diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,23062 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.2249183938333266,
+  "eval_steps": 10000,
+  "global_step": 82000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.548603911230614,
+      "learning_rate": 3e-09,
+      "loss": 1.3664,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 17.74263326459735,
+      "learning_rate": 1.55e-08,
+      "loss": 1.3938,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.768157816871117,
+      "learning_rate": 2.8000000000000003e-08,
+      "loss": 1.3875,
+      "step": 75
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.18791431777122,
+      "learning_rate": 4.05e-08,
+      "loss": 1.2869,
+      "step": 100
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.288361662728292,
+      "learning_rate": 5.3000000000000005e-08,
+      "loss": 1.3836,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 19.62961424039089,
+      "learning_rate": 6.550000000000001e-08,
+      "loss": 1.3198,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 18.510724860772605,
+      "learning_rate": 7.8e-08,
+      "loss": 1.3883,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 18.00659579211933,
+      "learning_rate": 9.050000000000001e-08,
+      "loss": 1.2896,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.149030358259694,
+      "learning_rate": 1.0300000000000001e-07,
+      "loss": 1.2758,
+      "step": 225
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 17.719461286476776,
+      "learning_rate": 1.1550000000000001e-07,
+      "loss": 1.1934,
+      "step": 250
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 18.479532395253678,
+      "learning_rate": 1.28e-07,
+      "loss": 1.2667,
+      "step": 275
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 18.189268236304354,
+      "learning_rate": 1.4050000000000002e-07,
+      "loss": 1.1188,
+      "step": 300
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 18.485150666624385,
+      "learning_rate": 1.53e-07,
+      "loss": 1.0843,
+      "step": 325
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.161860822592118,
+      "learning_rate": 1.655e-07,
+      "loss": 0.9748,
+      "step": 350
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 16.77864844044536,
+      "learning_rate": 1.78e-07,
+      "loss": 0.7618,
+      "step": 375
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.284839665193802,
+      "learning_rate": 1.9050000000000002e-07,
+      "loss": 0.6402,
+      "step": 400
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.563585737686541,
+      "learning_rate": 2.03e-07,
+      "loss": 0.6147,
+      "step": 425
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.195526034976105,
+      "learning_rate": 2.1550000000000002e-07,
+      "loss": 0.5645,
+      "step": 450
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.813975597971679,
+      "learning_rate": 2.2800000000000003e-07,
+      "loss": 0.564,
+      "step": 475
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.223184796437847,
+      "learning_rate": 2.405e-07,
+      "loss": 0.5439,
+      "step": 500
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.65715823434448,
+      "learning_rate": 2.53e-07,
+      "loss": 0.5262,
+      "step": 525
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.140278336499202,
+      "learning_rate": 2.6550000000000004e-07,
+      "loss": 0.4831,
+      "step": 550
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.595125220315392,
+      "learning_rate": 2.7800000000000003e-07,
+      "loss": 0.4605,
+      "step": 575
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 10.287013706634347,
+      "learning_rate": 2.905e-07,
+      "loss": 0.4882,
+      "step": 600
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.461372129814126,
+      "learning_rate": 3.0300000000000005e-07,
+      "loss": 0.4616,
+      "step": 625
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.492888534763829,
+      "learning_rate": 3.1550000000000004e-07,
+      "loss": 0.4572,
+      "step": 650
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.17219796215868,
+      "learning_rate": 3.280000000000001e-07,
+      "loss": 0.4422,
+      "step": 675
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.431269877118583,
+      "learning_rate": 3.405e-07,
+      "loss": 0.4441,
+      "step": 700
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.077076398765083,
+      "learning_rate": 3.53e-07,
+      "loss": 0.4324,
+      "step": 725
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.42463034481312,
+      "learning_rate": 3.6550000000000004e-07,
+      "loss": 0.4757,
+      "step": 750
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.002220474801542,
+      "learning_rate": 3.78e-07,
+      "loss": 0.4235,
+      "step": 775
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.249177989263307,
+      "learning_rate": 3.9050000000000006e-07,
+      "loss": 0.3844,
+      "step": 800
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.222274718732235,
+      "learning_rate": 4.0300000000000005e-07,
+      "loss": 0.4303,
+      "step": 825
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.29593645258072,
+      "learning_rate": 4.155e-07,
+      "loss": 0.3783,
+      "step": 850
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.0820275737306,
+      "learning_rate": 4.28e-07,
+      "loss": 0.3866,
+      "step": 875
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.284235367838265,
+      "learning_rate": 4.405e-07,
+      "loss": 0.3863,
+      "step": 900
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.990489183242782,
+      "learning_rate": 4.5300000000000005e-07,
+      "loss": 0.3868,
+      "step": 925
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.826149813494483,
+      "learning_rate": 4.6550000000000003e-07,
+      "loss": 0.3962,
+      "step": 950
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.652924783202419,
+      "learning_rate": 4.78e-07,
+      "loss": 0.4151,
+      "step": 975
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.231122893213122,
+      "learning_rate": 4.905000000000001e-07,
+      "loss": 0.4139,
+      "step": 1000
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.284357637366583,
+      "learning_rate": 5.03e-07,
+      "loss": 0.4003,
+      "step": 1025
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.764967682696836,
+      "learning_rate": 5.155e-07,
+      "loss": 0.3424,
+      "step": 1050
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.623487096739643,
+      "learning_rate": 5.280000000000001e-07,
+      "loss": 0.3726,
+      "step": 1075
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.118786193989378,
+      "learning_rate": 5.405000000000001e-07,
+      "loss": 0.3628,
+      "step": 1100
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.266772083450553,
+      "learning_rate": 5.53e-07,
+      "loss": 0.3585,
+      "step": 1125
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.9161160573400835,
+      "learning_rate": 5.655e-07,
+      "loss": 0.3638,
+      "step": 1150
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.957547049197,
+      "learning_rate": 5.78e-07,
+      "loss": 0.3543,
+      "step": 1175
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.64987127577368,
+      "learning_rate": 5.905e-07,
+      "loss": 0.3635,
+      "step": 1200
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.6784759653959345,
+      "learning_rate": 6.030000000000001e-07,
+      "loss": 0.3822,
+      "step": 1225
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.353852424599333,
+      "learning_rate": 6.155000000000001e-07,
+      "loss": 0.3715,
+      "step": 1250
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.886970263184239,
+      "learning_rate": 6.28e-07,
+      "loss": 0.3751,
+      "step": 1275
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.1464634900527635,
+      "learning_rate": 6.405e-07,
+      "loss": 0.3637,
+      "step": 1300
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.365126116348344,
+      "learning_rate": 6.53e-07,
+      "loss": 0.3717,
+      "step": 1325
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.264665698841982,
+      "learning_rate": 6.655e-07,
+      "loss": 0.3815,
+      "step": 1350
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.817525071240404,
+      "learning_rate": 6.78e-07,
+      "loss": 0.3192,
+      "step": 1375
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.541821815933655,
+      "learning_rate": 6.905000000000001e-07,
+      "loss": 0.3682,
+      "step": 1400
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.775244243834967,
+      "learning_rate": 7.030000000000001e-07,
+      "loss": 0.3583,
+      "step": 1425
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.699420120871023,
+      "learning_rate": 7.155000000000001e-07,
+      "loss": 0.353,
+      "step": 1450
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.3256513159082886,
+      "learning_rate": 7.280000000000001e-07,
+      "loss": 0.3475,
+      "step": 1475
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.954227102446838,
+      "learning_rate": 7.405000000000002e-07,
+      "loss": 0.3531,
+      "step": 1500
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.268304372701325,
+      "learning_rate": 7.530000000000001e-07,
+      "loss": 0.3641,
+      "step": 1525
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.594146689998704,
+      "learning_rate": 7.655000000000001e-07,
+      "loss": 0.3539,
+      "step": 1550
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.380277920462982,
+      "learning_rate": 7.78e-07,
+      "loss": 0.3398,
+      "step": 1575
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.249645134391981,
+      "learning_rate": 7.905e-07,
+      "loss": 0.3317,
+      "step": 1600
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.168651318042285,
+      "learning_rate": 8.03e-07,
+      "loss": 0.3347,
+      "step": 1625
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.779448193291338,
+      "learning_rate": 8.155000000000001e-07,
+      "loss": 0.3574,
+      "step": 1650
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.413939794725913,
+      "learning_rate": 8.280000000000001e-07,
+      "loss": 0.3444,
+      "step": 1675
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.011851874436536,
+      "learning_rate": 8.405e-07,
+      "loss": 0.346,
+      "step": 1700
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.4563682729389615,
+      "learning_rate": 8.53e-07,
+      "loss": 0.3496,
+      "step": 1725
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.042812209726682,
+      "learning_rate": 8.655000000000001e-07,
+      "loss": 0.3251,
+      "step": 1750
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.982650134613024,
+      "learning_rate": 8.780000000000001e-07,
+      "loss": 0.3312,
+      "step": 1775
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.208414233430044,
+      "learning_rate": 8.905000000000001e-07,
+      "loss": 0.3214,
+      "step": 1800
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.3406319348355105,
+      "learning_rate": 9.030000000000001e-07,
+      "loss": 0.351,
+      "step": 1825
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.869539965379989,
+      "learning_rate": 9.155000000000002e-07,
+      "loss": 0.3277,
+      "step": 1850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.619686331747831,
+      "learning_rate": 9.28e-07,
+      "loss": 0.3419,
+      "step": 1875
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.690019286737048,
+      "learning_rate": 9.405e-07,
+      "loss": 0.3313,
+      "step": 1900
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.40223765276817,
+      "learning_rate": 9.53e-07,
+      "loss": 0.3259,
+      "step": 1925
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.092936931428083,
+      "learning_rate": 9.655000000000001e-07,
+      "loss": 0.325,
+      "step": 1950
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.59220421800555,
+      "learning_rate": 9.78e-07,
+      "loss": 0.3364,
+      "step": 1975
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.215626189268482,
+      "learning_rate": 9.905e-07,
+      "loss": 0.3372,
+      "step": 2000
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.7827726836036035,
+      "learning_rate": 1.0030000000000002e-06,
+      "loss": 0.2945,
+      "step": 2025
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.846888472808441,
+      "learning_rate": 1.0155e-06,
+      "loss": 0.3352,
+      "step": 2050
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.524600018661036,
+      "learning_rate": 1.0280000000000002e-06,
+      "loss": 0.325,
+      "step": 2075
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.029269384160894,
+      "learning_rate": 1.0405e-06,
+      "loss": 0.3445,
+      "step": 2100
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.541151743266427,
+      "learning_rate": 1.0530000000000001e-06,
+      "loss": 0.3487,
+      "step": 2125
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 9.597600892762339,
+      "learning_rate": 1.0655000000000002e-06,
+      "loss": 0.3384,
+      "step": 2150
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.130328659028095,
+      "learning_rate": 1.078e-06,
+      "loss": 0.3482,
+      "step": 2175
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.4990076386556,
+      "learning_rate": 1.0905e-06,
+      "loss": 0.3026,
+      "step": 2200
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.422295132082161,
+      "learning_rate": 1.103e-06,
+      "loss": 0.3057,
+      "step": 2225
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.017101554617902,
+      "learning_rate": 1.1155e-06,
+      "loss": 0.3233,
+      "step": 2250
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.25880061391202,
+      "learning_rate": 1.128e-06,
+      "loss": 0.3046,
+      "step": 2275
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.812137488778935,
+      "learning_rate": 1.1405000000000001e-06,
+      "loss": 0.3125,
+      "step": 2300
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.88024638341902,
+      "learning_rate": 1.153e-06,
+      "loss": 0.3322,
+      "step": 2325
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 9.480678112460733,
+      "learning_rate": 1.1655000000000001e-06,
+      "loss": 0.3506,
+      "step": 2350
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.625242578507233,
+      "learning_rate": 1.1780000000000002e-06,
+      "loss": 0.3081,
+      "step": 2375
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.471850154248505,
+      "learning_rate": 1.1905e-06,
+      "loss": 0.3411,
+      "step": 2400
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.700212579759201,
+      "learning_rate": 1.2030000000000002e-06,
+      "loss": 0.2955,
+      "step": 2425
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.640668754634842,
+      "learning_rate": 1.2155e-06,
+      "loss": 0.3147,
+      "step": 2450
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.142799254759943,
+      "learning_rate": 1.2280000000000001e-06,
+      "loss": 0.3223,
+      "step": 2475
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.273025415582895,
+      "learning_rate": 1.2405e-06,
+      "loss": 0.3183,
+      "step": 2500
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.15480159136244,
+      "learning_rate": 1.2530000000000001e-06,
+      "loss": 0.3042,
+      "step": 2525
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.986248278662067,
+      "learning_rate": 1.2655e-06,
+      "loss": 0.2898,
+      "step": 2550
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.9880022706315765,
+      "learning_rate": 1.278e-06,
+      "loss": 0.3342,
+      "step": 2575
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.797582197495177,
+      "learning_rate": 1.2905000000000002e-06,
+      "loss": 0.3204,
+      "step": 2600
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.63428878101681,
+      "learning_rate": 1.303e-06,
+      "loss": 0.3106,
+      "step": 2625
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.016874596706454,
+      "learning_rate": 1.3155000000000002e-06,
+      "loss": 0.3034,
+      "step": 2650
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.761116991346277,
+      "learning_rate": 1.328e-06,
+      "loss": 0.3125,
+      "step": 2675
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.147210364841152,
+      "learning_rate": 1.3405000000000001e-06,
+      "loss": 0.3212,
+      "step": 2700
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.198210069471468,
+      "learning_rate": 1.3530000000000002e-06,
+      "loss": 0.3005,
+      "step": 2725
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.62021097760392,
+      "learning_rate": 1.3655e-06,
+      "loss": 0.3095,
+      "step": 2750
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.498105116813674,
+      "learning_rate": 1.3780000000000002e-06,
+      "loss": 0.2968,
+      "step": 2775
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.096234425980665,
+      "learning_rate": 1.3905000000000003e-06,
+      "loss": 0.3013,
+      "step": 2800
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.529807578803446,
+      "learning_rate": 1.4030000000000002e-06,
+      "loss": 0.3285,
+      "step": 2825
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.090935769510644,
+      "learning_rate": 1.4155000000000003e-06,
+      "loss": 0.3186,
+      "step": 2850
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.142203590894807,
+      "learning_rate": 1.4280000000000001e-06,
+      "loss": 0.3283,
+      "step": 2875
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.5553560310705885,
+      "learning_rate": 1.4405000000000002e-06,
+      "loss": 0.3078,
+      "step": 2900
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.030443253782761,
+      "learning_rate": 1.4530000000000003e-06,
+      "loss": 0.3259,
+      "step": 2925
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.45791822784214,
+      "learning_rate": 1.4655000000000002e-06,
+      "loss": 0.3071,
+      "step": 2950
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.169665682759633,
+      "learning_rate": 1.478e-06,
+      "loss": 0.3059,
+      "step": 2975
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.903163712753615,
+      "learning_rate": 1.4905e-06,
+      "loss": 0.3201,
+      "step": 3000
