
+      "step": 15120
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.7400749039907617e-05,
+      "loss": 1.1983,
+      "step": 15125
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.734689913004568e-05,
+      "loss": 1.1715,
+      "step": 15130
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.7293124759122882e-05,
+      "loss": 1.1548,
+      "step": 15135
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.723942597628504e-05,
+      "loss": 1.1605,
+      "step": 15140
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.7185802830608978e-05,
+      "loss": 1.1958,
+      "step": 15145
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.7132255371102358e-05,
+      "loss": 1.1007,
+      "step": 15150
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.707878364670368e-05,
+      "loss": 1.08,
+      "step": 15155
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.7025387706282237e-05,
+      "loss": 1.1766,
+      "step": 15160
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.6972067598637973e-05,
+      "loss": 1.1762,
+      "step": 15165
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.6918823372501713e-05,
+      "loss": 1.1726,
+      "step": 15170
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.6865655076534726e-05,
+      "loss": 1.1205,
+      "step": 15175
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.681256275932903e-05,
+      "loss": 1.1836,
+      "step": 15180
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.6759546469407082e-05,
+      "loss": 1.1649,
+      "step": 15185
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.6706606255222035e-05,
+      "loss": 1.1409,
+      "step": 15190
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.6653742165157304e-05,
+      "loss": 1.1462,
+      "step": 15195
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.6600954247526913e-05,
+      "loss": 1.1258,
+      "step": 15200
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.654824255057509e-05,
+      "loss": 1.0984,
+      "step": 15205
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.6495607122476618e-05,
+      "loss": 1.2316,
+      "step": 15210
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.644304801133638e-05,
+      "loss": 1.164,
+      "step": 15215
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18359375,
+      "learning_rate": 1.639056526518964e-05,
+      "loss": 1.146,
+      "step": 15220
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.6338158932001747e-05,
+      "loss": 1.1302,
+      "step": 15225
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.628582905966839e-05,
+      "loss": 1.1973,
+      "step": 15230
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.62335756960152e-05,
+      "loss": 1.1351,
+      "step": 15235
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.6181398888797982e-05,
+      "loss": 1.1337,
+      "step": 15240
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.6129298685702586e-05,
+      "loss": 1.0967,
+      "step": 15245
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.607727513434475e-05,
+      "loss": 1.1385,
+      "step": 15250
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.6025328282270315e-05,
+      "loss": 1.1293,
+      "step": 15255
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.597345817695486e-05,
+      "loss": 1.2423,
+      "step": 15260
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.5921664865803955e-05,
+      "loss": 1.1544,
+      "step": 15265
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.5869948396152844e-05,
+      "loss": 1.1054,
+      "step": 15270
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.5818308815266734e-05,
+      "loss": 1.1825,
+      "step": 15275
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18359375,
+      "learning_rate": 1.5766746170340398e-05,
+      "loss": 1.1645,
+      "step": 15280
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.5715260508498387e-05,
+      "loss": 1.1658,
+      "step": 15285
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.566385187679479e-05,
+      "loss": 1.1489,
+      "step": 15290
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.5612520322213466e-05,
+      "loss": 1.1312,
+      "step": 15295
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.5561265891667654e-05,
+      "loss": 1.1652,
+      "step": 15300
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.551008863200024e-05,
+      "loss": 1.1594,
+      "step": 15305
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.545898858998346e-05,
+      "loss": 1.1512,
+      "step": 15310
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.5407965812319125e-05,
+      "loss": 1.1489,
+      "step": 15315
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.5357020345638306e-05,
+      "loss": 1.1412,
+      "step": 15320
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.530615223650147e-05,
+      "loss": 1.107,
+      "step": 15325
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.5255361531398382e-05,
+      "loss": 1.2239,
+      "step": 15330
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.5204648276748101e-05,
+      "loss": 1.1057,
+      "step": 15335
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.5154012518898786e-05,
+      "loss": 1.1529,
+      "step": 15340
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.5103454304127896e-05,
+      "loss": 1.1616,
+      "step": 15345
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.505297367864199e-05,
+      "loss": 1.1154,
+      "step": 15350
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.5002570688576622e-05,
+      "loss": 1.1565,
+      "step": 15355
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.4952245379996565e-05,
+      "loss": 1.1727,
+      "step": 15360
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.4901997798895439e-05,
+      "loss": 1.2091,
+      "step": 15365
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.4851827991195922e-05,
+      "loss": 1.1739,
+      "step": 15370
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 1.480173600274951e-05,
+      "loss": 1.1919,
+      "step": 15375
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.4751721879336755e-05,
+      "loss": 1.1953,
+      "step": 15380
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.4701785666666857e-05,
+      "loss": 1.1684,
+      "step": 15385
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.465192741037793e-05,
+      "loss": 1.173,
+      "step": 15390
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.4602147156036761e-05,
+      "loss": 1.1794,
+      "step": 15395
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.4552444949138954e-05,
+      "loss": 1.0663,
+      "step": 15400
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.4502820835108677e-05,
+      "loss": 1.2011,
+      "step": 15405
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.445327485929877e-05,
+      "loss": 1.2647,
+      "step": 15410
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.4403807066990694e-05,
+      "loss": 1.1713,
+      "step": 15415
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.4354417503394412e-05,
+      "loss": 1.1608,
+      "step": 15420
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.4305106213648366e-05,
+      "loss": 1.1372,
+      "step": 15425
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.425587324281954e-05,
+      "loss": 1.1375,
+      "step": 15430
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.4206718635903304e-05,
+      "loss": 1.1521,
+      "step": 15435
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.4157642437823337e-05,
+      "loss": 1.1572,
+      "step": 15440
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.4108644693431816e-05,
+      "loss": 1.1404,
+      "step": 15445
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 1.4059725447509053e-05,
+      "loss": 1.1601,
+      "step": 15450
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.4010884744763742e-05,
+      "loss": 1.131,
+      "step": 15455
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.3962122629832674e-05,
+      "loss": 1.1325,
+      "step": 15460
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.3913439147280982e-05,
+      "loss": 1.1895,
+      "step": 15465
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.3864834341601762e-05,
+      "loss": 1.1374,
+      "step": 15470
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.3816308257216325e-05,
+      "loss": 1.1611,
+      "step": 15475
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.376786093847392e-05,
+      "loss": 1.1853,
+      "step": 15480
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.3719492429651992e-05,
+      "loss": 1.1183,
+      "step": 15485
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.367120277495576e-05,
+      "loss": 1.2144,
+      "step": 15490
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.3622992018518532e-05,
+      "loss": 1.2231,
+      "step": 15495
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.3574860204401352e-05,
+      "loss": 1.1949,
+      "step": 15500
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.352680737659331e-05,
+      "loss": 1.1698,
+      "step": 15505
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.347883357901114e-05,
+      "loss": 1.1875,
+      "step": 15510
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.343093885549943e-05,
+      "loss": 1.2028,
+      "step": 15515
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.3383123249830488e-05,
+      "loss": 1.2279,
+      "step": 15520