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4235199170734267,
+      "learning_rate": 1.5025e-06,
+      "loss": 0.3193,
+      "step": 3025
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.388562008472956,
+      "learning_rate": 1.5145e-06,
+      "loss": 0.3154,
+      "step": 3050
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.32478874239276,
+      "learning_rate": 1.5270000000000002e-06,
+      "loss": 0.3341,
+      "step": 3075
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.882897860372599,
+      "learning_rate": 1.5395000000000003e-06,
+      "loss": 0.2996,
+      "step": 3100
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.830461306218142,
+      "learning_rate": 1.5520000000000001e-06,
+      "loss": 0.2997,
+      "step": 3125
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.340070140671299,
+      "learning_rate": 1.5645000000000002e-06,
+      "loss": 0.2984,
+      "step": 3150
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.491658642899819,
+      "learning_rate": 1.577e-06,
+      "loss": 0.3165,
+      "step": 3175
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.69433147934288,
+      "learning_rate": 1.5895000000000002e-06,
+      "loss": 0.3031,
+      "step": 3200
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.314351227889603,
+      "learning_rate": 1.6020000000000003e-06,
+      "loss": 0.3018,
+      "step": 3225
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.310085290349688,
+      "learning_rate": 1.6145000000000002e-06,
+      "loss": 0.2997,
+      "step": 3250
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.305879482033044,
+      "learning_rate": 1.6270000000000003e-06,
+      "loss": 0.3054,
+      "step": 3275
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.40127326870802,
+      "learning_rate": 1.6395000000000004e-06,
+      "loss": 0.3134,
+      "step": 3300
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.850845319267836,
+      "learning_rate": 1.6520000000000002e-06,
+      "loss": 0.3062,
+      "step": 3325
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.047475032917652,
+      "learning_rate": 1.6645e-06,
+      "loss": 0.2863,
+      "step": 3350
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.725844976704757,
+      "learning_rate": 1.677e-06,
+      "loss": 0.2891,
+      "step": 3375
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.78153953615269,
+      "learning_rate": 1.6895e-06,
+      "loss": 0.3248,
+      "step": 3400
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.696888074356876,
+      "learning_rate": 1.702e-06,
+      "loss": 0.3188,
+      "step": 3425
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.676273130360383,
+      "learning_rate": 1.7145e-06,
+      "loss": 0.2988,
+      "step": 3450
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.006703021039653,
+      "learning_rate": 1.7270000000000002e-06,
+      "loss": 0.3086,
+      "step": 3475
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.465714704473822,
+      "learning_rate": 1.7395e-06,
+      "loss": 0.3052,
+      "step": 3500
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.170356640181524,
+      "learning_rate": 1.7520000000000001e-06,
+      "loss": 0.2844,
+      "step": 3525
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.411466808472138,
+      "learning_rate": 1.7645e-06,
+      "loss": 0.2994,
+      "step": 3550
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.624191447880747,
+      "learning_rate": 1.777e-06,
+      "loss": 0.3387,
+      "step": 3575
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.697489245961635,
+      "learning_rate": 1.7895000000000002e-06,
+      "loss": 0.2976,
+      "step": 3600
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.358874355083804,
+      "learning_rate": 1.802e-06,
+      "loss": 0.3221,
+      "step": 3625
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.701960236286209,
+      "learning_rate": 1.8145000000000002e-06,
+      "loss": 0.2778,
+      "step": 3650
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.093145317658292,
+      "learning_rate": 1.8270000000000003e-06,
+      "loss": 0.2809,
+      "step": 3675
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.5735424824748065,
+      "learning_rate": 1.8395000000000001e-06,
+      "loss": 0.3094,
+      "step": 3700
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.111126859679747,
+      "learning_rate": 1.8520000000000002e-06,
+      "loss": 0.2811,
+      "step": 3725
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.603638409237136,
+      "learning_rate": 1.8645e-06,
+      "loss": 0.3065,
+      "step": 3750
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 9.4063632414015,
+      "learning_rate": 1.8770000000000002e-06,
+      "loss": 0.3027,
+      "step": 3775
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.494637512263619,
+      "learning_rate": 1.8895000000000003e-06,
+      "loss": 0.3136,
+      "step": 3800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.048441681493421,
+      "learning_rate": 1.9020000000000002e-06,
+      "loss": 0.3337,
+      "step": 3825
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.438034199516792,
+      "learning_rate": 1.9145e-06,
+      "loss": 0.269,
+      "step": 3850
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.358960525686304,
+      "learning_rate": 1.9270000000000004e-06,
+      "loss": 0.3209,
+      "step": 3875
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.925504929075971,
+      "learning_rate": 1.9395000000000002e-06,
+      "loss": 0.2988,
+      "step": 3900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.739147879633613,
+      "learning_rate": 1.952e-06,
+      "loss": 0.2722,
+      "step": 3925
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.776015161103489,
+      "learning_rate": 1.9645000000000004e-06,
+      "loss": 0.2798,
+      "step": 3950
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.056155062567179,
+      "learning_rate": 1.977e-06,
+      "loss": 0.3053,
+      "step": 3975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.121259851550022,
+      "learning_rate": 1.9895e-06,
+      "loss": 0.2838,
+      "step": 4000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.30300122513391,
+      "learning_rate": 2.002e-06,
+      "loss": 0.3092,
+      "step": 4025
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.035093216193755,
+      "learning_rate": 2.0145e-06,
+      "loss": 0.2989,
+      "step": 4050
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.312870093933501,
+      "learning_rate": 2.0270000000000002e-06,
+      "loss": 0.2909,
+      "step": 4075
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.1161318992697105,
+      "learning_rate": 2.0395e-06,
+      "loss": 0.3155,
+      "step": 4100
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.985023001843178,
+      "learning_rate": 2.052e-06,
+      "loss": 0.2811,
+      "step": 4125
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.490009552023306,
+      "learning_rate": 2.0645000000000003e-06,
+      "loss": 0.2943,
+      "step": 4150
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.036226568442008,
+      "learning_rate": 2.077e-06,
+      "loss": 0.3051,
+      "step": 4175
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.160882683593939,
+      "learning_rate": 2.0895e-06,
+      "loss": 0.3128,
+      "step": 4200
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.899026627729313,
+      "learning_rate": 2.102e-06,
+      "loss": 0.3007,
+      "step": 4225
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.667556679571071,
+      "learning_rate": 2.1145000000000003e-06,
+      "loss": 0.294,
+      "step": 4250
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.45891735416569,
+      "learning_rate": 2.127e-06,
+      "loss": 0.3072,
+      "step": 4275
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.247465785827324,
+      "learning_rate": 2.1395e-06,
+      "loss": 0.2987,
+      "step": 4300
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.925041488918864,
+      "learning_rate": 2.1520000000000003e-06,
+      "loss": 0.3038,
+      "step": 4325
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.697636253311961,
+      "learning_rate": 2.1645e-06,
+      "loss": 0.2896,
+      "step": 4350
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.671845123322663,
+      "learning_rate": 2.177e-06,
+      "loss": 0.3013,
+      "step": 4375
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 14.42965685715927,
+      "learning_rate": 2.1895000000000004e-06,
+      "loss": 0.2935,
+      "step": 4400
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.619313005480762,
+      "learning_rate": 2.2020000000000003e-06,
+      "loss": 0.2822,
+      "step": 4425
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.96836481945949,
+      "learning_rate": 2.2145e-06,
+      "loss": 0.3013,
+      "step": 4450
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.013522956861229,
+      "learning_rate": 2.2270000000000004e-06,
+      "loss": 0.2998,
+      "step": 4475
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.564283110102572,
+      "learning_rate": 2.2395000000000003e-06,
+      "loss": 0.3069,
+      "step": 4500
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.412696409963094,
+      "learning_rate": 2.252e-06,
+      "loss": 0.2911,
+      "step": 4525
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.532716257217559,
+      "learning_rate": 2.2645000000000005e-06,
+      "loss": 0.293,
+      "step": 4550
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.44868242354979,
+      "learning_rate": 2.2770000000000004e-06,
+      "loss": 0.2785,
+      "step": 4575
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.10635902763617,
+      "learning_rate": 2.2895e-06,
+      "loss": 0.2966,
+      "step": 4600
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.249805780154833,
+      "learning_rate": 2.302e-06,
+      "loss": 0.3102,
+      "step": 4625
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.204571957255101,
+      "learning_rate": 2.3145e-06,
+      "loss": 0.309,
+      "step": 4650
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.803854804655206,
+      "learning_rate": 2.327e-06,
+      "loss": 0.2814,
+      "step": 4675
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.440776762453003,
+      "learning_rate": 2.3395000000000002e-06,
+      "loss": 0.3056,
+      "step": 4700
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.15879339209754,
+      "learning_rate": 2.352e-06,
+      "loss": 0.2964,
+      "step": 4725
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.891657048350974,
+      "learning_rate": 2.3645e-06,
+      "loss": 0.2815,
+      "step": 4750
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.240599749344253,
+      "learning_rate": 2.3770000000000003e-06,
+      "loss": 0.2854,
+      "step": 4775
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.975988999533506,
+      "learning_rate": 2.3895e-06,
+      "loss": 0.2888,
+      "step": 4800
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.013535222990837,
+      "learning_rate": 2.402e-06,
+      "loss": 0.2954,
+      "step": 4825
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.029743044779613,
+      "learning_rate": 2.4145000000000003e-06,
+      "loss": 0.2884,
+      "step": 4850
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.956958714016763,
+      "learning_rate": 2.4270000000000002e-06,
+      "loss": 0.2969,
+      "step": 4875
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.823749748330595,
+      "learning_rate": 2.4395e-06,
+      "loss": 0.3081,
+      "step": 4900
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.2241357450122425,
+      "learning_rate": 2.4520000000000004e-06,
+      "loss": 0.2992,
+      "step": 4925
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.16911141354443,
+      "learning_rate": 2.4645000000000003e-06,
+      "loss": 0.3245,
+      "step": 4950
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.060217712197034,
+      "learning_rate": 2.477e-06,
+      "loss": 0.3118,
+      "step": 4975
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.425673795449125,
+      "learning_rate": 2.4895e-06,
+      "loss": 0.2662,
+      "step": 5000
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.595958329136231,
+      "learning_rate": 2.502e-06,
+      "loss": 0.3158,
+      "step": 5025
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6550675982874363,
+      "learning_rate": 2.5140000000000004e-06,
+      "loss": 0.28,
+      "step": 5050
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.79782451562284,
+      "learning_rate": 2.5265e-06,
+      "loss": 0.2879,
+      "step": 5075
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.451642556039346,
+      "learning_rate": 2.539e-06,
+      "loss": 0.305,
+      "step": 5100
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.916829339590843,
+      "learning_rate": 2.5515e-06,
+      "loss": 0.271,
+      "step": 5125
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.245252699817357,
+      "learning_rate": 2.5640000000000004e-06,
+      "loss": 0.2763,
+      "step": 5150
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.212402917259702,
+      "learning_rate": 2.5765000000000002e-06,
+      "loss": 0.2763,
+      "step": 5175
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.823573409166045,
+      "learning_rate": 2.5890000000000005e-06,
+      "loss": 0.3001,
+      "step": 5200
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.940683369097169,
+      "learning_rate": 2.6015e-06,
+      "loss": 0.3116,
+      "step": 5225
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.3976420123516755,
+      "learning_rate": 2.6140000000000003e-06,
+      "loss": 0.2606,
+      "step": 5250
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.791330529626596,
+      "learning_rate": 2.6265e-06,
+      "loss": 0.2826,
+      "step": 5275
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.076143804816086,
+      "learning_rate": 2.6390000000000005e-06,
+      "loss": 0.2833,
+      "step": 5300
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.190822919416679,
+      "learning_rate": 2.6515000000000004e-06,
+      "loss": 0.3373,
+      "step": 5325
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.8692012310423465,
+      "learning_rate": 2.6640000000000007e-06,
+      "loss": 0.3403,
+      "step": 5350
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.442368544772186,
+      "learning_rate": 2.6765e-06,
+      "loss": 0.2975,
+      "step": 5375
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.524857704750757,
+      "learning_rate": 2.689e-06,
+      "loss": 0.2758,
+      "step": 5400
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.21051104357922,
+      "learning_rate": 2.7015000000000003e-06,
+      "loss": 0.2862,
+      "step": 5425
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.6759388918523985,
+      "learning_rate": 2.7139999999999998e-06,
+      "loss": 0.2744,
+      "step": 5450
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.630174118449949,
+      "learning_rate": 2.7265e-06,
+      "loss": 0.2988,
+      "step": 5475
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.357716937591224,
+      "learning_rate": 2.739e-06,
+      "loss": 0.2827,
+      "step": 5500
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.779528956514759,
+      "learning_rate": 2.7515000000000003e-06,
+      "loss": 0.2589,
+      "step": 5525
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.02168399367326,
+      "learning_rate": 2.764e-06,
+      "loss": 0.2819,
+      "step": 5550
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.823990587562972,
+      "learning_rate": 2.7765000000000004e-06,
+      "loss": 0.3137,
+      "step": 5575
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.019705933767146,
+      "learning_rate": 2.789e-06,
+      "loss": 0.3069,
+      "step": 5600
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.7318630027454995,
+      "learning_rate": 2.8015e-06,
+      "loss": 0.2665,
+      "step": 5625
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.11880507789445,
+      "learning_rate": 2.814e-06,
+      "loss": 0.2872,
+      "step": 5650
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.7977850369642505,
+      "learning_rate": 2.8265000000000004e-06,
+      "loss": 0.283,
+      "step": 5675
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.341136733880947,
+      "learning_rate": 2.8390000000000003e-06,
+      "loss": 0.3007,
+      "step": 5700
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.970881329915707,
+      "learning_rate": 2.8515000000000006e-06,
+      "loss": 0.2784,
+      "step": 5725
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.729318544130669,
+      "learning_rate": 2.864e-06,
+      "loss": 0.3107,
+      "step": 5750
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.546507761712345,
+      "learning_rate": 2.8765000000000003e-06,
+      "loss": 0.2822,
+      "step": 5775
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.978696715726333,
+      "learning_rate": 2.889e-06,
+      "loss": 0.288,
+      "step": 5800
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.228311134929747,
+      "learning_rate": 2.9015000000000005e-06,
+      "loss": 0.2903,
+      "step": 5825
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.909081531042711,
+      "learning_rate": 2.914e-06,
+      "loss": 0.3001,
+      "step": 5850
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.066473702285178,
+      "learning_rate": 2.9265000000000003e-06,
+      "loss": 0.3,
+      "step": 5875
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.981022806434303,
+      "learning_rate": 2.939e-06,
+      "loss": 0.2623,
+      "step": 5900
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.451955876526818,
+      "learning_rate": 2.9515000000000005e-06,
+      "loss": 0.3001,
+      "step": 5925
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.213926487478616,
+      "learning_rate": 2.9640000000000003e-06,
+      "loss": 0.2856,
+      "step": 5950
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.590418657876296,
+      "learning_rate": 2.9765000000000006e-06,
+      "loss": 0.3048,
+      "step": 5975
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.618581560821221,
+      "learning_rate": 2.989e-06,
+      "loss": 0.2802,
+      "step": 6000
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.747177181686857,
+      "learning_rate": 3.0015e-06,
+      "loss": 0.3047,
+      "step": 6025
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.702779308707192,
+      "learning_rate": 3.0140000000000003e-06,
+      "loss": 0.2845,
+      "step": 6050
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.466401914069188,
+      "learning_rate": 3.026e-06,
+      "loss": 0.2889,
+      "step": 6075
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.750175939483835,
+      "learning_rate": 3.0385000000000002e-06,
+      "loss": 0.2998,
+      "step": 6100
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.052264883280283,
+      "learning_rate": 3.051e-06,
+      "loss": 0.3153,
+      "step": 6125
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.1350193715664165,
+      "learning_rate": 3.0635000000000004e-06,
+      "loss": 0.2599,
+      "step": 6150
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.948910715297065,
+      "learning_rate": 3.0760000000000003e-06,
+      "loss": 0.2711,
+      "step": 6175
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.252587211408633,
+      "learning_rate": 3.0885000000000006e-06,
+      "loss": 0.295,
+      "step": 6200
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.165520107137274,
+      "learning_rate": 3.101e-06,
+      "loss": 0.2763,
+      "step": 6225
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.4996014769823915,
+      "learning_rate": 3.1135000000000003e-06,
+      "loss": 0.2849,
+      "step": 6250
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.964362624237183,
+      "learning_rate": 3.1260000000000002e-06,
+      "loss": 0.3028,
+      "step": 6275
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.169773589103691,
+      "learning_rate": 3.1385000000000005e-06,
+      "loss": 0.278,
+      "step": 6300
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.406835628812196,
+      "learning_rate": 3.151e-06,
+      "loss": 0.2558,
+      "step": 6325
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.084464446709841,
+      "learning_rate": 3.1635000000000003e-06,
+      "loss": 0.284,
+      "step": 6350
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.694245128142803,
+      "learning_rate": 3.176e-06,
+      "loss": 0.2611,
+      "step": 6375
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.564775176663976,
+      "learning_rate": 3.1885000000000005e-06,
+      "loss": 0.277,
+      "step": 6400
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.356766955805626,
+      "learning_rate": 3.2010000000000004e-06,
+      "loss": 0.2585,
+      "step": 6425
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.649934835176111,
+      "learning_rate": 3.2135000000000007e-06,
+      "loss": 0.2556,
+      "step": 6450
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.635433701155168,
+      "learning_rate": 3.226e-06,
+      "loss": 0.291,
+      "step": 6475
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.724811843119688,
+      "learning_rate": 3.2385000000000004e-06,
+      "loss": 0.2774,
+      "step": 6500
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.629267878070488,
+      "learning_rate": 3.2510000000000003e-06,
+      "loss": 0.288,
+      "step": 6525
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.037964444193315,
+      "learning_rate": 3.2635e-06,
+      "loss": 0.2723,
+      "step": 6550
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.182370121813334,
+      "learning_rate": 3.2760000000000005e-06,
+      "loss": 0.3106,
+      "step": 6575
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.1528459015462245,
+      "learning_rate": 3.2885e-06,
+      "loss": 0.2692,
+      "step": 6600
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.500105672722975,
+      "learning_rate": 3.3010000000000002e-06,
+      "loss": 0.3031,
+      "step": 6625
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.834985512213181,
+      "learning_rate": 3.3135e-06,
+      "loss": 0.2664,
+      "step": 6650
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.19309701510752,
+      "learning_rate": 3.3260000000000004e-06,
+      "loss": 0.2723,
+      "step": 6675
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.139308511625616,
+      "learning_rate": 3.3385e-06,
+      "loss": 0.2839,
+      "step": 6700
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.3241076676376435,
+      "learning_rate": 3.351e-06,
+      "loss": 0.2806,
+      "step": 6725
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.98434229153364,
+      "learning_rate": 3.3635e-06,
+      "loss": 0.28,
+      "step": 6750
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.129264792301614,
+      "learning_rate": 3.3760000000000004e-06,
+      "loss": 0.282,
+      "step": 6775
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.170862565343718,
+      "learning_rate": 3.3885000000000003e-06,
+      "loss": 0.2858,
+      "step": 6800
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.838114805347313,
+      "learning_rate": 3.4010000000000006e-06,
+      "loss": 0.2821,
+      "step": 6825
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.030750778829712,
+      "learning_rate": 3.4135e-06,
+      "loss": 0.2925,
+      "step": 6850
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.186794776187053,
+      "learning_rate": 3.4260000000000003e-06,
+      "loss": 0.2813,
+      "step": 6875
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.6235944995301645,
+      "learning_rate": 3.4385e-06,
+      "loss": 0.2884,
+      "step": 6900
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.471229594532035,
+      "learning_rate": 3.4510000000000005e-06,
+      "loss": 0.2759,
+      "step": 6925
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.0819247117845,
+      "learning_rate": 3.4635000000000004e-06,
+      "loss": 0.2738,
+      "step": 6950
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.998542491050096,
+      "learning_rate": 3.4760000000000007e-06,
+      "loss": 0.2878,
+      "step": 6975
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.461630412552012,
+      "learning_rate": 3.4885e-06,
+      "loss": 0.3053,
+      "step": 7000
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.586574763749033,
+      "learning_rate": 3.5010000000000004e-06,
+      "loss": 0.2919,
+      "step": 7025
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.792709877852477,
+      "learning_rate": 3.5135000000000003e-06,
+      "loss": 0.2641,
+      "step": 7050
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.911279814176992,
+      "learning_rate": 3.5255e-06,
+      "loss": 0.2902,
+      "step": 7075
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.067894245771899,
+      "learning_rate": 3.5380000000000003e-06,
+      "loss": 0.2668,
+      "step": 7100
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.1730720185264785,
+      "learning_rate": 3.5505e-06,
+      "loss": 0.2602,
+      "step": 7125
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.9493924947782295,
+      "learning_rate": 3.5630000000000004e-06,
+      "loss": 0.2878,
+      "step": 7150
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.022243763558953,
+      "learning_rate": 3.5755e-06,
+      "loss": 0.2792,
+      "step": 7175
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.59839736126787,
+      "learning_rate": 3.588e-06,
+      "loss": 0.2716,
+      "step": 7200
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.033887406465633,
+      "learning_rate": 3.6005e-06,
+      "loss": 0.3156,
+      "step": 7225
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.642200365588752,
+      "learning_rate": 3.6130000000000004e-06,
+      "loss": 0.2713,
+      "step": 7250
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.796078080375151,
+      "learning_rate": 3.6255000000000003e-06,
+      "loss": 0.2759,
+      "step": 7275
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.564454612601875,
+      "learning_rate": 3.6380000000000006e-06,
+      "loss": 0.2598,
+      "step": 7300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.505268037249782,
+      "learning_rate": 3.6505e-06,
+      "loss": 0.2723,
+      "step": 7325
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.1652467945628,
+      "learning_rate": 3.6630000000000003e-06,
+      "loss": 0.2671,
+      "step": 7350
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.08578394589941,
+      "learning_rate": 3.6755000000000002e-06,
+      "loss": 0.2741,
+      "step": 7375
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.409058891602932,
+      "learning_rate": 3.6880000000000005e-06,
+      "loss": 0.2785,
+      "step": 7400
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.599948826653968,
+      "learning_rate": 3.7005000000000004e-06,
+      "loss": 0.2881,
+      "step": 7425
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.3632193157404755,
+      "learning_rate": 3.7130000000000007e-06,
+      "loss": 0.2754,
+      "step": 7450
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.580890830497889,
+      "learning_rate": 3.7255e-06,
+      "loss": 0.2763,
+      "step": 7475
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.429129214199958,
+      "learning_rate": 3.7380000000000005e-06,
+      "loss": 0.27,
+      "step": 7500
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.94826201614837,
+      "learning_rate": 3.7505000000000003e-06,
+      "loss": 0.2718,
+      "step": 7525
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.004602730126053,
+      "learning_rate": 3.7630000000000006e-06,
+      "loss": 0.2996,
+      "step": 7550
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.70321711644467,
+      "learning_rate": 3.7755e-06,
+      "loss": 0.2533,
+      "step": 7575
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.926960517500858,
+      "learning_rate": 3.7880000000000004e-06,
+      "loss": 0.266,
+      "step": 7600
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.099979599842278,
+      "learning_rate": 3.8005000000000003e-06,
+      "loss": 0.278,
+      "step": 7625
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.394314379220583,
+      "learning_rate": 3.813e-06,
+      "loss": 0.2794,
+      "step": 7650
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.984983861009723,
+      "learning_rate": 3.8255e-06,
+      "loss": 0.2889,
+      "step": 7675
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.498279020164819,
+      "learning_rate": 3.838e-06,
+      "loss": 0.2793,
+      "step": 7700
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.672802599011468,
+      "learning_rate": 3.850500000000001e-06,
+      "loss": 0.2954,
+      "step": 7725
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.429422338538009,
+      "learning_rate": 3.863e-06,
+      "loss": 0.3035,
+      "step": 7750
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.278266658380528,
+      "learning_rate": 3.8755e-06,
+      "loss": 0.2887,
+      "step": 7775
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.933838932503291,
+      "learning_rate": 3.888e-06,
+      "loss": 0.27,
+      "step": 7800
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.341557859998104,
+      "learning_rate": 3.9005e-06,
+      "loss": 0.2874,
+      "step": 7825
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.368800106240496,
+      "learning_rate": 3.9130000000000005e-06,
+      "loss": 0.2697,
+      "step": 7850
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.531746542220086,
+      "learning_rate": 3.925500000000001e-06,
+      "loss": 0.2704,
+      "step": 7875
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.948769494377511,
+      "learning_rate": 3.938e-06,
+      "loss": 0.266,
+      "step": 7900
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.0940656271040545,
+      "learning_rate": 3.9505000000000005e-06,
+      "loss": 0.2859,
+      "step": 7925
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.508339585723285,
+      "learning_rate": 3.963e-06,
+      "loss": 0.2851,
+      "step": 7950
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.14929249851529,
+      "learning_rate": 3.9755e-06,
+      "loss": 0.2823,
+      "step": 7975
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.3779482123770475,
+      "learning_rate": 3.988000000000001e-06,
+      "loss": 0.2698,
+      "step": 8000
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.074063079478101,
+      "learning_rate": 4.000500000000001e-06,
+      "loss": 0.2819,
+      "step": 8025
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.362443834444672,
+      "learning_rate": 4.013e-06,
+      "loss": 0.2623,
+      "step": 8050
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.51778892694598,
+      "learning_rate": 4.0250000000000004e-06,
+      "loss": 0.2563,
+      "step": 8075
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.3055285454811845,
+      "learning_rate": 4.037500000000001e-06,
+      "loss": 0.2698,
+      "step": 8100
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.9267504904096375,
+      "learning_rate": 4.05e-06,
+      "loss": 0.2522,
+      "step": 8125
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.966682070450956,
+      "learning_rate": 4.0625000000000005e-06,
+      "loss": 0.2976,
+      "step": 8150
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.548744530829551,
+      "learning_rate": 4.075e-06,
+      "loss": 0.2904,
+      "step": 8175
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.753040626050456,
+      "learning_rate": 4.0875e-06,
+      "loss": 0.2843,
+      "step": 8200
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.200730763133159,
+      "learning_rate": 4.1e-06,
+      "loss": 0.2648,
+      "step": 8225
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.411993308970435,
+      "learning_rate": 4.1125e-06,
+      "loss": 0.2949,
+      "step": 8250
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.226554884823071,
+      "learning_rate": 4.125e-06,
+      "loss": 0.2606,
+      "step": 8275
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.5307361977359895,
+      "learning_rate": 4.137500000000001e-06,
+      "loss": 0.275,
+      "step": 8300
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.482533214728566,
+      "learning_rate": 4.15e-06,
+      "loss": 0.2997,
+      "step": 8325
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.674706679093602,
+      "learning_rate": 4.1625e-06,
+      "loss": 0.2955,
+      "step": 8350
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.992206297283636,
+      "learning_rate": 4.175e-06,
+      "loss": 0.2743,
+      "step": 8375
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.042025737938246,
+      "learning_rate": 4.1875e-06,
+      "loss": 0.2671,
+      "step": 8400
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.055250933615206,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.2577,
+      "step": 8425
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.836279970090582,
+      "learning_rate": 4.212500000000001e-06,
+      "loss": 0.2959,
+      "step": 8450
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.347072802266105,
+      "learning_rate": 4.225e-06,
+      "loss": 0.3041,
+      "step": 8475
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.126442118336176,
+      "learning_rate": 4.2375000000000005e-06,
+      "loss": 0.2882,
+      "step": 8500
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.18958310982247,
+      "learning_rate": 4.25e-06,
+      "loss": 0.3015,
+      "step": 8525
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.197927352860307,
+      "learning_rate": 4.2625e-06,
+      "loss": 0.2791,
+      "step": 8550
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.522494327067332,
+      "learning_rate": 4.2750000000000006e-06,
+      "loss": 0.2736,
+      "step": 8575
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.237403417855199,
+      "learning_rate": 4.287500000000001e-06,
+      "loss": 0.2924,
+      "step": 8600
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.3182172892190565,
+      "learning_rate": 4.3e-06,
+      "loss": 0.2824,
+      "step": 8625
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.816379144060154,
+      "learning_rate": 4.312500000000001e-06,
+      "loss": 0.2843,
+      "step": 8650
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.895548452165421,
+      "learning_rate": 4.325e-06,
+      "loss": 0.284,
+      "step": 8675
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.654941433869707,
+      "learning_rate": 4.3375e-06,
+      "loss": 0.2705,
+      "step": 8700
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.469132180677638,
+      "learning_rate": 4.350000000000001e-06,
+      "loss": 0.2729,
+      "step": 8725
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.44234959433828,
+      "learning_rate": 4.362500000000001e-06,
+      "loss": 0.2644,
+      "step": 8750
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.838324842695495,
+      "learning_rate": 4.3750000000000005e-06,
+      "loss": 0.2676,
+      "step": 8775
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.085540132092943,
+      "learning_rate": 4.3875e-06,
+      "loss": 0.2869,
+      "step": 8800
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.795754152011421,
+      "learning_rate": 4.4e-06,
+      "loss": 0.276,
+      "step": 8825
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.102982969922736,
+      "learning_rate": 4.4125000000000005e-06,
+      "loss": 0.2595,
+      "step": 8850
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.700990648549398,
+      "learning_rate": 4.425e-06,
+      "loss": 0.2812,
+      "step": 8875
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.312933654551669,
+      "learning_rate": 4.4375e-06,
+      "loss": 0.2852,
+      "step": 8900
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.379757647366656,
+      "learning_rate": 4.450000000000001e-06,
+      "loss": 0.2779,
+      "step": 8925
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.792550628543418,
+      "learning_rate": 4.4625e-06,
+      "loss": 0.2793,
+      "step": 8950
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.945063099925784,
+      "learning_rate": 4.475e-06,
+      "loss": 0.291,
+      "step": 8975
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.365832685836148,
+      "learning_rate": 4.4875e-06,
+      "loss": 0.275,
+      "step": 9000