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.3335386805704308e-05,
+      "loss": 1.1933,
+      "step": 15525
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.3287729566748553e-05,
+      "loss": 1.0972,
+      "step": 15530
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.3240151576518446e-05,
+      "loss": 1.2459,
+      "step": 15535
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.3192652878496869e-05,
+      "loss": 1.1849,
+      "step": 15540
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.3145233516094103e-05,
+      "loss": 1.1269,
+      "step": 15545
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.3097893532648108e-05,
+      "loss": 1.153,
+      "step": 15550
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.3050632971424116e-05,
+      "loss": 1.2444,
+      "step": 15555
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.3003451875614913e-05,
+      "loss": 1.178,
+      "step": 15560
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.2956350288340524e-05,
+      "loss": 1.145,
+      "step": 15565
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.2909328252648468e-05,
+      "loss": 1.132,
+      "step": 15570
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18359375,
+      "learning_rate": 1.2862385811513411e-05,
+      "loss": 1.2389,
+      "step": 15575
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.2815523007837393e-05,
+      "loss": 1.1711,
+      "step": 15580
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.2768739884449532e-05,
+      "loss": 1.1258,
+      "step": 15585
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 1.2722036484106314e-05,
+      "loss": 1.1635,
+      "step": 15590
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.2675412849491198e-05,
+      "loss": 1.1924,
+      "step": 15595
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.2628869023214807e-05,
+      "loss": 1.1779,
+      "step": 15600
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.2582405047814827e-05,
+      "loss": 1.1109,
+      "step": 15605
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.2536020965755968e-05,
+      "loss": 1.1628,
+      "step": 15610
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.2489716819429941e-05,
+      "loss": 1.1961,
+      "step": 15615
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.2443492651155332e-05,
+      "loss": 1.2238,
+      "step": 15620
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.2397348503177686e-05,
+      "loss": 1.1817,
+      "step": 15625
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.2351284417669428e-05,
+      "loss": 1.2217,
+      "step": 15630
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.2305300436729793e-05,
+      "loss": 1.1884,
+      "step": 15635
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.2259396602384776e-05,
+      "loss": 1.1059,
+      "step": 15640
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.2213572956587182e-05,
+      "loss": 1.1857,
+      "step": 15645
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.2167829541216459e-05,
+      "loss": 1.2348,
+      "step": 15650
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.2122166398078827e-05,
+      "loss": 1.1035,
+      "step": 15655
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.207658356890704e-05,
+      "loss": 1.2017,
+      "step": 15660
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.2031081095360542e-05,
+      "loss": 1.14,
+      "step": 15665
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.1985659019025208e-05,
+      "loss": 1.1444,
+      "step": 15670
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.1940317381413635e-05,
+      "loss": 1.1573,
+      "step": 15675
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.1895056223964707e-05,
+      "loss": 1.1853,
+      "step": 15680
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.1849875588043879e-05,
+      "loss": 1.1688,
+      "step": 15685
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.1804775514942946e-05,
+      "loss": 1.1349,
+      "step": 15690
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.1759756045880155e-05,
+      "loss": 1.2029,
+      "step": 15695
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.1714817221999975e-05,
+      "loss": 1.1874,
+      "step": 15700
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.1669959084373261e-05,
+      "loss": 1.1484,
+      "step": 15705
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.1625181673997087e-05,
+      "loss": 1.1444,
+      "step": 15710
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.1580485031794775e-05,
+      "loss": 1.2489,
+      "step": 15715
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.1535869198615824e-05,
+      "loss": 1.2006,
+      "step": 15720
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.1491334215235826e-05,
+      "loss": 1.1459,
+      "step": 15725
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.1446880122356541e-05,
+      "loss": 1.1095,
+      "step": 15730
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.1402506960605808e-05,
+      "loss": 1.2269,
+      "step": 15735
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.1358214770537479e-05,
+      "loss": 1.1456,
+      "step": 15740
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.1314003592631362e-05,
+      "loss": 1.1032,
+      "step": 15745
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.1269873467293313e-05,
+      "loss": 1.1541,
+      "step": 15750
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.1225824434855015e-05,
+      "loss": 1.1225,
+      "step": 15755
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.1181856535574154e-05,
+      "loss": 1.1481,
+      "step": 15760
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.1137969809634141e-05,
+      "loss": 1.1179,
+      "step": 15765
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.1094164297144294e-05,
+      "loss": 1.1666,
+      "step": 15770
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.185546875,
+      "learning_rate": 1.1050440038139664e-05,
+      "loss": 1.2062,
+      "step": 15775
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.1006797072581065e-05,
+      "loss": 1.1807,
+      "step": 15780
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.096323544035498e-05,
+      "loss": 1.2034,
+      "step": 15785
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.16015625,
+      "learning_rate": 1.0919755181273584e-05,
+      "loss": 1.0936,
+      "step": 15790
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.0876356335074688e-05,
+      "loss": 1.2038,
+      "step": 15795
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.0833038941421691e-05,
+      "loss": 1.1436,
+      "step": 15800
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.078980303990358e-05,
+      "loss": 1.1527,
+      "step": 15805
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.0746648670034765e-05,
+      "loss": 1.1403,
+      "step": 15810
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.0703575871255245e-05,
+      "loss": 1.1163,
+      "step": 15815
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.0660584682930441e-05,
+      "loss": 1.1601,
+      "step": 15820
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18359375,
+      "learning_rate": 1.061767514435119e-05,
+      "loss": 1.1993,
+      "step": 15825
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.057484729473367e-05,
+      "loss": 1.1943,
+      "step": 15830
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.0532101173219433e-05,
+      "loss": 1.2041,
+      "step": 15835
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.0489436818875331e-05,
+      "loss": 1.1697,
+      "step": 15840
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.044685427069354e-05,
+      "loss": 1.1648,
+      "step": 15845
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.0404353567591351e-05,
+      "loss": 1.1325,
+      "step": 15850
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.0361934748411373e-05,
+      "loss": 1.2276,
+      "step": 15855
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.031959785192127e-05,
+      "loss": 1.1691,
+      "step": 15860
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.027734291681397e-05,
+      "loss": 1.1603,
+      "step": 15865
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.0235169981707371e-05,
+      "loss": 1.2138,
+      "step": 15870
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.0193079085144485e-05,
+      "loss": 1.2319,
+      "step": 15875
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.0151070265593343e-05,
+      "loss": 1.192,
+      "step": 15880
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.0109143561446955e-05,
+      "loss": 1.1216,
+      "step": 15885
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.0067299011023312e-05,
+      "loss": 1.146,
+      "step": 15890
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.002553665256526e-05,
+      "loss": 1.1802,
+      "step": 15895
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.169921875,
+      "learning_rate": 9.983856524240588e-06,
+      "loss": 1.1755,
+      "step": 15900
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.942258664141901e-06,
+      "loss": 1.1626,
+      "step": 15905
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1796875,
+      "learning_rate": 9.900743110286648e-06,
+      "loss": 1.177,
+      "step": 15910
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.859309900616987e-06,
+      "loss": 1.1246,
+      "step": 15915