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.188639752527868,
+      "learning_rate": 4.5e-06,
+      "loss": 0.3043,
+      "step": 9025
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.383012165041406,
+      "learning_rate": 4.5125e-06,
+      "loss": 0.2791,
+      "step": 9050
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.5648061171749745,
+      "learning_rate": 4.5245000000000005e-06,
+      "loss": 0.2812,
+      "step": 9075
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.718625044735663,
+      "learning_rate": 4.537e-06,
+      "loss": 0.2817,
+      "step": 9100
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.578986129922235,
+      "learning_rate": 4.5495e-06,
+      "loss": 0.3059,
+      "step": 9125
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.950562990896707,
+      "learning_rate": 4.5620000000000005e-06,
+      "loss": 0.2991,
+      "step": 9150
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.6274318125177105,
+      "learning_rate": 4.574500000000001e-06,
+      "loss": 0.2905,
+      "step": 9175
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.868725872074689,
+      "learning_rate": 4.587e-06,
+      "loss": 0.3046,
+      "step": 9200
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.514705728312666,
+      "learning_rate": 4.599500000000001e-06,
+      "loss": 0.2778,
+      "step": 9225
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.37394309344846,
+      "learning_rate": 4.612e-06,
+      "loss": 0.2796,
+      "step": 9250
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.710149784112031,
+      "learning_rate": 4.6245e-06,
+      "loss": 0.2921,
+      "step": 9275
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.545639520621682,
+      "learning_rate": 4.637000000000001e-06,
+      "loss": 0.268,
+      "step": 9300
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.688131287423445,
+      "learning_rate": 4.6495e-06,
+      "loss": 0.2524,
+      "step": 9325
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.135284919662475,
+      "learning_rate": 4.6620000000000004e-06,
+      "loss": 0.276,
+      "step": 9350
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.688180517230278,
+      "learning_rate": 4.6745e-06,
+      "loss": 0.2862,
+      "step": 9375
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.795335976277467,
+      "learning_rate": 4.687e-06,
+      "loss": 0.2888,
+      "step": 9400
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.746370462450612,
+      "learning_rate": 4.6995000000000005e-06,
+      "loss": 0.2662,
+      "step": 9425
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.769522511707671,
+      "learning_rate": 4.712000000000001e-06,
+      "loss": 0.2984,
+      "step": 9450
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.346844756223433,
+      "learning_rate": 4.7245e-06,
+      "loss": 0.2596,
+      "step": 9475
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.3781982999908795,
+      "learning_rate": 4.7370000000000006e-06,
+      "loss": 0.2792,
+      "step": 9500
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.069856946236931,
+      "learning_rate": 4.7495e-06,
+      "loss": 0.2762,
+      "step": 9525
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.564725582025151,
+      "learning_rate": 4.762e-06,
+      "loss": 0.2868,
+      "step": 9550
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.831607241976059,
+      "learning_rate": 4.774500000000001e-06,
+      "loss": 0.2633,
+      "step": 9575
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.993570224205531,
+      "learning_rate": 4.787000000000001e-06,
+      "loss": 0.2907,
+      "step": 9600
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.473582625195408,
+      "learning_rate": 4.7995e-06,
+      "loss": 0.2939,
+      "step": 9625
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.8096853747215675,
+      "learning_rate": 4.812000000000001e-06,
+      "loss": 0.2694,
+      "step": 9650
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.871920620924059,
+      "learning_rate": 4.8245e-06,
+      "loss": 0.2796,
+      "step": 9675
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.914335893785346,
+      "learning_rate": 4.8370000000000004e-06,
+      "loss": 0.2843,
+      "step": 9700
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.459462930577671,
+      "learning_rate": 4.8495e-06,
+      "loss": 0.273,
+      "step": 9725
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.001699192626203,
+      "learning_rate": 4.862e-06,
+      "loss": 0.2588,
+      "step": 9750
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.407971853644161,
+      "learning_rate": 4.8745000000000005e-06,
+      "loss": 0.2705,
+      "step": 9775
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.359611800340205,
+      "learning_rate": 4.887000000000001e-06,
+      "loss": 0.2818,
+      "step": 9800
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.385058355738006,
+      "learning_rate": 4.8995e-06,
+      "loss": 0.2821,
+      "step": 9825
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.8879065529077135,
+      "learning_rate": 4.9120000000000006e-06,
+      "loss": 0.2879,
+      "step": 9850
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.521614128029462,
+      "learning_rate": 4.9245e-06,
+      "loss": 0.2701,
+      "step": 9875
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.22001469157038,
+      "learning_rate": 4.937e-06,
+      "loss": 0.2814,
+      "step": 9900
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.645954802312944,
+      "learning_rate": 4.949500000000001e-06,
+      "loss": 0.2788,
+      "step": 9925
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.975705959262324,
+      "learning_rate": 4.962e-06,
+      "loss": 0.2745,
+      "step": 9950
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.7617079563479345,
+      "learning_rate": 4.9745e-06,
+      "loss": 0.2882,
+      "step": 9975
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.349127933186041,
+      "learning_rate": 4.987e-06,
+      "loss": 0.2717,
+      "step": 10000
+    },
+    {
+      "epoch": 0.39,
+      "eval_loss": 0.414306640625,
+      "eval_runtime": 11566.661,
+      "eval_samples_per_second": 0.818,
+      "eval_steps_per_second": 0.051,
+      "eval_wer": 0.1341312356224764,
+      "step": 10000
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.011108784934462,
+      "learning_rate": 4.9995e-06,
+      "loss": 0.2786,
+      "step": 10025
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.8292619191203014,
+      "learning_rate": 4.999959866220736e-06,
+      "loss": 0.2727,
+      "step": 10050
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.723216450366968,
+      "learning_rate": 4.999919732441472e-06,
+      "loss": 0.2894,
+      "step": 10075
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.24048520411389,
+      "learning_rate": 4.999877926421406e-06,
+      "loss": 0.2752,
+      "step": 10100
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.342763382928159,
+      "learning_rate": 4.999836120401338e-06,
+      "loss": 0.2667,
+      "step": 10125
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.50752715283013,
+      "learning_rate": 4.999794314381271e-06,
+      "loss": 0.2939,
+      "step": 10150
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.204906569121375,
+      "learning_rate": 4.9997525083612046e-06,
+      "loss": 0.2718,
+      "step": 10175
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.752575646886806,
+      "learning_rate": 4.999710702341137e-06,
+      "loss": 0.2747,
+      "step": 10200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.796030259684187,
+      "learning_rate": 4.999668896321071e-06,
+      "loss": 0.2902,
+      "step": 10225
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.053195409587687,
+      "learning_rate": 4.9996270903010035e-06,
+      "loss": 0.3108,
+      "step": 10250
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.059127631079693,
+      "learning_rate": 4.999585284280937e-06,
+      "loss": 0.2775,
+      "step": 10275
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.383155938148739,
+      "learning_rate": 4.99954347826087e-06,
+      "loss": 0.3066,
+      "step": 10300
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.561402837665798,
+      "learning_rate": 4.999501672240803e-06,
+      "loss": 0.2798,
+      "step": 10325
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.225986990831043,
+      "learning_rate": 4.999459866220736e-06,
+      "loss": 0.2957,
+      "step": 10350
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.381995878211679,
+      "learning_rate": 4.9994180602006695e-06,
+      "loss": 0.2824,
+      "step": 10375
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.422499531575439,
+      "learning_rate": 4.999376254180602e-06,
+      "loss": 0.2682,
+      "step": 10400
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.429046461471528,
+      "learning_rate": 4.999334448160536e-06,
+      "loss": 0.2909,
+      "step": 10425
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.823072159394408,
+      "learning_rate": 4.9992926421404685e-06,
+      "loss": 0.2588,
+      "step": 10450
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.64555406872812,
+      "learning_rate": 4.999250836120402e-06,
+      "loss": 0.2675,
+      "step": 10475
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.285451019452604,
+      "learning_rate": 4.999209030100335e-06,
+      "loss": 0.2743,
+      "step": 10500
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.804120728206725,
+      "learning_rate": 4.999167224080268e-06,
+      "loss": 0.288,
+      "step": 10525
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.098440922169357,
+      "learning_rate": 4.999125418060201e-06,
+      "loss": 0.272,
+      "step": 10550
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.546611829047939,
+      "learning_rate": 4.9990836120401345e-06,
+      "loss": 0.2883,
+      "step": 10575
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.622357797847333,
+      "learning_rate": 4.999041806020067e-06,
+      "loss": 0.2577,
+      "step": 10600
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.001323909432522,
+      "learning_rate": 4.999000000000001e-06,
+      "loss": 0.2925,
+      "step": 10625
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.211276512696046,
+      "learning_rate": 4.9989581939799335e-06,
+      "loss": 0.2711,
+      "step": 10650
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.565052395271628,
+      "learning_rate": 4.998916387959867e-06,
+      "loss": 0.2988,
+      "step": 10675
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.296027910196759,
+      "learning_rate": 4.9988745819398e-06,
+      "loss": 0.2773,
+      "step": 10700
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.344517029275454,
+      "learning_rate": 4.998832775919733e-06,
+      "loss": 0.257,
+      "step": 10725
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.746065840605949,
+      "learning_rate": 4.998790969899666e-06,
+      "loss": 0.2832,
+      "step": 10750
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.779174959859481,
+      "learning_rate": 4.9987491638795995e-06,
+      "loss": 0.281,
+      "step": 10775
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.109047891868748,
+      "learning_rate": 4.998707357859532e-06,
+      "loss": 0.2972,
+      "step": 10800
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.426495294412173,
+      "learning_rate": 4.998665551839466e-06,
+      "loss": 0.2689,
+      "step": 10825
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.24421870740098,
+      "learning_rate": 4.9986237458193984e-06,
+      "loss": 0.2659,
+      "step": 10850
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.081586642876751,
+      "learning_rate": 4.998581939799331e-06,
+      "loss": 0.2672,
+      "step": 10875
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.126266886484015,
+      "learning_rate": 4.998540133779265e-06,
+      "loss": 0.2675,
+      "step": 10900
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.726554912145362,
+      "learning_rate": 4.998498327759197e-06,
+      "loss": 0.2881,
+      "step": 10925
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.749427772911198,
+      "learning_rate": 4.998456521739131e-06,
+      "loss": 0.297,
+      "step": 10950
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.598962835292076,
+      "learning_rate": 4.998414715719064e-06,
+      "loss": 0.2978,
+      "step": 10975
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.200710604353594,
+      "learning_rate": 4.998372909698997e-06,
+      "loss": 0.2692,
+      "step": 11000
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.702673268567435,
+      "learning_rate": 4.99833110367893e-06,
+      "loss": 0.2944,
+      "step": 11025
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.625028895002169,
+      "learning_rate": 4.998289297658863e-06,
+      "loss": 0.2885,
+      "step": 11050
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.820444832366389,
+      "learning_rate": 4.998249163879598e-06,
+      "loss": 0.2942,
+      "step": 11075
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.567604834669745,
+      "learning_rate": 4.998207357859532e-06,
+      "loss": 0.2742,
+      "step": 11100
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.423735455535033,
+      "learning_rate": 4.998165551839465e-06,
+      "loss": 0.2846,
+      "step": 11125
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.999616850642319,
+      "learning_rate": 4.998123745819398e-06,
+      "loss": 0.2782,
+      "step": 11150
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.030451571678514,
+      "learning_rate": 4.998081939799332e-06,
+      "loss": 0.3123,
+      "step": 11175
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.9115247740062875,
+      "learning_rate": 4.9980401337792644e-06,
+      "loss": 0.2536,
+      "step": 11200
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.072557616156972,
+      "learning_rate": 4.997998327759198e-06,
+      "loss": 0.2915,
+      "step": 11225
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.711391798767144,
+      "learning_rate": 4.997956521739131e-06,
+      "loss": 0.2736,
+      "step": 11250
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.645455721566767,
+      "learning_rate": 4.997914715719064e-06,
+      "loss": 0.288,
+      "step": 11275
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.186122138761158,
+      "learning_rate": 4.997872909698997e-06,
+      "loss": 0.2697,
+      "step": 11300
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.00329145846993,
+      "learning_rate": 4.9978311036789305e-06,
+      "loss": 0.2924,
+      "step": 11325
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.312139760803962,
+      "learning_rate": 4.997789297658863e-06,
+      "loss": 0.2756,
+      "step": 11350
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.8874897753934965,
+      "learning_rate": 4.997747491638797e-06,
+      "loss": 0.2703,
+      "step": 11375
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.211561229810027,
+      "learning_rate": 4.997705685618729e-06,
+      "loss": 0.2724,
+      "step": 11400
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.444390049063497,
+      "learning_rate": 4.997663879598663e-06,
+      "loss": 0.2877,
+      "step": 11425
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.170573267490551,
+      "learning_rate": 4.997622073578596e-06,
+      "loss": 0.259,
+      "step": 11450
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.632322155697068,
+      "learning_rate": 4.997580267558529e-06,
+      "loss": 0.2823,
+      "step": 11475
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.152734308462785,
+      "learning_rate": 4.997538461538462e-06,
+      "loss": 0.2607,
+      "step": 11500
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.107228957537137,
+      "learning_rate": 4.9974966555183954e-06,
+      "loss": 0.2746,
+      "step": 11525
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.472917184920453,
+      "learning_rate": 4.997454849498328e-06,
+      "loss": 0.2728,
+      "step": 11550
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.83478467904063,
+      "learning_rate": 4.997413043478262e-06,
+      "loss": 0.2751,
+      "step": 11575
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.279889559058508,
+      "learning_rate": 4.997371237458194e-06,
+      "loss": 0.2659,
+      "step": 11600
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.228962539059828,
+      "learning_rate": 4.997329431438128e-06,
+      "loss": 0.2946,
+      "step": 11625
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.329870520493551,
+      "learning_rate": 4.997287625418061e-06,
+      "loss": 0.2752,
+      "step": 11650
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.644595157995969,
+      "learning_rate": 4.997245819397994e-06,
+      "loss": 0.2782,
+      "step": 11675
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.922993574833506,
+      "learning_rate": 4.997204013377927e-06,
+      "loss": 0.277,
+      "step": 11700
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.510974001369662,
+      "learning_rate": 4.99716220735786e-06,
+      "loss": 0.2752,
+      "step": 11725
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.9210546984842365,
+      "learning_rate": 4.997120401337793e-06,
+      "loss": 0.2634,
+      "step": 11750
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.079776486981066,
+      "learning_rate": 4.997078595317727e-06,
+      "loss": 0.2647,
+      "step": 11775
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.305360379026096,
+      "learning_rate": 4.997036789297659e-06,
+      "loss": 0.2988,
+      "step": 11800
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.735797578726933,
+      "learning_rate": 4.996994983277592e-06,
+      "loss": 0.2484,
+      "step": 11825
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.43170191366388,
+      "learning_rate": 4.996953177257526e-06,
+      "loss": 0.2619,
+      "step": 11850
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.402374846991611,
+      "learning_rate": 4.996911371237458e-06,
+      "loss": 0.2746,
+      "step": 11875
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.531743806890973,
+      "learning_rate": 4.996869565217392e-06,
+      "loss": 0.2665,
+      "step": 11900
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.839865759331419,
+      "learning_rate": 4.9968277591973245e-06,
+      "loss": 0.2702,
+      "step": 11925
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.825681200003497,
+      "learning_rate": 4.996785953177258e-06,
+      "loss": 0.2531,
+      "step": 11950
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.125879741698381,
+      "learning_rate": 4.996744147157191e-06,
+      "loss": 0.2704,
+      "step": 11975
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.663308145081337,
+      "learning_rate": 4.996702341137124e-06,
+      "loss": 0.2653,
+      "step": 12000
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.184543925229218,
+      "learning_rate": 4.996660535117057e-06,
+      "loss": 0.2786,
+      "step": 12025
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.865850646784692,
+      "learning_rate": 4.996618729096991e-06,
+      "loss": 0.2781,
+      "step": 12050
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.13013469457717,
+      "learning_rate": 4.9965785953177256e-06,
+      "loss": 0.2725,
+      "step": 12075
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.6434695533203305,
+      "learning_rate": 4.996536789297659e-06,
+      "loss": 0.2852,
+      "step": 12100
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.9799882279278265,
+      "learning_rate": 4.996494983277592e-06,
+      "loss": 0.2699,
+      "step": 12125
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.060616926983353,
+      "learning_rate": 4.996453177257525e-06,
+      "loss": 0.2752,
+      "step": 12150
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.746693589734832,
+      "learning_rate": 4.996411371237458e-06,
+      "loss": 0.2624,
+      "step": 12175
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.229583616728782,
+      "learning_rate": 4.996369565217392e-06,
+      "loss": 0.2515,
+      "step": 12200
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.999145743899038,
+      "learning_rate": 4.996327759197324e-06,
+      "loss": 0.2775,
+      "step": 12225
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.927319188266453,
+      "learning_rate": 4.996285953177258e-06,
+      "loss": 0.2846,
+      "step": 12250
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.874400599413486,
+      "learning_rate": 4.996244147157191e-06,
+      "loss": 0.2758,
+      "step": 12275
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.024826067369364,
+      "learning_rate": 4.996202341137124e-06,
+      "loss": 0.2733,
+      "step": 12300
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.724310991701934,
+      "learning_rate": 4.996160535117058e-06,
+      "loss": 0.2661,
+      "step": 12325
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.190847168134574,
+      "learning_rate": 4.99611872909699e-06,
+      "loss": 0.3009,
+      "step": 12350
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.981795774000632,
+      "learning_rate": 4.996076923076924e-06,
+      "loss": 0.2714,
+      "step": 12375
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.982768726819535,
+      "learning_rate": 4.9960351170568566e-06,
+      "loss": 0.2452,
+      "step": 12400
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.30852279239966,
+      "learning_rate": 4.99599331103679e-06,
+      "loss": 0.2584,
+      "step": 12425
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.858438111584143,
+      "learning_rate": 4.995951505016723e-06,
+      "loss": 0.2505,
+      "step": 12450
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.174552548362399,
+      "learning_rate": 4.995909698996656e-06,
+      "loss": 0.2532,
+      "step": 12475
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.0465648379078125,
+      "learning_rate": 4.995867892976589e-06,
+      "loss": 0.265,
+      "step": 12500
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.988153680687835,
+      "learning_rate": 4.995826086956523e-06,
+      "loss": 0.2432,
+      "step": 12525
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.264922296051626,
+      "learning_rate": 4.995784280936455e-06,
+      "loss": 0.2657,
+      "step": 12550
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.449845098914649,
+      "learning_rate": 4.995742474916389e-06,
+      "loss": 0.2708,
+      "step": 12575
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.009503417055907,
+      "learning_rate": 4.995702341137124e-06,
+      "loss": 0.2522,
+      "step": 12600
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.1804236684247345,
+      "learning_rate": 4.995660535117057e-06,
+      "loss": 0.2539,
+      "step": 12625
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.543935093810235,
+      "learning_rate": 4.99561872909699e-06,
+      "loss": 0.2672,
+      "step": 12650
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.471815945203869,
+      "learning_rate": 4.995576923076924e-06,
+      "loss": 0.2952,
+      "step": 12675
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.501833406890489,
+      "learning_rate": 4.995535117056856e-06,
+      "loss": 0.2667,
+      "step": 12700
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.343214654642585,
+      "learning_rate": 4.99549331103679e-06,
+      "loss": 0.2872,
+      "step": 12725
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.271797540434532,
+      "learning_rate": 4.9954515050167226e-06,
+      "loss": 0.2683,
+      "step": 12750
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.599139231991323,
+      "learning_rate": 4.995409698996656e-06,
+      "loss": 0.263,
+      "step": 12775
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.002476348463217,
+      "learning_rate": 4.995367892976589e-06,
+      "loss": 0.2783,
+      "step": 12800
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.525297588148386,
+      "learning_rate": 4.995326086956522e-06,
+      "loss": 0.2871,
+      "step": 12825
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.708477915446077,
+      "learning_rate": 4.995284280936455e-06,
+      "loss": 0.2691,
+      "step": 12850
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.6843516447099445,
+      "learning_rate": 4.995242474916389e-06,
+      "loss": 0.2871,
+      "step": 12875
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.115815653336138,
+      "learning_rate": 4.995200668896321e-06,
+      "loss": 0.2425,
+      "step": 12900
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.516585844765604,
+      "learning_rate": 4.995158862876255e-06,
+      "loss": 0.2576,
+      "step": 12925
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.532827849171224,
+      "learning_rate": 4.995117056856188e-06,
+      "loss": 0.2921,
+      "step": 12950
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.0067926091682144,
+      "learning_rate": 4.99507525083612e-06,
+      "loss": 0.2559,
+      "step": 12975
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.750130129977456,
+      "learning_rate": 4.995033444816054e-06,
+      "loss": 0.264,
+      "step": 13000
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.321056688567065,
+      "learning_rate": 4.9949916387959865e-06,
+      "loss": 0.2627,
+      "step": 13025
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.063866279377709,
+      "learning_rate": 4.99494983277592e-06,
+      "loss": 0.2723,
+      "step": 13050
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.685284661380941,
+      "learning_rate": 4.994908026755853e-06,
+      "loss": 0.2791,
+      "step": 13075
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.645607817332848,
+      "learning_rate": 4.994866220735786e-06,
+      "loss": 0.2569,
+      "step": 13100
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.818177466809242,
+      "learning_rate": 4.994824414715719e-06,
+      "loss": 0.2652,
+      "step": 13125
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.567556521092891,
+      "learning_rate": 4.9947826086956525e-06,
+      "loss": 0.2722,
+      "step": 13150
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.049853304178241,
+      "learning_rate": 4.994740802675585e-06,
+      "loss": 0.2521,
+      "step": 13175
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.525962556248073,
+      "learning_rate": 4.994698996655519e-06,
+      "loss": 0.2782,
+      "step": 13200
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.2979035414331035,
+      "learning_rate": 4.9946571906354515e-06,
+      "loss": 0.2672,
+      "step": 13225
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.494928513364084,
+      "learning_rate": 4.994615384615385e-06,
+      "loss": 0.2705,
+      "step": 13250
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.084597225504904,
+      "learning_rate": 4.994573578595318e-06,
+      "loss": 0.2883,
+      "step": 13275
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.072120776961249,
+      "learning_rate": 4.994531772575251e-06,
+      "loss": 0.2588,
+      "step": 13300
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.741904082206531,
+      "learning_rate": 4.994489966555184e-06,
+      "loss": 0.2687,
+      "step": 13325
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.687864286901733,
+      "learning_rate": 4.9944481605351175e-06,
+      "loss": 0.2618,
+      "step": 13350
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.279456079341556,
+      "learning_rate": 4.994406354515051e-06,
+      "loss": 0.2788,
+      "step": 13375
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.034957788594076,
+      "learning_rate": 4.994364548494984e-06,
+      "loss": 0.2425,
+      "step": 13400
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.552783387380151,
+      "learning_rate": 4.994322742474917e-06,
+      "loss": 0.251,
+      "step": 13425
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.177948746144479,
+      "learning_rate": 4.99428093645485e-06,
+      "loss": 0.2782,
+      "step": 13450
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.068512146023676,
+      "learning_rate": 4.9942391304347835e-06,
+      "loss": 0.2608,
+      "step": 13475
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.744283896920329,
+      "learning_rate": 4.994197324414716e-06,
+      "loss": 0.2564,
+      "step": 13500
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.126944376370206,
+      "learning_rate": 4.99415551839465e-06,
+      "loss": 0.2673,
+      "step": 13525
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.907130426240397,
+      "learning_rate": 4.9941137123745825e-06,
+      "loss": 0.2547,
+      "step": 13550
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.4639597192774865,
+      "learning_rate": 4.994071906354516e-06,
+      "loss": 0.2795,
+      "step": 13575
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.672162745544116,
+      "learning_rate": 4.994030100334449e-06,
+      "loss": 0.289,
+      "step": 13600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.72396760446355,
+      "learning_rate": 4.993988294314382e-06,
+      "loss": 0.2711,
+      "step": 13625
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.555801546793623,
+      "learning_rate": 4.993946488294315e-06,
+      "loss": 0.2666,
+      "step": 13650
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.9331030728145016,
+      "learning_rate": 4.993904682274248e-06,
+      "loss": 0.243,
+      "step": 13675
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.11331974297062,
+      "learning_rate": 4.99386287625418e-06,
+      "loss": 0.2442,
+      "step": 13700
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.512882588773397,
+      "learning_rate": 4.993821070234114e-06,
+      "loss": 0.2622,
+      "step": 13725
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.5372514495460905,
+      "learning_rate": 4.993779264214047e-06,
+      "loss": 0.299,
+      "step": 13750
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.888509838266094,
+      "learning_rate": 4.99373745819398e-06,
+      "loss": 0.2575,
+      "step": 13775
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.748557291602811,
+      "learning_rate": 4.993695652173914e-06,
+      "loss": 0.2668,
+      "step": 13800
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.6182807926195215,
+      "learning_rate": 4.993653846153846e-06,
+      "loss": 0.2569,
+      "step": 13825
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.067375657021155,
+      "learning_rate": 4.99361204013378e-06,
+      "loss": 0.2417,
+      "step": 13850
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.0814238478732365,
+      "learning_rate": 4.993570234113713e-06,
+      "loss": 0.284,
+      "step": 13875
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.849375990146647,
+      "learning_rate": 4.993528428093646e-06,
+      "loss": 0.2586,
+      "step": 13900
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.420173287461275,
+      "learning_rate": 4.993486622073579e-06,
+      "loss": 0.2685,
+      "step": 13925
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.510360107745434,
+      "learning_rate": 4.9934448160535124e-06,
+      "loss": 0.2622,
+      "step": 13950
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.744495569021254,
+      "learning_rate": 4.993403010033445e-06,
+      "loss": 0.2654,
+      "step": 13975
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.241624718908685,
+      "learning_rate": 4.993361204013379e-06,
+      "loss": 0.2575,
+      "step": 14000
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.964854463517818,
+      "learning_rate": 4.993319397993311e-06,
+      "loss": 0.2601,
+      "step": 14025
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.98667622104724,
+      "learning_rate": 4.993277591973245e-06,
+      "loss": 0.2631,
+      "step": 14050
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.991177855529151,
+      "learning_rate": 4.993235785953178e-06,
+      "loss": 0.2711,
+      "step": 14075
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.205578681842165,
+      "learning_rate": 4.993193979933111e-06,
+      "loss": 0.3043,
+      "step": 14100
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.597102209370411,
+      "learning_rate": 4.993152173913044e-06,
+      "loss": 0.292,
+      "step": 14125
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.747292937249573,
+      "learning_rate": 4.993110367892977e-06,
+      "loss": 0.2496,
+      "step": 14150
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.567122798131745,
+      "learning_rate": 4.99306856187291e-06,
+      "loss": 0.2898,
+      "step": 14175
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.425348552599907,
+      "learning_rate": 4.993026755852844e-06,
+      "loss": 0.2675,
+      "step": 14200
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.923945998272752,
+      "learning_rate": 4.992984949832776e-06,
+      "loss": 0.2693,
+      "step": 14225
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.085703696247624,
+      "learning_rate": 4.99294314381271e-06,
+      "loss": 0.2661,
+      "step": 14250
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.73816088361352,
+      "learning_rate": 4.992901337792643e-06,
+      "loss": 0.2694,
+      "step": 14275
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.809164188200779,
+      "learning_rate": 4.992859531772576e-06,
+      "loss": 0.2719,
+      "step": 14300
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.3598512977073,
+      "learning_rate": 4.992817725752509e-06,
+      "loss": 0.2774,
+      "step": 14325
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.216046308562436,
+      "learning_rate": 4.992775919732442e-06,
+      "loss": 0.247,
+      "step": 14350
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.329199858263562,
+      "learning_rate": 4.992734113712375e-06,
+      "loss": 0.2868,
+      "step": 14375
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.210460931264835,
+      "learning_rate": 4.992692307692308e-06,
+      "loss": 0.2683,
+      "step": 14400
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.7667612517073765,
+      "learning_rate": 4.992650501672241e-06,
+      "loss": 0.2535,
+      "step": 14425
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.9612302730938405,
+      "learning_rate": 4.992608695652174e-06,
+      "loss": 0.2529,
+      "step": 14450
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.596575617259898,
+      "learning_rate": 4.9925668896321076e-06,
+      "loss": 0.253,
+      "step": 14475
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.62301372614655,
+      "learning_rate": 4.99252508361204e-06,
+      "loss": 0.2504,
+      "step": 14500
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.950147967461413,
+      "learning_rate": 4.992483277591974e-06,
+      "loss": 0.2682,
+      "step": 14525
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.881464976782473,
+      "learning_rate": 4.9924414715719065e-06,
+      "loss": 0.2473,
+      "step": 14550
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.705896744318513,
+      "learning_rate": 4.99239966555184e-06,
+      "loss": 0.2468,
+      "step": 14575
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.6827213444842526,
+      "learning_rate": 4.992357859531773e-06,
+      "loss": 0.3114,
+      "step": 14600
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.249966633309746,
+      "learning_rate": 4.992317725752509e-06,
+      "loss": 0.2633,
+      "step": 14625
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.42278315835385,
+      "learning_rate": 4.992275919732441e-06,
+      "loss": 0.2854,
+      "step": 14650
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.655143231460296,
+      "learning_rate": 4.992234113712375e-06,
+      "loss": 0.2655,
+      "step": 14675
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.858506902274973,
+      "learning_rate": 4.9921923076923075e-06,
+      "loss": 0.2913,
+      "step": 14700
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.606517868907838,
+      "learning_rate": 4.992150501672241e-06,
+      "loss": 0.2562,
+      "step": 14725
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.8892802718869275,
+      "learning_rate": 4.992108695652174e-06,
+      "loss": 0.2653,
+      "step": 14750
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.395026098564766,
+      "learning_rate": 4.992066889632107e-06,
+      "loss": 0.2717,
+      "step": 14775
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.043189975643226,