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.817959072999904e-06,
+      "loss": 1.2089,
+      "step": 15920
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.77669066522705e-06,
+      "loss": 1.1896,
+      "step": 15925
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.735504715014764e-06,
+      "loss": 1.186,
+      "step": 15930
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.694401260003993e-06,
+      "loss": 1.1843,
+      "step": 15935
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.653380337760354e-06,
+      "loss": 1.1994,
+      "step": 15940
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.612441985773912e-06,
+      "loss": 1.2127,
+      "step": 15945
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 9.571586241459462e-06,
+      "loss": 1.1564,
+      "step": 15950
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.530813142156104e-06,
+      "loss": 1.1474,
+      "step": 15955
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.490122725127538e-06,
+      "loss": 1.1543,
+      "step": 15960
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.449515027561828e-06,
+      "loss": 1.1735,
+      "step": 15965
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.166015625,
+      "learning_rate": 9.408990086571479e-06,
+      "loss": 1.153,
+      "step": 15970
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.368547939193373e-06,
+      "loss": 1.1329,
+      "step": 15975
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.328188622388644e-06,
+      "loss": 1.1603,
+      "step": 15980
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.287912173042811e-06,
+      "loss": 1.1533,
+      "step": 15985
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.24771862796563e-06,
+      "loss": 1.1527,
+      "step": 15990
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.207608023891101e-06,
+      "loss": 1.1962,
+      "step": 15995
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.16758039747736e-06,
+      "loss": 1.1322,
+      "step": 16000
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.127635785306788e-06,
+      "loss": 1.1952,
+      "step": 16005
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.087774223885848e-06,
+      "loss": 1.1636,
+      "step": 16010
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.169921875,
+      "learning_rate": 9.047995749645145e-06,
+      "loss": 1.1521,
+      "step": 16015
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.008300398939274e-06,
+      "loss": 1.27,
+      "step": 16020
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.968688208046938e-06,
+      "loss": 1.1969,
+      "step": 16025
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.929159213170802e-06,
+      "loss": 1.177,
+      "step": 16030
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.171875,
+      "learning_rate": 8.88971345043751e-06,
+      "loss": 1.0645,
+      "step": 16035
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.85035095589759e-06,
+      "loss": 1.0861,
+      "step": 16040
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.811071765525537e-06,
+      "loss": 1.1567,
+      "step": 16045
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.771875915219674e-06,
+      "loss": 1.1757,
+      "step": 16050
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.732763440802183e-06,
+      "loss": 1.1666,
+      "step": 16055
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.693734378019003e-06,
+      "loss": 1.1707,
+      "step": 16060
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.654788762539879e-06,
+      "loss": 1.1437,
+      "step": 16065
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.61592662995827e-06,
+      "loss": 1.1796,
+      "step": 16070
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 8.577148015791348e-06,
+      "loss": 1.161,
+      "step": 16075
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 8.538452955479982e-06,
+      "loss": 1.1462,
+      "step": 16080
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1640625,
+      "learning_rate": 8.499841484388615e-06,
+      "loss": 1.1826,
+      "step": 16085
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.461313637805335e-06,
+      "loss": 1.094,
+      "step": 16090
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.17578125,
+      "learning_rate": 8.422869450941807e-06,
+      "loss": 1.2613,
+      "step": 16095
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.384508958933256e-06,
+      "loss": 1.1028,
+      "step": 16100
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.346232196838333e-06,
+      "loss": 1.1893,
+      "step": 16105
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.308039199639261e-06,
+      "loss": 1.2048,
+      "step": 16110
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.269930002241655e-06,
+      "loss": 1.1932,
+      "step": 16115
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.231904639474585e-06,
+      "loss": 1.2113,
+      "step": 16120
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.193963146090432e-06,
+      "loss": 1.2568,
+      "step": 16125
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.156105556764982e-06,
+      "loss": 1.1633,
+      "step": 16130
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 8.118331906097342e-06,
+      "loss": 1.1578,
+      "step": 16135
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.080642228609891e-06,
+      "loss": 1.175,
+      "step": 16140
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.04303655874824e-06,
+      "loss": 1.228,
+      "step": 16145
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.005514930881242e-06,
+      "loss": 1.1441,
+      "step": 16150
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.968077379300964e-06,
+      "loss": 1.1286,
+      "step": 16155
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 7.930723938222595e-06,
+      "loss": 1.1959,
+      "step": 16160
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.893454641784493e-06,
+      "loss": 1.2265,
+      "step": 16165
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.856269524048066e-06,
+      "loss": 1.2004,
+      "step": 16170
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.819168618997819e-06,
+      "loss": 1.2231,
+      "step": 16175
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.782151960541306e-06,
+      "loss": 1.185,
+      "step": 16180
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.745219582509066e-06,
+      "loss": 1.1759,
+      "step": 16185
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.708371518654589e-06,
+      "loss": 1.2174,
+      "step": 16190
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.671607802654345e-06,
+      "loss": 1.1044,
+      "step": 16195
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.634928468107705e-06,
+      "loss": 1.2369,
+      "step": 16200
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 7.59833354853694e-06,
+      "loss": 1.1629,
+      "step": 16205
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.56182307738712e-06,
+      "loss": 1.2227,
+      "step": 16210
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.525397088026187e-06,
+      "loss": 1.2655,
+      "step": 16215
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.489055613744833e-06,
+      "loss": 1.1773,
+      "step": 16220
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.452798687756568e-06,
+      "loss": 1.1289,
+      "step": 16225
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.4166263431975635e-06,
+      "loss": 1.1562,
+      "step": 16230
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.380538613126709e-06,
+      "loss": 1.1724,
+      "step": 16235
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.344535530525598e-06,
+      "loss": 1.2103,
+      "step": 16240
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.3086171282984115e-06,
+      "loss": 1.243,
+      "step": 16245
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.272783439272013e-06,
+      "loss": 1.2094,
+      "step": 16250
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.237034496195727e-06,
+      "loss": 1.1357,
+      "step": 16255
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.201370331741541e-06,
+      "loss": 1.1915,
+      "step": 16260
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 7.165790978503895e-06,
+      "loss": 1.183,
+      "step": 16265
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.1302964689997535e-06,
+      "loss": 1.132,
+      "step": 16270
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.0948868356685125e-06,
+      "loss": 1.2421,
+      "step": 16275
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.059562110871998e-06,
+      "loss": 1.0783,
+      "step": 16280
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.024322326894461e-06,
+      "loss": 1.2316,
+      "step": 16285
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.989167515942519e-06,
+      "loss": 1.2171,
+      "step": 16290
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.954097710145091e-06,
+      "loss": 1.2011,
+      "step": 16295
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.919112941553452e-06,
+      "loss": 1.2027,
+      "step": 16300
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.884213242141136e-06,
+      "loss": 1.1424,
+      "step": 16305
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.849398643803962e-06,
+      "loss": 1.181,
+      "step": 16310
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.814669178359912e-06,
+      "loss": 1.1685,