+      "learning_rate": 4.99202508361204e-06,
+      "loss": 0.2587,
+      "step": 14800
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.36609725720899,
+      "learning_rate": 4.9919832775919736e-06,
+      "loss": 0.272,
+      "step": 14825
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.645686026598199,
+      "learning_rate": 4.991941471571906e-06,
+      "loss": 0.2333,
+      "step": 14850
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.55507184083137,
+      "learning_rate": 4.99189966555184e-06,
+      "loss": 0.2509,
+      "step": 14875
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.345846578476869,
+      "learning_rate": 4.991857859531773e-06,
+      "loss": 0.2613,
+      "step": 14900
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.935348438197945,
+      "learning_rate": 4.991816053511706e-06,
+      "loss": 0.2816,
+      "step": 14925
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.803960018432536,
+      "learning_rate": 4.99177424749164e-06,
+      "loss": 0.2548,
+      "step": 14950
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.42912623970475,
+      "learning_rate": 4.991732441471572e-06,
+      "loss": 0.2529,
+      "step": 14975
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.975233188396705,
+      "learning_rate": 4.991690635451506e-06,
+      "loss": 0.2881,
+      "step": 15000
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.22507074373668,
+      "learning_rate": 4.9916488294314385e-06,
+      "loss": 0.2552,
+      "step": 15025
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.14001014159123,
+      "learning_rate": 4.991607023411372e-06,
+      "loss": 0.2684,
+      "step": 15050
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.034722792372274,
+      "learning_rate": 4.991565217391305e-06,
+      "loss": 0.2564,
+      "step": 15075
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.342705074428043,
+      "learning_rate": 4.991523411371238e-06,
+      "loss": 0.2542,
+      "step": 15100
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.154437791807671,
+      "learning_rate": 4.991481605351171e-06,
+      "loss": 0.2728,
+      "step": 15125
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.466502948123444,
+      "learning_rate": 4.9914397993311046e-06,
+      "loss": 0.2759,
+      "step": 15150
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.730270409539509,
+      "learning_rate": 4.991397993311037e-06,
+      "loss": 0.2671,
+      "step": 15175
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.13478451695111,
+      "learning_rate": 4.991356187290971e-06,
+      "loss": 0.2639,
+      "step": 15200
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.63611014055058,
+      "learning_rate": 4.9913143812709035e-06,
+      "loss": 0.2511,
+      "step": 15225
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.486679281838134,
+      "learning_rate": 4.991272575250837e-06,
+      "loss": 0.2615,
+      "step": 15250
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.609027754294971,
+      "learning_rate": 4.99123076923077e-06,
+      "loss": 0.2521,
+      "step": 15275
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.300744039135444,
+      "learning_rate": 4.991188963210703e-06,
+      "loss": 0.2734,
+      "step": 15300
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.302779818219576,
+      "learning_rate": 4.991147157190635e-06,
+      "loss": 0.2771,
+      "step": 15325
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.909444498848482,
+      "learning_rate": 4.991105351170569e-06,
+      "loss": 0.2675,
+      "step": 15350
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.43122899967809,
+      "learning_rate": 4.991063545150502e-06,
+      "loss": 0.2618,
+      "step": 15375
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.228246630467435,
+      "learning_rate": 4.991021739130435e-06,
+      "loss": 0.2687,
+      "step": 15400
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.064148750557926,
+      "learning_rate": 4.9909799331103685e-06,
+      "loss": 0.2428,
+      "step": 15425
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.255096086616142,
+      "learning_rate": 4.990938127090301e-06,
+      "loss": 0.2667,
+      "step": 15450
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.709327247522713,
+      "learning_rate": 4.990896321070235e-06,
+      "loss": 0.2732,
+      "step": 15475
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.184508161358503,
+      "learning_rate": 4.9908545150501674e-06,
+      "loss": 0.2719,
+      "step": 15500
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.922987038531951,
+      "learning_rate": 4.990812709030101e-06,
+      "loss": 0.2797,
+      "step": 15525
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.626376622087191,
+      "learning_rate": 4.990770903010034e-06,
+      "loss": 0.2479,
+      "step": 15550
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.514060439157896,
+      "learning_rate": 4.990729096989967e-06,
+      "loss": 0.2405,
+      "step": 15575
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.373488145308544,
+      "learning_rate": 4.9906872909699e-06,
+      "loss": 0.248,
+      "step": 15600
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.741275477526847,
+      "learning_rate": 4.990647157190636e-06,
+      "loss": 0.253,
+      "step": 15625
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.095305164848746,
+      "learning_rate": 4.9906053511705685e-06,
+      "loss": 0.2635,
+      "step": 15650
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.587285383976124,
+      "learning_rate": 4.990563545150502e-06,
+      "loss": 0.2713,
+      "step": 15675
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.9812853760264115,
+      "learning_rate": 4.990521739130435e-06,
+      "loss": 0.2796,
+      "step": 15700
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.944806936412515,
+      "learning_rate": 4.990479933110368e-06,
+      "loss": 0.2599,
+      "step": 15725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.171130869044126,
+      "learning_rate": 4.990438127090301e-06,
+      "loss": 0.2489,
+      "step": 15750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.206236751191819,
+      "learning_rate": 4.9903963210702345e-06,
+      "loss": 0.2717,
+      "step": 15775
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.974957589959287,
+      "learning_rate": 4.990354515050167e-06,
+      "loss": 0.2632,
+      "step": 15800
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.464737013981678,
+      "learning_rate": 4.990312709030101e-06,
+      "loss": 0.2562,
+      "step": 15825
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.039455268506516,
+      "learning_rate": 4.9902709030100334e-06,
+      "loss": 0.2589,
+      "step": 15850
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.93453838570339,
+      "learning_rate": 4.990229096989967e-06,
+      "loss": 0.254,
+      "step": 15875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.429570899110624,
+      "learning_rate": 4.9901872909699e-06,
+      "loss": 0.2755,
+      "step": 15900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.473116184036116,
+      "learning_rate": 4.990145484949833e-06,
+      "loss": 0.255,
+      "step": 15925
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.245501608124624,
+      "learning_rate": 4.990103678929766e-06,
+      "loss": 0.2323,
+      "step": 15950
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.1289839856825346,
+      "learning_rate": 4.9900618729096995e-06,
+      "loss": 0.2623,
+      "step": 15975
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.717814516086991,
+      "learning_rate": 4.990020066889632e-06,
+      "loss": 0.2414,
+      "step": 16000
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.751851025204562,
+      "learning_rate": 4.989978260869566e-06,
+      "loss": 0.2621,
+      "step": 16025
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.6353456530210755,
+      "learning_rate": 4.989936454849499e-06,
+      "loss": 0.2621,
+      "step": 16050
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.035407794665361,
+      "learning_rate": 4.989894648829432e-06,
+      "loss": 0.2706,
+      "step": 16075
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.175886486573118,
+      "learning_rate": 4.9898528428093655e-06,
+      "loss": 0.2556,
+      "step": 16100
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.306973626946458,
+      "learning_rate": 4.989811036789298e-06,
+      "loss": 0.2532,
+      "step": 16125
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.739524436385124,
+      "learning_rate": 4.989769230769232e-06,
+      "loss": 0.243,
+      "step": 16150
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.743554132202793,
+      "learning_rate": 4.9897274247491644e-06,
+      "loss": 0.2571,
+      "step": 16175
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.507860974296257,
+      "learning_rate": 4.989685618729098e-06,
+      "loss": 0.2498,
+      "step": 16200
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.521567618683169,
+      "learning_rate": 4.989643812709031e-06,
+      "loss": 0.2406,
+      "step": 16225
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.287306100434563,
+      "learning_rate": 4.989602006688964e-06,
+      "loss": 0.2548,
+      "step": 16250
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.605554498785768,
+      "learning_rate": 4.989560200668896e-06,
+      "loss": 0.2735,
+      "step": 16275
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.763241090962355,
+      "learning_rate": 4.98951839464883e-06,
+      "loss": 0.2306,
+      "step": 16300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.236303105187178,
+      "learning_rate": 4.989476588628762e-06,
+      "loss": 0.2406,
+      "step": 16325
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.893192357925432,
+      "learning_rate": 4.989434782608696e-06,
+      "loss": 0.2643,
+      "step": 16350
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.7801672791546945,
+      "learning_rate": 4.9893929765886286e-06,
+      "loss": 0.2557,
+      "step": 16375
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.509427372226329,
+      "learning_rate": 4.989351170568562e-06,
+      "loss": 0.2408,
+      "step": 16400
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.145083984273696,
+      "learning_rate": 4.989309364548495e-06,
+      "loss": 0.2475,
+      "step": 16425
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.961929779970909,
+      "learning_rate": 4.989267558528428e-06,
+      "loss": 0.2543,
+      "step": 16450
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.237616244924249,
+      "learning_rate": 4.989225752508362e-06,
+      "loss": 0.2693,
+      "step": 16475
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.071941176628237,
+      "learning_rate": 4.989183946488295e-06,
+      "loss": 0.2672,
+      "step": 16500
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.99881469452368,
+      "learning_rate": 4.989142140468228e-06,
+      "loss": 0.252,
+      "step": 16525
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.824243326080856,
+      "learning_rate": 4.989100334448161e-06,
+      "loss": 0.2457,
+      "step": 16550
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.86991972909777,
+      "learning_rate": 4.989058528428094e-06,
+      "loss": 0.266,
+      "step": 16575
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.545283659754445,
+      "learning_rate": 4.989016722408027e-06,
+      "loss": 0.2781,
+      "step": 16600
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.925067126921487,
+      "learning_rate": 4.988976588628763e-06,
+      "loss": 0.2954,
+      "step": 16625
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.58821701479221,
+      "learning_rate": 4.988934782608696e-06,
+      "loss": 0.2494,
+      "step": 16650
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.185520155813638,
+      "learning_rate": 4.988892976588629e-06,
+      "loss": 0.2307,
+      "step": 16675
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.039375868942344,
+      "learning_rate": 4.988851170568562e-06,
+      "loss": 0.2454,
+      "step": 16700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.8875232231482615,
+      "learning_rate": 4.988809364548495e-06,
+      "loss": 0.2556,
+      "step": 16725
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.585235280866713,
+      "learning_rate": 4.988767558528428e-06,
+      "loss": 0.2687,
+      "step": 16750
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.796602212489863,
+      "learning_rate": 4.988725752508362e-06,
+      "loss": 0.2573,
+      "step": 16775
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.1627805195183685,
+      "learning_rate": 4.988683946488294e-06,
+      "loss": 0.2932,
+      "step": 16800
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.552009029354894,
+      "learning_rate": 4.988642140468228e-06,
+      "loss": 0.2707,
+      "step": 16825
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.466725002249184,
+      "learning_rate": 4.988600334448161e-06,
+      "loss": 0.2336,
+      "step": 16850
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.186062157421561,
+      "learning_rate": 4.988558528428094e-06,
+      "loss": 0.2538,
+      "step": 16875
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.080865810052987,
+      "learning_rate": 4.988516722408027e-06,
+      "loss": 0.2464,
+      "step": 16900
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.570893467223135,
+      "learning_rate": 4.98847491638796e-06,
+      "loss": 0.2555,
+      "step": 16925
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.9998578151996504,
+      "learning_rate": 4.988433110367893e-06,
+      "loss": 0.2685,
+      "step": 16950
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.714104620985236,
+      "learning_rate": 4.988391304347827e-06,
+      "loss": 0.2579,
+      "step": 16975
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.204684812582076,
+      "learning_rate": 4.988349498327759e-06,
+      "loss": 0.2814,
+      "step": 17000
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.184079656953484,
+      "learning_rate": 4.988307692307693e-06,
+      "loss": 0.2413,
+      "step": 17025
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.268387433526686,
+      "learning_rate": 4.988265886287626e-06,
+      "loss": 0.2817,
+      "step": 17050
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.257289805906091,
+      "learning_rate": 4.988224080267559e-06,
+      "loss": 0.2595,
+      "step": 17075
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.0383919249393045,
+      "learning_rate": 4.988182274247492e-06,
+      "loss": 0.2666,
+      "step": 17100
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.291305025444263,
+      "learning_rate": 4.988140468227425e-06,
+      "loss": 0.2654,
+      "step": 17125
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.340130958791915,
+      "learning_rate": 4.988098662207359e-06,
+      "loss": 0.2525,
+      "step": 17150
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.911613542246811,
+      "learning_rate": 4.988056856187292e-06,
+      "loss": 0.2788,
+      "step": 17175
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.477361220204573,
+      "learning_rate": 4.988015050167225e-06,
+      "loss": 0.2522,
+      "step": 17200
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.792368739302577,
+      "learning_rate": 4.987973244147157e-06,
+      "loss": 0.2724,
+      "step": 17225
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.35695367909829,
+      "learning_rate": 4.9879314381270906e-06,
+      "loss": 0.2761,
+      "step": 17250
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.629047017884108,
+      "learning_rate": 4.987889632107023e-06,
+      "loss": 0.2635,
+      "step": 17275
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.331465117440273,
+      "learning_rate": 4.987847826086957e-06,
+      "loss": 0.2414,
+      "step": 17300
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.367627157757305,
+      "learning_rate": 4.9878060200668895e-06,
+      "loss": 0.2572,
+      "step": 17325
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.1172264981864295,
+      "learning_rate": 4.987764214046823e-06,
+      "loss": 0.2577,
+      "step": 17350
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.355305084531797,
+      "learning_rate": 4.987724080267559e-06,
+      "loss": 0.2792,
+      "step": 17375
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.842697216641398,
+      "learning_rate": 4.987682274247492e-06,
+      "loss": 0.2617,
+      "step": 17400
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.743056099274135,
+      "learning_rate": 4.987640468227425e-06,
+      "loss": 0.2488,
+      "step": 17425
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.36774370696684,
+      "learning_rate": 4.987598662207358e-06,
+      "loss": 0.2777,
+      "step": 17450
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.700174425799506,
+      "learning_rate": 4.987556856187291e-06,
+      "loss": 0.2445,
+      "step": 17475
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.5291577524920905,
+      "learning_rate": 4.987515050167224e-06,
+      "loss": 0.2391,
+      "step": 17500
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.0434604227152455,
+      "learning_rate": 4.987473244147158e-06,
+      "loss": 0.2868,
+      "step": 17525
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.471288503953715,
+      "learning_rate": 4.98743143812709e-06,
+      "loss": 0.294,
+      "step": 17550
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.967607026828377,
+      "learning_rate": 4.987389632107024e-06,
+      "loss": 0.2582,
+      "step": 17575
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.493557381929855,
+      "learning_rate": 4.9873478260869566e-06,
+      "loss": 0.2337,
+      "step": 17600
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.557469040585438,
+      "learning_rate": 4.98730602006689e-06,
+      "loss": 0.2703,
+      "step": 17625
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.523565121186858,
+      "learning_rate": 4.987264214046823e-06,
+      "loss": 0.2577,
+      "step": 17650
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.148642857874085,
+      "learning_rate": 4.987222408026756e-06,
+      "loss": 0.252,
+      "step": 17675
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.875580443467786,
+      "learning_rate": 4.987180602006689e-06,
+      "loss": 0.278,
+      "step": 17700
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.99454936628081,
+      "learning_rate": 4.987138795986623e-06,
+      "loss": 0.2721,
+      "step": 17725
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.555180518575604,
+      "learning_rate": 4.987096989966555e-06,
+      "loss": 0.2533,
+      "step": 17750
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.395100579611774,
+      "learning_rate": 4.987055183946489e-06,
+      "loss": 0.2625,
+      "step": 17775
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.563405837664269,
+      "learning_rate": 4.9870133779264215e-06,
+      "loss": 0.2622,
+      "step": 17800
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.043632738488831,
+      "learning_rate": 4.986971571906355e-06,
+      "loss": 0.2845,
+      "step": 17825
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.119469778219465,
+      "learning_rate": 4.986929765886288e-06,
+      "loss": 0.251,
+      "step": 17850
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.75507347596451,
+      "learning_rate": 4.986887959866221e-06,
+      "loss": 0.251,
+      "step": 17875
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.826211048039959,
+      "learning_rate": 4.986846153846154e-06,
+      "loss": 0.2691,
+      "step": 17900
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.010072111561531,
+      "learning_rate": 4.9868043478260876e-06,
+      "loss": 0.2366,
+      "step": 17925
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.859353986359138,
+      "learning_rate": 4.98676254180602e-06,
+      "loss": 0.2688,
+      "step": 17950
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.436740266722265,
+      "learning_rate": 4.986720735785954e-06,
+      "loss": 0.2287,
+      "step": 17975
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.699368648294974,
+      "learning_rate": 4.9866789297658865e-06,
+      "loss": 0.2623,
+      "step": 18000
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.286788999671542,
+      "learning_rate": 4.98663712374582e-06,
+      "loss": 0.2675,
+      "step": 18025
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.383383566653747,
+      "learning_rate": 4.986595317725753e-06,
+      "loss": 0.2483,
+      "step": 18050
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.558217801754021,
+      "learning_rate": 4.986553511705686e-06,
+      "loss": 0.2542,
+      "step": 18075
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.38291780085364,
+      "learning_rate": 4.986511705685619e-06,
+      "loss": 0.2497,
+      "step": 18100
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.520472839091616,
+      "learning_rate": 4.9864698996655525e-06,
+      "loss": 0.2574,
+      "step": 18125
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.77899108452888,
+      "learning_rate": 4.986428093645485e-06,
+      "loss": 0.264,
+      "step": 18150
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.778959356886794,
+      "learning_rate": 4.986386287625418e-06,
+      "loss": 0.2524,
+      "step": 18175
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.918405607392925,
+      "learning_rate": 4.9863444816053515e-06,
+      "loss": 0.2635,
+      "step": 18200
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.031548207897417,
+      "learning_rate": 4.986302675585284e-06,
+      "loss": 0.2426,
+      "step": 18225
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.352776823368371,
+      "learning_rate": 4.986260869565218e-06,
+      "loss": 0.2621,
+      "step": 18250
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.416316266136449,
+      "learning_rate": 4.9862190635451504e-06,
+      "loss": 0.2337,
+      "step": 18275
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.902376325347049,
+      "learning_rate": 4.986177257525084e-06,
+      "loss": 0.2764,
+      "step": 18300
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.577162546382801,
+      "learning_rate": 4.986135451505017e-06,
+      "loss": 0.2647,
+      "step": 18325
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.429483200140283,
+      "learning_rate": 4.98609364548495e-06,
+      "loss": 0.2539,
+      "step": 18350
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.472015436354258,
+      "learning_rate": 4.986051839464883e-06,
+      "loss": 0.2194,
+      "step": 18375
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.304795171586728,
+      "learning_rate": 4.9860100334448165e-06,
+      "loss": 0.2506,
+      "step": 18400
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.5728897025692525,
+      "learning_rate": 4.985968227424749e-06,
+      "loss": 0.2471,
+      "step": 18425
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.522313447060331,
+      "learning_rate": 4.985926421404683e-06,
+      "loss": 0.2708,
+      "step": 18450
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.393407912312494,
+      "learning_rate": 4.985884615384615e-06,
+      "loss": 0.2504,
+      "step": 18475
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.290470046509902,
+      "learning_rate": 4.985842809364549e-06,
+      "loss": 0.2418,
+      "step": 18500
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.8204122499023567,
+      "learning_rate": 4.985801003344482e-06,
+      "loss": 0.2303,
+      "step": 18525
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.705257208518913,
+      "learning_rate": 4.985759197324415e-06,
+      "loss": 0.2552,
+      "step": 18550
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.239833066925982,
+      "learning_rate": 4.985717391304348e-06,
+      "loss": 0.2349,
+      "step": 18575
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.007178185695814,
+      "learning_rate": 4.9856755852842814e-06,
+      "loss": 0.2746,
+      "step": 18600
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.604459846104763,
+      "learning_rate": 4.985633779264214e-06,
+      "loss": 0.2556,
+      "step": 18625
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.923135167645596,
+      "learning_rate": 4.985591973244148e-06,
+      "loss": 0.2602,
+      "step": 18650
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.086240722877576,
+      "learning_rate": 4.985550167224081e-06,
+      "loss": 0.2445,
+      "step": 18675
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.46646029604406,
+      "learning_rate": 4.985508361204014e-06,
+      "loss": 0.253,
+      "step": 18700
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.479558717023913,
+      "learning_rate": 4.9854665551839475e-06,
+      "loss": 0.2642,
+      "step": 18725
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.161950185397198,
+      "learning_rate": 4.98542474916388e-06,
+      "loss": 0.2664,
+      "step": 18750
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.149826660631525,
+      "learning_rate": 4.985382943143814e-06,
+      "loss": 0.2735,
+      "step": 18775
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.098059065958003,
+      "learning_rate": 4.985341137123746e-06,
+      "loss": 0.2626,
+      "step": 18800
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.7322272581947376,
+      "learning_rate": 4.98529933110368e-06,
+      "loss": 0.2706,
+      "step": 18825
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.485082925954047,
+      "learning_rate": 4.985257525083613e-06,
+      "loss": 0.265,
+      "step": 18850
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8109358920665106,
+      "learning_rate": 4.985215719063545e-06,
+      "loss": 0.2364,
+      "step": 18875
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.94763670161086,
+      "learning_rate": 4.985173913043478e-06,
+      "loss": 0.238,
+      "step": 18900
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.52671801682198,
+      "learning_rate": 4.985132107023412e-06,
+      "loss": 0.2535,
+      "step": 18925
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.79617606507355,
+      "learning_rate": 4.985090301003344e-06,
+      "loss": 0.2319,
+      "step": 18950
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.661160862165919,
+      "learning_rate": 4.985048494983278e-06,
+      "loss": 0.25,
+      "step": 18975
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.823787487368952,
+      "learning_rate": 4.9850066889632105e-06,
+      "loss": 0.2494,
+      "step": 19000
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.041962533822711,
+      "learning_rate": 4.984964882943144e-06,
+      "loss": 0.2534,
+      "step": 19025
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.137209039165049,
+      "learning_rate": 4.984923076923077e-06,
+      "loss": 0.2485,
+      "step": 19050
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.157303270364183,
+      "learning_rate": 4.98488127090301e-06,
+      "loss": 0.2495,
+      "step": 19075
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.804140711641237,
+      "learning_rate": 4.984839464882944e-06,
+      "loss": 0.2547,
+      "step": 19100
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.783724624022202,
+      "learning_rate": 4.9847976588628766e-06,
+      "loss": 0.2346,
+      "step": 19125
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.69877397127529,
+      "learning_rate": 4.98475585284281e-06,
+      "loss": 0.2892,
+      "step": 19150
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.811523610920984,
+      "learning_rate": 4.984714046822743e-06,
+      "loss": 0.2576,
+      "step": 19175
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.632437319589204,
+      "learning_rate": 4.984672240802676e-06,
+      "loss": 0.2688,
+      "step": 19200
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.287424341539971,
+      "learning_rate": 4.984630434782609e-06,
+      "loss": 0.2582,
+      "step": 19225
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.27438775796841,
+      "learning_rate": 4.984588628762543e-06,
+      "loss": 0.2517,
+      "step": 19250
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.312578396357088,
+      "learning_rate": 4.984546822742475e-06,
+      "loss": 0.2659,
+      "step": 19275
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.561038478654943,
+      "learning_rate": 4.984505016722409e-06,
+      "loss": 0.2669,
+      "step": 19300
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.883831211163112,
+      "learning_rate": 4.9844632107023416e-06,
+      "loss": 0.2644,
+      "step": 19325
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.7206392552751435,
+      "learning_rate": 4.984421404682275e-06,
+      "loss": 0.2419,
+      "step": 19350
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.772561430178686,
+      "learning_rate": 4.98438127090301e-06,
+      "loss": 0.2545,
+      "step": 19375
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.808370166675897,
+      "learning_rate": 4.984339464882944e-06,
+      "loss": 0.2463,
+      "step": 19400
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.8556307395440825,
+      "learning_rate": 4.984297658862876e-06,
+      "loss": 0.2811,
+      "step": 19425
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.305858961213936,
+      "learning_rate": 4.98425585284281e-06,
+      "loss": 0.2662,
+      "step": 19450
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.59470367618029,
+      "learning_rate": 4.9842140468227426e-06,
+      "loss": 0.2878,
+      "step": 19475
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.3608941320554395,
+      "learning_rate": 4.984172240802676e-06,
+      "loss": 0.2566,
+      "step": 19500
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.757478418973163,
+      "learning_rate": 4.984130434782609e-06,
+      "loss": 0.2448,
+      "step": 19525
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.286220221553772,
+      "learning_rate": 4.984088628762542e-06,
+      "loss": 0.2359,
+      "step": 19550
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.690642809026132,
+      "learning_rate": 4.984046822742475e-06,
+      "loss": 0.2417,
+      "step": 19575
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.924557118890291,
+      "learning_rate": 4.984005016722409e-06,
+      "loss": 0.2663,
+      "step": 19600
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.77094033240899,
+      "learning_rate": 4.983963210702341e-06,
+      "loss": 0.2548,
+      "step": 19625
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.358984573786143,
+      "learning_rate": 4.983921404682275e-06,
+      "loss": 0.2619,
+      "step": 19650
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.7311555484725965,
+      "learning_rate": 4.9838795986622075e-06,
+      "loss": 0.2821,
+      "step": 19675
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.141037303387008,
+      "learning_rate": 4.983837792642141e-06,
+      "loss": 0.2697,
+      "step": 19700
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.218327752735989,
+      "learning_rate": 4.983795986622074e-06,
+      "loss": 0.2596,
+      "step": 19725
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.478272133179351,
+      "learning_rate": 4.983754180602007e-06,
+      "loss": 0.2788,
+      "step": 19750
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.3508943589897315,
+      "learning_rate": 4.983712374581941e-06,
+      "loss": 0.2787,
+      "step": 19775
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.109200564645706,
+      "learning_rate": 4.983670568561874e-06,
+      "loss": 0.226,
+      "step": 19800
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.18181645808782,
+      "learning_rate": 4.983628762541806e-06,
+      "loss": 0.2721,
+      "step": 19825
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.996776147055529,
+      "learning_rate": 4.983586956521739e-06,
+      "loss": 0.2465,
+      "step": 19850
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.761328825995187,
+      "learning_rate": 4.9835451505016725e-06,
+      "loss": 0.2453,
+      "step": 19875
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.828048723654273,
+      "learning_rate": 4.983503344481605e-06,
+      "loss": 0.2522,
+      "step": 19900
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.580918856632557,
+      "learning_rate": 4.983461538461539e-06,
+      "loss": 0.2744,
+      "step": 19925
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.91882009627752,
+      "learning_rate": 4.9834197324414715e-06,
+      "loss": 0.2666,
+      "step": 19950
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.32076952903993,
+      "learning_rate": 4.983377926421405e-06,
+      "loss": 0.265,
+      "step": 19975
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.971506292016713,
+      "learning_rate": 4.983336120401338e-06,
+      "loss": 0.2646,
+      "step": 20000
+    },
+    {
+      "epoch": 0.79,
+      "eval_loss": 0.4140625,
+      "eval_runtime": 11517.8974,
+      "eval_samples_per_second": 0.822,
+      "eval_steps_per_second": 0.051,
+      "eval_wer": 0.1283560960700095,
+      "step": 20000
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.597765424127864,
+      "learning_rate": 4.983294314381271e-06,
+      "loss": 0.2477,
+      "step": 20025
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.125850278326169,
+      "learning_rate": 4.983252508361204e-06,
+      "loss": 0.2679,
+      "step": 20050
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.937530192709086,
+      "learning_rate": 4.9832107023411375e-06,
+      "loss": 0.2524,
+      "step": 20075
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.879107953793692,
+      "learning_rate": 4.98316889632107e-06,
+      "loss": 0.2629,
+      "step": 20100
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.726138480598868,
+      "learning_rate": 4.983127090301004e-06,
+      "loss": 0.2541,
+      "step": 20125
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.90017454045631,
+      "learning_rate": 4.9830852842809364e-06,
+      "loss": 0.263,
+      "step": 20150
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.581872676278387,
+      "learning_rate": 4.98304347826087e-06,
+      "loss": 0.2568,
+      "step": 20175
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.54004763181376,
+      "learning_rate": 4.9830016722408035e-06,
+      "loss": 0.237,
+      "step": 20200
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.106351027498988,
+      "learning_rate": 4.982959866220736e-06,
+      "loss": 0.2521,
+      "step": 20225
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.218144145112456,
+      "learning_rate": 4.98291806020067e-06,
+      "loss": 0.2719,
+      "step": 20250
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.68621541939387,
+      "learning_rate": 4.9828762541806025e-06,
+      "loss": 0.2378,
+      "step": 20275
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.990952354340057,
+      "learning_rate": 4.982834448160536e-06,
+      "loss": 0.2557,
+      "step": 20300
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.0326266609134045,
+      "learning_rate": 4.982792642140469e-06,
+      "loss": 0.2745,
+      "step": 20325
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.628421286827086,
+      "learning_rate": 4.982750836120402e-06,
+      "loss": 0.2832,
+      "step": 20350
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.890967827336047,
+      "learning_rate": 4.982710702341137e-06,
+      "loss": 0.2428,
+      "step": 20375
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.757261412656723,
+      "learning_rate": 4.982668896321071e-06,
+      "loss": 0.2363,
+      "step": 20400
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.790150924630483,
+      "learning_rate": 4.9826270903010035e-06,
+      "loss": 0.2841,
+      "step": 20425
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.334436845925394,
+      "learning_rate": 4.982585284280937e-06,
+      "loss": 0.2335,
+      "step": 20450
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.496338521482485,
+      "learning_rate": 4.98254347826087e-06,
+      "loss": 0.256,
+      "step": 20475
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.1183227106736355,
+      "learning_rate": 4.982501672240803e-06,
+      "loss": 0.2316,
+      "step": 20500
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.6851818591062955,