+      "step": 16315
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.7800248775492185e-06,
+      "loss": 1.1517,
+      "step": 16320
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.745465773034243e-06,
+      "loss": 1.1752,
+      "step": 16325
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.16796875,
+      "learning_rate": 6.710991896399521e-06,
+      "loss": 1.1186,
+      "step": 16330
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 6.676603279151683e-06,
+      "loss": 1.1278,
+      "step": 16335
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.642299952719388e-06,
+      "loss": 1.2376,
+      "step": 16340
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.6080819484534265e-06,
+      "loss": 1.1657,
+      "step": 16345
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.573949297626558e-06,
+      "loss": 1.1629,
+      "step": 16350
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.539902031433565e-06,
+      "loss": 1.1644,
+      "step": 16355
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.16796875,
+      "learning_rate": 6.505940180991155e-06,
+      "loss": 1.1675,
+      "step": 16360
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.472063777338e-06,
+      "loss": 1.1836,
+      "step": 16365
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.43827285143469e-06,
+      "loss": 1.1358,
+      "step": 16370
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 6.404567434163688e-06,
+      "loss": 1.1718,
+      "step": 16375
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.370947556329265e-06,
+      "loss": 1.1502,
+      "step": 16380
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.337413248657576e-06,
+      "loss": 1.1677,
+      "step": 16385
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.303964541796547e-06,
+      "loss": 1.1864,
+      "step": 16390
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.270601466315873e-06,
+      "loss": 1.1558,
+      "step": 16395
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.237324052706961e-06,
+      "loss": 1.1839,
+      "step": 16400
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 6.2041323313829725e-06,
+      "loss": 1.1195,
+      "step": 16405
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.171026332678731e-06,
+      "loss": 1.1968,
+      "step": 16410
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 6.138006086850734e-06,
+      "loss": 1.1062,
+      "step": 16415
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.105071624077053e-06,
+      "loss": 1.1991,
+      "step": 16420
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.07222297445742e-06,
+      "loss": 1.1643,
+      "step": 16425
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.039460168013122e-06,
+      "loss": 1.1206,
+      "step": 16430
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.173828125,
+      "learning_rate": 6.006783234686975e-06,
+      "loss": 1.1193,
+      "step": 16435
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.974192204343354e-06,
+      "loss": 1.2072,
+      "step": 16440
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.941687106768068e-06,
+      "loss": 1.1326,
+      "step": 16445
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.909267971668419e-06,
+      "loss": 1.1982,
+      "step": 16450
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.876934828673153e-06,
+      "loss": 1.1706,
+      "step": 16455
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.8446877073324305e-06,
+      "loss": 1.1214,
+      "step": 16460
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 5.812526637117765e-06,
+      "loss": 1.1113,
+      "step": 16465
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.780451647422036e-06,
+      "loss": 1.2201,
+      "step": 16470
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.748462767559471e-06,
+      "loss": 1.1673,
+      "step": 16475
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.716560026765594e-06,
+      "loss": 1.1072,
+      "step": 16480
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.684743454197173e-06,
+      "loss": 1.1669,
+      "step": 16485
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.6530130789322565e-06,
+      "loss": 1.2506,
+      "step": 16490
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.6213689299701055e-06,
+      "loss": 1.2053,
+      "step": 16495
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.589811036231196e-06,
+      "loss": 1.1743,
+      "step": 16500
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.558339426557124e-06,
+      "loss": 1.1414,
+      "step": 16505
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.5269541297106705e-06,
+      "loss": 1.1321,
+      "step": 16510
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.495655174375724e-06,
+      "loss": 1.0863,
+      "step": 16515
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.464442589157259e-06,
+      "loss": 1.1304,
+      "step": 16520
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.433316402581323e-06,
+      "loss": 1.1899,
+      "step": 16525
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.402276643094961e-06,
+      "loss": 1.1078,
+      "step": 16530
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.166015625,
+      "learning_rate": 5.371323339066303e-06,
+      "loss": 1.1161,
+      "step": 16535
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.177734375,
+      "learning_rate": 5.340456518784387e-06,
+      "loss": 1.2131,
+      "step": 16540
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 5.30967621045928e-06,
+      "loss": 1.1079,
+      "step": 16545
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.2789824422219246e-06,
+      "loss": 1.2354,
+      "step": 16550
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.248375242124215e-06,
+      "loss": 1.1894,
+      "step": 16555
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.217854638138897e-06,
+      "loss": 1.1514,
+      "step": 16560
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.187420658159614e-06,
+      "loss": 1.1632,
+      "step": 16565
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.157073330000794e-06,
+      "loss": 1.1349,
+      "step": 16570
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18359375,
+      "learning_rate": 5.126812681397697e-06,
+      "loss": 1.1648,
+      "step": 16575
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.096638740006365e-06,
+      "loss": 1.2359,
+      "step": 16580
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 5.066551533403607e-06,
+      "loss": 1.1126,
+      "step": 16585
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.171875,
+      "learning_rate": 5.036551089086916e-06,
+      "loss": 1.1309,
+      "step": 16590
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.006637434474537e-06,
+      "loss": 1.1842,
+      "step": 16595
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1796875,
+      "learning_rate": 4.976810596905379e-06,
+      "loss": 1.1839,
+      "step": 16600
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.947070603639015e-06,
+      "loss": 1.1687,
+      "step": 16605
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.917417481855657e-06,
+      "loss": 1.2529,
+      "step": 16610
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.185546875,
+      "learning_rate": 4.887851258656062e-06,
+      "loss": 1.2056,
+      "step": 16615
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.858371961061647e-06,
+      "loss": 1.1376,
+      "step": 16620
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.828979616014351e-06,
+      "loss": 1.2936,
+      "step": 16625
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.79967425037664e-06,
+      "loss": 1.1438,
+      "step": 16630
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.19921875,
+      "learning_rate": 4.770455890931469e-06,
+      "loss": 1.1084,
+      "step": 16635
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.7413245643823435e-06,
+      "loss": 1.2458,
+      "step": 16640
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.712280297353156e-06,
+      "loss": 1.1415,
+      "step": 16645
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 4.6833231163882716e-06,
+      "loss": 1.1359,
+      "step": 16650
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.654453047952434e-06,
+      "loss": 1.1625,
+      "step": 16655
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.6256701184308135e-06,
+      "loss": 1.103,
+      "step": 16660
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.5969743541289025e-06,
+      "loss": 1.1005,
+      "step": 16665
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.5683657812725655e-06,
+      "loss": 1.1641,
+      "step": 16670
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.5398444260079555e-06,
+      "loss": 1.1248,
+      "step": 16675
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.511410314401509e-06,
+      "loss": 1.235,
+      "step": 16680
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.483063472439952e-06,
+      "loss": 1.1876,
+      "step": 16685
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 4.4548039260302486e-06,
+      "loss": 1.1554,
+      "step": 16690
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.426631700999584e-06,
+      "loss": 1.1349,
+      "step": 16695
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.398546823095295e-06,
+      "loss": 1.1381,
+      "step": 16700
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.370549317984951e-06,
+      "loss": 1.163,
+      "step": 16705
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1796875,
+      "learning_rate": 4.342639211256227e-06,
+      "loss": 1.1566,
+      "step": 16710
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.314816528416943e-06,
+      "loss": 1.223,
+      "step": 16715
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 4.287081294895001e-06,
+      "loss": 1.1395,
+      "step": 16720
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.2594335360384264e-06,
+      "loss": 1.1159,
+      "step": 16725
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.231873277115217e-06,
+      "loss": 1.169,
+      "step": 16730
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.204400543313503e-06,
+      "loss": 1.1806,
+      "step": 16735
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.17701535974131e-06,
+      "loss": 1.1525,
+      "step": 16740
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.149717751426762e-06,
+      "loss": 1.2027,
+      "step": 16745
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.122507743317861e-06,
+      "loss": 1.1434,
+      "step": 16750
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 4.095385360282611e-06,
+      "loss": 1.1302,
+      "step": 16755
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.068350627108875e-06,
+      "loss": 1.0971,
+      "step": 16760
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 4.041403568504443e-06,
+      "loss": 1.231,
+      "step": 16765
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.166015625,
+      "learning_rate": 4.014544209096971e-06,
+      "loss": 1.2159,
+      "step": 16770
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.98777257343399e-06,
+      "loss": 1.1385,
+      "step": 16775
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 3.961088685982806e-06,
+      "loss": 1.1849,
+      "step": 16780
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.9344925711305655e-06,
+      "loss": 1.1613,
+      "step": 16785
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.907984253184182e-06,
+      "loss": 1.1847,
+      "step": 16790
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.881563756370344e-06,
+      "loss": 1.1473,
+      "step": 16795
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.16015625,
+      "learning_rate": 3.855231104835477e-06,
+      "loss": 1.0595,
+      "step": 16800
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.16796875,
+      "learning_rate": 3.828986322645678e-06,
+      "loss": 1.1727,
+      "step": 16805
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.802829433786825e-06,
+      "loss": 1.1347,
+      "step": 16810
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.173828125,
+      "learning_rate": 3.7767604621643747e-06,
+      "loss": 1.2203,
+      "step": 16815
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.16796875,
+      "learning_rate": 3.750779431603513e-06,
+      "loss": 1.1079,
+      "step": 16820
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.7248863658489787e-06,
+      "loss": 1.2136,
+      "step": 16825
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.699081288565187e-06,
+      "loss": 1.1902,
+      "step": 16830
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.6733642233360952e-06,
+      "loss": 1.1588,
+      "step": 16835
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.647735193665247e-06,
+      "loss": 1.2034,
+      "step": 16840
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.173828125,
+      "learning_rate": 3.622194222975672e-06,
+      "loss": 1.2295,
+      "step": 16845
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.596741334610032e-06,
+      "loss": 1.1129,
+      "step": 16850
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.5713765518303743e-06,
+      "loss": 1.1785,
+      "step": 16855
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.177734375,
+      "learning_rate": 3.546099897818289e-06,
+      "loss": 1.243,
+      "step": 16860
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.5209113956747862e-06,
+      "loss": 1.1932,
+      "step": 16865
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.4958110684203405e-06,
+      "loss": 1.1639,
+      "step": 16870
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.177734375,
+      "learning_rate": 3.4707989389948457e-06,
+      "loss": 1.1287,
+      "step": 16875
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.4458750302575503e-06,
+      "loss": 1.1221,
+      "step": 16880
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.421039364987122e-06,
+      "loss": 1.0995,
+      "step": 16885
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.396291965881515e-06,
+      "loss": 1.1231,
+      "step": 16890
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.3716328555581154e-06,
+      "loss": 1.2138,
+      "step": 16895
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.3470620565535294e-06,
+      "loss": 1.2367,
+      "step": 16900
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.322579591323682e-06,
+      "loss": 1.2178,
+      "step": 16905
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.298185482243765e-06,
+      "loss": 1.1914,
+      "step": 16910
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 3.2738797516082663e-06,
+      "loss": 1.183,
+      "step": 16915
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.2496624216308057e-06,
+      "loss": 1.204,
+      "step": 16920
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.225533514444312e-06,
+      "loss": 1.17,
+      "step": 16925
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 3.2014930521008126e-06,
+      "loss": 1.0692,
+      "step": 16930
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.1775410565715866e-06,
+      "loss": 1.0283,
+      "step": 16935
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.1536775497470028e-06,
+      "loss": 1.1455,
+      "step": 16940
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.173828125,
+      "learning_rate": 3.1299025534365813e-06,
+      "loss": 1.1927,
+      "step": 16945
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.1062160893689297e-06,
+      "loss": 1.19,
+      "step": 16950
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.082618179191765e-06,
+      "loss": 1.1795,
+      "step": 16955
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.0591088444718697e-06,
+      "loss": 1.155,
+      "step": 16960
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 3.035688106695067e-06,
+      "loss": 1.1472,
+      "step": 16965
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.0123559872662244e-06,
+      "loss": 1.1495,
+      "step": 16970
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.989112507509195e-06,
+      "loss": 1.2443,
+      "step": 16975
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.9659576886668317e-06,
+      "loss": 1.2309,
+      "step": 16980
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.9428915519009616e-06,
+      "loss": 1.1494,
+      "step": 16985
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.919914118292366e-06,
+      "loss": 1.1678,
+      "step": 16990
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.8970254088407368e-06,
+      "loss": 1.1647,
+      "step": 16995
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.8742254444647178e-06,
+      "loss": 1.1213,
+      "step": 17000
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.8515142460018075e-06,
+      "loss": 1.1535,
+      "step": 17005
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.8288918342084027e-06,
+      "loss": 1.1511,
+      "step": 17010
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.8063582297597313e-06,
+      "loss": 1.1657,
+      "step": 17015
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.7839134532498977e-06,
+      "loss": 1.2301,
+      "step": 17020
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.761557525191771e-06,
+      "loss": 1.0988,
+      "step": 17025
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.739290466017075e-06,
+      "loss": 1.2522,
+      "step": 17030
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.717112296076263e-06,
+      "loss": 1.1256,
+      "step": 17035
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.6950230356385996e-06,
+      "loss": 1.1679,
+      "step": 17040
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.6730227048920365e-06,
+      "loss": 1.1421,
+      "step": 17045
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.6511113239433004e-06,
+      "loss": 1.1202,
+      "step": 17050
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.6292889128178065e-06,
+      "loss": 1.1374,
+      "step": 17055
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.6075554914596456e-06,
+      "loss": 1.1719,
+      "step": 17060
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.5859110797315846e-06,
+      "loss": 1.2577,
+      "step": 17065
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.564355697415055e-06,
+      "loss": 1.2182,
+      "step": 17070
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.5428893642101216e-06,
+      "loss": 1.1502,
+      "step": 17075
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.5215120997354235e-06,
+      "loss": 1.2016,
+      "step": 17080
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.5002239235282664e-06,
+      "loss": 1.1663,
+      "step": 17085
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.479024855044476e-06,
+      "loss": 1.1409,
+      "step": 17090
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.4579149136584876e-06,
+      "loss": 1.2097,
+      "step": 17095
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.436894118663213e-06,
+      "loss": 1.1563,
+      "step": 17100
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.415962489270196e-06,
+      "loss": 1.1454,
+      "step": 17105
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.395120044609367e-06,
+      "loss": 1.125,
+      "step": 17110
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.374366803729244e-06,
+      "loss": 1.1371,
+      "step": 17115
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.3537027855967676e-06,
+      "loss": 1.1338,
+      "step": 17120
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.333128009097374e-06,
+      "loss": 1.2073,
+      "step": 17125
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.312642493034889e-06,
+      "loss": 1.0865,
+      "step": 17130
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.2922462561316026e-06,
+      "loss": 1.1977,
+      "step": 17135
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.271939317028182e-06,
+      "loss": 1.1992,
+      "step": 17140
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.251721694283726e-06,
+      "loss": 1.1972,
+      "step": 17145
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.231593406375654e-06,
+      "loss": 1.2196,
+      "step": 17150
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.2115544716997617e-06,
+      "loss": 1.2578,
+      "step": 17155
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.1916049085701886e-06,
+      "loss": 1.1821,
+      "step": 17160
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.171744735219372e-06,
+      "loss": 1.1414,
+      "step": 17165
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.1519739697980933e-06,
+      "loss": 1.1677,
+      "step": 17170
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.1322926303753877e-06,
+      "loss": 1.1222,
+      "step": 17175
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.1127007349385773e-06,
+      "loss": 1.1948,
+      "step": 17180
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.0931983013932065e-06,
+      "loss": 1.1847,
+      "step": 17185
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.0737853475631285e-06,
+      "loss": 1.1168,
+      "step": 17190
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.054461891190351e-06,
+      "loss": 1.1159,
+      "step": 17195
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.0352279499351147e-06,
+      "loss": 1.2277,
+      "step": 17200
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.016083541375824e-06,
+      "loss": 1.0641,
+      "step": 17205
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.9970286830091167e-06,
+      "loss": 1.2393,
+      "step": 17210
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.9780633922497404e-06,
+      "loss": 1.1617,
+      "step": 17215
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.9591876864305856e-06,
+      "loss": 1.159,
+      "step": 17220
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.9404015828026756e-06,
+      "loss": 1.0684,
+      "step": 17225
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.921705098535165e-06,
+      "loss": 1.1426,
+      "step": 17230
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.9030982507152628e-06,
+      "loss": 1.1149,
+      "step": 17235
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.8845810563483002e-06,
+      "loss": 1.0976,
+      "step": 17240
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.8661535323576286e-06,
+      "loss": 1.0789,
+      "step": 17245
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.8478156955846871e-06,
+      "loss": 1.1253,
+      "step": 17250
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.829567562788914e-06,
+      "loss": 1.1965,
+      "step": 17255
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.8114091506477914e-06,
+      "loss": 1.1672,
+      "step": 17260
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.7933404757567884e-06,
+      "loss": 1.1776,
+      "step": 17265
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.77536155462934e-06,
+      "loss": 1.1156,
+      "step": 17270
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.7574724036969026e-06,
+      "loss": 1.1717,
+      "step": 17275
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.7396730393088644e-06,
+      "loss": 1.1301,
+      "step": 17280
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.7219634777325354e-06,
+      "loss": 1.2481,
+      "step": 17285
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.7043437351531798e-06,
+      "loss": 1.1498,
+      "step": 17290
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.686813827673961e-06,
+      "loss": 1.1509,
+      "step": 17295
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.6693737713159519e-06,
+      "loss": 1.1795,
+      "step": 17300
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.6520235820180918e-06,
+      "loss": 1.2174,
+      "step": 17305
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.6347632756371855e-06,
+      "loss": 1.1921,
+      "step": 17310
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.6175928679479258e-06,
+      "loss": 1.1467,
+      "step": 17315
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.6005123746428042e-06,
+      "loss": 1.1252,
+      "step": 17320
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.583521811332145e-06,
+      "loss": 1.216,
+      "step": 17325
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.5666211935441044e-06,
+      "loss": 1.1271,
+      "step": 17330
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.5498105367246274e-06,
+      "loss": 1.1873,
+      "step": 17335
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.533089856237413e-06,
+      "loss": 1.2123,
+      "step": 17340
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.5164591673639595e-06,
+      "loss": 1.267,
+      "step": 17345
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.4999184853035197e-06,
+      "loss": 1.1982,
+      "step": 17350
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.4834678251730572e-06,
+      "loss": 1.1584,
+      "step": 17355
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.4671072020073118e-06,
+      "loss": 1.2699,
+      "step": 17360
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.4508366307586786e-06,
+      "loss": 1.2251,
+      "step": 17365
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.4346561262972957e-06,
+      "loss": 1.1924,
+      "step": 17370
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.4185657034109567e-06,
+      "loss": 1.2191,
+      "step": 17375
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.402565376805165e-06,
+      "loss": 1.1823,
+      "step": 17380
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.3866551611030342e-06,
+      "loss": 1.1107,
+      "step": 17385
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.3708350708453554e-06,
+      "loss": 1.2043,
+      "step": 17390
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.3551051204905408e-06,
+      "loss": 1.1392,
+      "step": 17395
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.339465324414635e-06,
+      "loss": 1.122,
+      "step": 17400
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.323915696911271e-06,
+      "loss": 1.1663,
+      "step": 17405
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.30845625219167e-06,
+      "loss": 1.2134,
+      "step": 17410
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.2930870043846633e-06,
+      "loss": 1.1754,
+      "step": 17415
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.277807967536615e-06,
+      "loss": 1.1718,
+      "step": 17420
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.2626191556114664e-06,
+      "loss": 1.1119,
+      "step": 17425
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.2475205824906689e-06,
+      "loss": 1.1386,
+      "step": 17430
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.23251226197324e-06,
+      "loss": 1.2457,
+      "step": 17435
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.2175942077756963e-06,
+      "loss": 1.1837,
+      "step": 17440
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.2027664335320655e-06,
+      "loss": 1.2327,
+      "step": 17445
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.1880289527938404e-06,
+      "loss": 1.2002,
+      "step": 17450
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.173381779030014e-06,
+      "loss": 1.1953,
+      "step": 17455
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.1588249256270334e-06,
+      "loss": 1.2091,
+      "step": 17460
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.1443584058888347e-06,
+      "loss": 1.1761,
+      "step": 17465
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.1299822330367305e-06,
+      "loss": 1.1212,
+      "step": 17470
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.1156964202095332e-06,
+      "loss": 1.2025,
+      "step": 17475
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.1015009804634102e-06,
+      "loss": 1.2314,
+      "step": 17480
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.0873959267719947e-06,
+      "loss": 1.1906,
+      "step": 17485
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.0733812720262525e-06,
+      "loss": 1.2927,
+      "step": 17490
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.059457029034583e-06,
+      "loss": 1.1541,
+      "step": 17495
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.045623210522717e-06,
+      "loss": 1.1553,
+      "step": 17500
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.0318798291337861e-06,
+      "loss": 1.1507,
+      "step": 17505
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.0182268974282206e-06,
+      "loss": 1.1781,
+      "step": 17510
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.0046644278838057e-06,
+      "loss": 1.1769,
+      "step": 17515
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.911924328956823e-07,
+      "loss": 1.1506,
+      "step": 17520
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.778109247762568e-07,
+      "loss": 1.1506,
+      "step": 17525
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.645199157552688e-07,
+      "loss": 1.2123,
+      "step": 17530
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.513194179797235e-07,
+      "loss": 1.1271,
+      "step": 17535
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.382094435139155e-07,
+      "loss": 1.1028,
+      "step": 17540
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.251900043394269e-07,
+      "loss": 1.1772,
+      "step": 17545
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 9.122611123550839e-07,
+      "loss": 1.1611,
+      "step": 17550
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.99422779376935e-07,
+      "loss": 1.1994,
+      "step": 17555
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.866750171382942e-07,
+      "loss": 1.1474,
+      "step": 17560
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 8.740178372896979e-07,
+      "loss": 1.1702,
+      "step": 17565
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 8.614512513988815e-07,
+      "loss": 1.1506,
+      "step": 17570
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 8.489752709508025e-07,
+      "loss": 1.1921,
+      "step": 17575
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.365899073475958e-07,
+      "loss": 1.132,
+      "step": 17580
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.242951719085734e-07,
+      "loss": 1.2136,
+      "step": 17585
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 8.120910758702582e-07,
+      "loss": 1.1054,
+      "step": 17590
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.999776303862838e-07,
+      "loss": 1.0366,
+      "step": 17595
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.879548465274833e-07,
+      "loss": 1.1015,
+      "step": 17600
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.760227352817895e-07,
+      "loss": 1.2512,
+      "step": 17605
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.641813075543014e-07,
+      "loss": 1.1721,
+      "step": 17610
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.524305741672178e-07,
+      "loss": 1.194,
+      "step": 17615
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 7.40770545859848e-07,
+      "loss": 1.158,
+      "step": 17620
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.16796875,
+      "learning_rate": 7.292012332886344e-07,
+      "loss": 1.0983,
+      "step": 17625
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.162109375,
+      "learning_rate": 7.177226470270748e-07,
+      "loss": 1.1755,
+      "step": 17630
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.063347975657664e-07,
+      "loss": 1.1139,
+      "step": 17635
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.950376953123727e-07,
+      "loss": 1.1781,
+      "step": 17640
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.166015625,
+      "learning_rate": 6.838313505916238e-07,
+      "loss": 1.2069,
+      "step": 17645
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.727157736453272e-07,
+      "loss": 1.215,
+      "step": 17650
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.616909746322897e-07,
+      "loss": 1.1989,
+      "step": 17655
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.507569636283961e-07,
+      "loss": 1.1836,
+      "step": 17660
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1796875,
+      "learning_rate": 6.399137506265307e-07,
+      "loss": 1.214,
+      "step": 17665
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.291613455365886e-07,
+      "loss": 1.2227,
+      "step": 17670
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.184997581855201e-07,
+      "loss": 1.0723,
+      "step": 17675
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.166015625,
+      "learning_rate": 6.079289983172199e-07,
+      "loss": 1.152,
+      "step": 17680
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.974490755926043e-07,
+      "loss": 1.2155,
+      "step": 17685
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.870599995895564e-07,
+      "loss": 1.1577,
+      "step": 17690
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 5.767617798029589e-07,
+      "loss": 1.0966,
+      "step": 17695
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.166015625,
+      "learning_rate": 5.665544256446054e-07,
+      "loss": 1.1387,
+      "step": 17700
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.56437946443289e-07,
+      "loss": 1.1982,
+      "step": 17705
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.464123514447472e-07,
+      "loss": 1.1425,
+      "step": 17710
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18359375,
+      "learning_rate": 5.364776498116398e-07,
+      "loss": 1.1517,
+      "step": 17715
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.266338506235701e-07,
+      "loss": 1.1926,
+      "step": 17720
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 5.168809628770532e-07,
+      "loss": 1.1041,
+      "step": 17725
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 5.072189954855034e-07,
+      "loss": 1.0947,
+      "step": 17730
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.976479572792903e-07,
+      "loss": 1.0862,
+      "step": 17735
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 4.88167857005628e-07,
+      "loss": 1.2524,
+      "step": 17740
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.787787033286417e-07,
+      "loss": 1.2265,
+      "step": 17745
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.694805048293338e-07,
+      "loss": 1.2045,
+      "step": 17750
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.602732700055956e-07,
+      "loss": 1.1682,
+      "step": 17755
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.181640625,
+      "learning_rate": 4.511570072721738e-07,
+      "loss": 1.2312,
+      "step": 17760
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.4213172496065936e-07,
+      "loss": 1.1783,
+      "step": 17765
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.3319743131949864e-07,
+      "loss": 1.1469,
+      "step": 17770
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.243541345140045e-07,
+      "loss": 1.1705,
+      "step": 17775
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.15601842626312e-07,
+      "loss": 1.0916,
+      "step": 17780
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.0694056365535584e-07,
+      "loss": 1.1413,
+      "step": 17785
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.9837030551694853e-07,
+      "loss": 1.2364,
+      "step": 17790
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.8989107604365803e-07,
+      "loss": 1.1832,
+      "step": 17795
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.8150288298489655e-07,
+      "loss": 1.1379,
+      "step": 17800
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.732057340068762e-07,
+      "loss": 1.0899,
+      "step": 17805
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.6499963669256453e-07,
+      "loss": 1.0986,
+      "step": 17810
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.173828125,
+      "learning_rate": 3.5688459854175127e-07,
+      "loss": 1.1834,
+      "step": 17815
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.173828125,
+      "learning_rate": 3.4886062697100375e-07,
+      "loss": 1.1115,
+      "step": 17820
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.409277293136337e-07,
+      "loss": 1.1083,
+      "step": 17825
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 3.3308591281973055e-07,
+      "loss": 1.1524,
+      "step": 17830
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 3.253351846561614e-07,
+      "loss": 1.1753,
+      "step": 17835
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 3.1767555190652667e-07,
+      "loss": 1.1314,
+      "step": 17840
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.10107021571171e-07,
+      "loss": 1.1214,
+      "step": 17845
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.026296005671725e-07,
+      "loss": 1.1708,
+      "step": 17850
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.9524329572837573e-07,
+      "loss": 1.1913,
+      "step": 17855
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.8794811380530304e-07,
+      "loss": 1.2385,
+      "step": 17860
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.8074406146523234e-07,
+      "loss": 1.1682,
+      "step": 17865
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.736311452921525e-07,
+      "loss": 1.1836,
+      "step": 17870
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.6660937178674127e-07,
+      "loss": 1.127,
+      "step": 17875
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.5967874736638756e-07,
+      "loss": 1.1737,
+      "step": 17880
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.5283927836519117e-07,
+      "loss": 1.1194,
+      "step": 17885
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.4609097103392984e-07,
+      "loss": 1.1573,
+      "step": 17890
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.3943383154005904e-07,
+      "loss": 1.1017,
+      "step": 17895
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.3286786596772302e-07,
+      "loss": 1.2186,
+      "step": 17900
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.2639308031774388e-07,
+      "loss": 1.1778,
+      "step": 17905
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.2000948050759917e-07,
+      "loss": 1.0945,
+      "step": 17910
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.1371707237144433e-07,
+      "loss": 1.1886,
+      "step": 17915
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.0751586166007919e-07,
+      "loss": 1.1772,
+      "step": 17920
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.0140585404094803e-07,
+      "loss": 1.2029,
+      "step": 17925
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.95387055098184e-07,
+      "loss": 1.0444,
+      "step": 17930
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.894594703324981e-07,
+      "loss": 1.1908,
+      "step": 17935
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.8362310516130132e-07,
+      "loss": 1.1938,
+      "step": 17940
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.7787796491859355e-07,
+      "loss": 1.1436,
+      "step": 17945
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.7222405485501912e-07,
+      "loss": 1.2447,
+      "step": 17950
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 1.666613801378447e-07,
+      "loss": 1.1373,
+      "step": 17955
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.6118994585094805e-07,
+      "loss": 1.2357,
+      "step": 17960
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.5580975699482914e-07,
+      "loss": 1.1555,
+      "step": 17965
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.5052081848657694e-07,
+      "loss": 1.1874,
+      "step": 17970
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.453231351599138e-07,
+      "loss": 1.2045,
+      "step": 17975
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.4021671176513984e-07,
+      "loss": 1.2113,
+      "step": 17980
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.3520155296914417e-07,
+      "loss": 1.2882,
+      "step": 17985
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.3027766335543812e-07,
+      "loss": 1.2075,
+      "step": 17990
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.254450474240776e-07,
+      "loss": 1.2115,
+      "step": 17995
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.207037095917407e-07,
+      "loss": 1.279,
+      "step": 18000
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.1605365419166126e-07,
+      "loss": 1.163,
+      "step": 18005
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.1149488547365083e-07,
+      "loss": 1.1874,
+      "step": 18010
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.0702740760409891e-07,
+      "loss": 1.1259,
+      "step": 18015
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.0265122466593946e-07,
+      "loss": 1.1281,
+      "step": 18020
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.836634065870654e-08,
+      "loss": 1.2164,
+      "step": 18025
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.41727594984676e-08,
+      "loss": 1.1758,
+      "step": 18030
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.007048501784576e-08,
+      "loss": 1.1462,
+      "step": 18035
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.605952096601977e-08,
+      "loss": 1.1391,
+      "step": 18040
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.213987100873511e-08,
+      "loss": 1.1093,
+      "step": 18045
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.831153872824847e-08,
+      "loss": 1.1492,
+      "step": 18050
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.457452762339445e-08,
+      "loss": 1.1813,
+      "step": 18055
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.092884110951881e-08,
+      "loss": 1.144,
+      "step": 18060
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.737448251852296e-08,
+      "loss": 1.2227,
+      "step": 18065
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.391145509883067e-08,
+      "loss": 1.1563,
+      "step": 18070
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.053976201538803e-08,
+      "loss": 1.2125,
+      "step": 18075
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.725940634969673e-08,
+      "loss": 1.0353,
+      "step": 18080
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 5.407039109974754e-08,
+      "loss": 1.1594,
+      "step": 18085
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.097271918007573e-08,
+      "loss": 1.186,
+      "step": 18090
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.7966393421727815e-08,
+      "loss": 1.2099,
+      "step": 18095
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.505141657226153e-08,
+      "loss": 1.1668,
+      "step": 18100
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.222779129576804e-08,
+      "loss": 1.1367,
+      "step": 18105
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.949552017282754e-08,
+      "loss": 1.1588,
+      "step": 18110
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.16796875,
+      "learning_rate": 3.685460570053145e-08,
+      "loss": 1.1962,
+      "step": 18115
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.169921875,
+      "learning_rate": 3.430505029249354e-08,
+      "loss": 1.1058,
+      "step": 18120
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.184685627882766e-08,
+      "loss": 1.2007,
+      "step": 18125
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 2.948002590612564e-08,
+      "loss": 1.1123,
+      "step": 18130
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.7204561337523803e-08,
+      "loss": 1.1671,
+      "step": 18135
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.5020464652603103e-08,
+      "loss": 1.2248,
+      "step": 18140
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.292773784748903e-08,
+      "loss": 1.1082,
+      "step": 18145
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.0926382834784984e-08,
+      "loss": 1.1874,
+      "step": 18150
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.9016401443572306e-08,
+      "loss": 1.0811,
+      "step": 18155
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.7197795419443552e-08,
+      "loss": 1.2035,
+      "step": 18160
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.54705664244581e-08,
+      "loss": 1.2233,
+      "step": 18165
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.3834716037197659e-08,
+      "loss": 1.1751,
+      "step": 18170
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.2290245752699659e-08,
+      "loss": 1.2549,
+      "step": 18175
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.0837156982501651e-08,
+      "loss": 1.1124,
+      "step": 18180
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1796875,
+      "learning_rate": 9.475451054619112e-09,
+      "loss": 1.1638,
+      "step": 18185
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.18359375,
+      "learning_rate": 8.20512921355654e-09,
+      "loss": 1.1538,
+      "step": 18190
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.026192620285254e-09,
+      "loss": 1.1048,
+      "step": 18195
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.938642352287804e-09,
+      "loss": 1.1768,
+      "step": 18200
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.942479403491351e-09,
+      "loss": 1.1869,
+      "step": 18205
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.037704684312083e-09,
+      "loss": 1.1892,
+      "step": 18210
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.2243190216552087e-09,
+      "loss": 1.1961,
+      "step": 18215
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.502323158903863e-09,
+      "loss": 1.0843,
+      "step": 18220
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.8717177558968957e-09,
+      "loss": 1.1736,
+      "step": 18225
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.3325033889621806e-09,
+      "loss": 1.1493,
+      "step": 18230
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1796875,
+      "learning_rate": 8.846805509166167e-10,
+      "loss": 1.1426,
+      "step": 18235
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.282496510328195e-10,
+      "loss": 1.216,
+      "step": 18240
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.6321101505022427e-10,
+      "loss": 1.1377,
+      "step": 18245
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.177734375,
+      "learning_rate": 8.956488520839302e-11,
+      "loss": 1.1538,
+      "step": 18250
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.31142020260478e-12,
+      "loss": 1.1785,
+      "step": 18255
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.1695406436920166,
+      "eval_runtime": 1922.8575,
+      "eval_samples_per_second": 8.407,
+      "eval_steps_per_second": 1.051,
+      "step": 18257
+    },
+    {
+      "epoch": 1.0,
+      "step": 18257,
+      "total_flos": 1.8793056999854572e+18,
+      "train_loss": 1.1836543822331609,
+      "train_runtime": 65654.1144,
+      "train_samples_per_second": 2.225,
+      "train_steps_per_second": 0.278
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 18257,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "total_flos": 1.8793056999854572e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}