+      "learning_rate": 4.982459866220736e-06,
+      "loss": 0.2858,
+      "step": 20525
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.243803882140677,
+      "learning_rate": 4.9824180602006695e-06,
+      "loss": 0.2509,
+      "step": 20550
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.557348957276566,
+      "learning_rate": 4.982376254180602e-06,
+      "loss": 0.255,
+      "step": 20575
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.057972306154081,
+      "learning_rate": 4.982334448160536e-06,
+      "loss": 0.2673,
+      "step": 20600
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.269049715317146,
+      "learning_rate": 4.9822926421404685e-06,
+      "loss": 0.2332,
+      "step": 20625
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.0641172747571215,
+      "learning_rate": 4.982250836120402e-06,
+      "loss": 0.2438,
+      "step": 20650
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.545638271569368,
+      "learning_rate": 4.982209030100335e-06,
+      "loss": 0.2507,
+      "step": 20675
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.833842229029364,
+      "learning_rate": 4.982167224080268e-06,
+      "loss": 0.2692,
+      "step": 20700
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.91131799728637,
+      "learning_rate": 4.982125418060201e-06,
+      "loss": 0.2708,
+      "step": 20725
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.81240765396023,
+      "learning_rate": 4.982083612040134e-06,
+      "loss": 0.2362,
+      "step": 20750
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.808809996905685,
+      "learning_rate": 4.982041806020067e-06,
+      "loss": 0.2398,
+      "step": 20775
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.063650018965221,
+      "learning_rate": 4.982e-06,
+      "loss": 0.2631,
+      "step": 20800
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.759033532600636,
+      "learning_rate": 4.9819581939799334e-06,
+      "loss": 0.2227,
+      "step": 20825
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.455666288438305,
+      "learning_rate": 4.981916387959866e-06,
+      "loss": 0.2445,
+      "step": 20850
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.701923172622271,
+      "learning_rate": 4.9818745819398e-06,
+      "loss": 0.2575,
+      "step": 20875
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.9618139565457255,
+      "learning_rate": 4.981832775919732e-06,
+      "loss": 0.2406,
+      "step": 20900
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.68671913016998,
+      "learning_rate": 4.981790969899666e-06,
+      "loss": 0.2665,
+      "step": 20925
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.722574500390105,
+      "learning_rate": 4.981749163879599e-06,
+      "loss": 0.2569,
+      "step": 20950
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.572541737429397,
+      "learning_rate": 4.981707357859532e-06,
+      "loss": 0.239,
+      "step": 20975
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.415209470257432,
+      "learning_rate": 4.981665551839465e-06,
+      "loss": 0.2451,
+      "step": 21000
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.020681059671357,
+      "learning_rate": 4.9816237458193984e-06,
+      "loss": 0.2469,
+      "step": 21025
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.046597623490536,
+      "learning_rate": 4.981581939799331e-06,
+      "loss": 0.2429,
+      "step": 21050
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.68155367027518,
+      "learning_rate": 4.981540133779265e-06,
+      "loss": 0.2215,
+      "step": 21075
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.289254570068965,
+      "learning_rate": 4.981498327759197e-06,
+      "loss": 0.2596,
+      "step": 21100
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.531907347250682,
+      "learning_rate": 4.981456521739131e-06,
+      "loss": 0.2724,
+      "step": 21125
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.70719685809795,
+      "learning_rate": 4.981414715719064e-06,
+      "loss": 0.246,
+      "step": 21150
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.033777182035272,
+      "learning_rate": 4.981372909698997e-06,
+      "loss": 0.2603,
+      "step": 21175
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.997263007017261,
+      "learning_rate": 4.98133110367893e-06,
+      "loss": 0.2549,
+      "step": 21200
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.397447058530203,
+      "learning_rate": 4.981289297658863e-06,
+      "loss": 0.2252,
+      "step": 21225
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.387609355487834,
+      "learning_rate": 4.981247491638796e-06,
+      "loss": 0.2424,
+      "step": 21250
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.86389992482405,
+      "learning_rate": 4.981207357859532e-06,
+      "loss": 0.2808,
+      "step": 21275
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.886969996279023,
+      "learning_rate": 4.9811655518394655e-06,
+      "loss": 0.2445,
+      "step": 21300
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.930646590051299,
+      "learning_rate": 4.981123745819398e-06,
+      "loss": 0.2385,
+      "step": 21325
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.658704856684161,
+      "learning_rate": 4.981081939799332e-06,
+      "loss": 0.2654,
+      "step": 21350
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.85826584313124,
+      "learning_rate": 4.981040133779264e-06,
+      "loss": 0.2787,
+      "step": 21375
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.2136211167108915,
+      "learning_rate": 4.980998327759198e-06,
+      "loss": 0.2496,
+      "step": 21400
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.292493595501454,
+      "learning_rate": 4.980956521739131e-06,
+      "loss": 0.2398,
+      "step": 21425
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.058955089554257,
+      "learning_rate": 4.980914715719064e-06,
+      "loss": 0.2399,
+      "step": 21450
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.676342839883485,
+      "learning_rate": 4.980872909698997e-06,
+      "loss": 0.2323,
+      "step": 21475
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.982168817113219,
+      "learning_rate": 4.9808311036789305e-06,
+      "loss": 0.2391,
+      "step": 21500
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.193462179348854,
+      "learning_rate": 4.980789297658863e-06,
+      "loss": 0.2386,
+      "step": 21525
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.014739296688169,
+      "learning_rate": 4.980747491638797e-06,
+      "loss": 0.2514,
+      "step": 21550
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.930490666345008,
+      "learning_rate": 4.980705685618729e-06,
+      "loss": 0.2643,
+      "step": 21575
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.295124932611308,
+      "learning_rate": 4.980663879598663e-06,
+      "loss": 0.2689,
+      "step": 21600
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.907807211766026,
+      "learning_rate": 4.980622073578596e-06,
+      "loss": 0.2589,
+      "step": 21625
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.178379073295326,
+      "learning_rate": 4.980580267558529e-06,
+      "loss": 0.2515,
+      "step": 21650
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.000008304241975,
+      "learning_rate": 4.980538461538462e-06,
+      "loss": 0.2544,
+      "step": 21675
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.7552188199845675,
+      "learning_rate": 4.980496655518395e-06,
+      "loss": 0.2474,
+      "step": 21700
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.291094061313578,
+      "learning_rate": 4.980454849498328e-06,
+      "loss": 0.2488,
+      "step": 21725
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.7158702845231,
+      "learning_rate": 4.980413043478261e-06,
+      "loss": 0.2377,
+      "step": 21750
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.371680856314009,
+      "learning_rate": 4.980371237458194e-06,
+      "loss": 0.2695,
+      "step": 21775
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.362994192905961,
+      "learning_rate": 4.980329431438127e-06,
+      "loss": 0.2622,
+      "step": 21800
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.382584959830635,
+      "learning_rate": 4.980287625418061e-06,
+      "loss": 0.2458,
+      "step": 21825
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.793487011383763,
+      "learning_rate": 4.980245819397993e-06,
+      "loss": 0.288,
+      "step": 21850
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.399312922304265,
+      "learning_rate": 4.980204013377927e-06,
+      "loss": 0.2615,
+      "step": 21875
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.081833307673779,
+      "learning_rate": 4.9801622073578596e-06,
+      "loss": 0.2694,
+      "step": 21900
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.915184300784432,
+      "learning_rate": 4.980120401337793e-06,
+      "loss": 0.2545,
+      "step": 21925
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.794109389200768,
+      "learning_rate": 4.980078595317726e-06,
+      "loss": 0.2422,
+      "step": 21950
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.638448141225639,
+      "learning_rate": 4.980036789297659e-06,
+      "loss": 0.2736,
+      "step": 21975
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.2006435060994844,
+      "learning_rate": 4.979994983277592e-06,
+      "loss": 0.2537,
+      "step": 22000
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.600066690029428,
+      "learning_rate": 4.979953177257526e-06,
+      "loss": 0.2654,
+      "step": 22025
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.687952294883638,
+      "learning_rate": 4.979911371237458e-06,
+      "loss": 0.2593,
+      "step": 22050
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.627401660974489,
+      "learning_rate": 4.979869565217392e-06,
+      "loss": 0.2711,
+      "step": 22075
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.065685866336755,
+      "learning_rate": 4.9798277591973245e-06,
+      "loss": 0.2387,
+      "step": 22100
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.682468578553179,
+      "learning_rate": 4.979785953177258e-06,
+      "loss": 0.2351,
+      "step": 22125
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.893827612677344,
+      "learning_rate": 4.979744147157191e-06,
+      "loss": 0.2501,
+      "step": 22150
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.300974420339823,
+      "learning_rate": 4.979702341137124e-06,
+      "loss": 0.2292,
+      "step": 22175
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.750330015989807,
+      "learning_rate": 4.979660535117057e-06,
+      "loss": 0.2534,
+      "step": 22200
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.921714939272224,
+      "learning_rate": 4.9796187290969906e-06,
+      "loss": 0.2378,
+      "step": 22225
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.104235634143888,
+      "learning_rate": 4.979576923076923e-06,
+      "loss": 0.239,
+      "step": 22250
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.365003485750156,
+      "learning_rate": 4.979535117056857e-06,
+      "loss": 0.2401,
+      "step": 22275
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.8943944369540064,
+      "learning_rate": 4.9794933110367895e-06,
+      "loss": 0.2703,
+      "step": 22300
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.740334114124547,
+      "learning_rate": 4.979451505016723e-06,
+      "loss": 0.2448,
+      "step": 22325
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.26645969958187,
+      "learning_rate": 4.979409698996656e-06,
+      "loss": 0.237,
+      "step": 22350
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.188296175194914,
+      "learning_rate": 4.979367892976589e-06,
+      "loss": 0.2439,
+      "step": 22375
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.28258281943557,
+      "learning_rate": 4.979326086956522e-06,
+      "loss": 0.2319,
+      "step": 22400
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.876641118702411,
+      "learning_rate": 4.979284280936455e-06,
+      "loss": 0.2256,
+      "step": 22425
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.9872075168697,
+      "learning_rate": 4.979242474916388e-06,
+      "loss": 0.2401,
+      "step": 22450
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.709817054749593,
+      "learning_rate": 4.979200668896321e-06,
+      "loss": 0.2318,
+      "step": 22475
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.821509746602487,
+      "learning_rate": 4.9791588628762545e-06,
+      "loss": 0.2556,
+      "step": 22500
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.68934415618322,
+      "learning_rate": 4.979117056856187e-06,
+      "loss": 0.2558,
+      "step": 22525
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.509947218331295,
+      "learning_rate": 4.979075250836121e-06,
+      "loss": 0.2365,
+      "step": 22550
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.58010564782228,
+      "learning_rate": 4.9790334448160534e-06,
+      "loss": 0.2585,
+      "step": 22575
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.350109880497084,
+      "learning_rate": 4.978991638795987e-06,
+      "loss": 0.2496,
+      "step": 22600
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.52596483622084,
+      "learning_rate": 4.97894983277592e-06,
+      "loss": 0.2513,
+      "step": 22625
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.5425308104361335,
+      "learning_rate": 4.978908026755853e-06,
+      "loss": 0.2317,
+      "step": 22650
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.042990505291717,
+      "learning_rate": 4.978866220735786e-06,
+      "loss": 0.2425,
+      "step": 22675
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.28155864508552,
+      "learning_rate": 4.9788244147157195e-06,
+      "loss": 0.2433,
+      "step": 22700
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.05219957106238,
+      "learning_rate": 4.978782608695652e-06,
+      "loss": 0.2582,
+      "step": 22725
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.3750393384628925,
+      "learning_rate": 4.978740802675586e-06,
+      "loss": 0.24,
+      "step": 22750
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.981026672978073,
+      "learning_rate": 4.978698996655518e-06,
+      "loss": 0.2564,
+      "step": 22775
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.8161776421109055,
+      "learning_rate": 4.978657190635452e-06,
+      "loss": 0.239,
+      "step": 22800
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.376213905893997,
+      "learning_rate": 4.978615384615385e-06,
+      "loss": 0.2611,
+      "step": 22825
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.15776727013437,
+      "learning_rate": 4.978573578595318e-06,
+      "loss": 0.2576,
+      "step": 22850
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.04185750311223,
+      "learning_rate": 4.978531772575252e-06,
+      "loss": 0.2482,
+      "step": 22875
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.950731538074712,
+      "learning_rate": 4.9784899665551844e-06,
+      "loss": 0.269,
+      "step": 22900
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.205593383648058,
+      "learning_rate": 4.978448160535118e-06,
+      "loss": 0.2401,
+      "step": 22925
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.548735869447201,
+      "learning_rate": 4.978406354515051e-06,
+      "loss": 0.2462,
+      "step": 22950
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.329571742913044,
+      "learning_rate": 4.978364548494984e-06,
+      "loss": 0.2636,
+      "step": 22975
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.536247610819822,
+      "learning_rate": 4.978322742474917e-06,
+      "loss": 0.2512,
+      "step": 23000
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.573520221822453,
+      "learning_rate": 4.9782809364548505e-06,
+      "loss": 0.2551,
+      "step": 23025
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.059324762232372,
+      "learning_rate": 4.978239130434783e-06,
+      "loss": 0.2677,
+      "step": 23050
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.4312911338529535,
+      "learning_rate": 4.978197324414717e-06,
+      "loss": 0.2647,
+      "step": 23075
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.905908003456633,
+      "learning_rate": 4.978155518394649e-06,
+      "loss": 0.241,
+      "step": 23100
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.305362513811501,
+      "learning_rate": 4.978113712374582e-06,
+      "loss": 0.2759,
+      "step": 23125
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.334507557305194,
+      "learning_rate": 4.978071906354515e-06,
+      "loss": 0.2471,
+      "step": 23150
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.167193712946236,
+      "learning_rate": 4.978030100334448e-06,
+      "loss": 0.2454,
+      "step": 23175
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.575127667668046,
+      "learning_rate": 4.977988294314381e-06,
+      "loss": 0.2166,
+      "step": 23200
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.829621265497541,
+      "learning_rate": 4.977948160535118e-06,
+      "loss": 0.2363,
+      "step": 23225
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.561630435288059,
+      "learning_rate": 4.9779063545150504e-06,
+      "loss": 0.2368,
+      "step": 23250
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.602771537076456,
+      "learning_rate": 4.977864548494984e-06,
+      "loss": 0.264,
+      "step": 23275
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.8921357782406565,
+      "learning_rate": 4.977822742474917e-06,
+      "loss": 0.2425,
+      "step": 23300
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.761815993674995,
+      "learning_rate": 4.97778093645485e-06,
+      "loss": 0.2628,
+      "step": 23325
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.789432658329115,
+      "learning_rate": 4.977739130434783e-06,
+      "loss": 0.2415,
+      "step": 23350
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.212070910404648,
+      "learning_rate": 4.977697324414716e-06,
+      "loss": 0.2606,
+      "step": 23375
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.645169198733312,
+      "learning_rate": 4.977655518394649e-06,
+      "loss": 0.2295,
+      "step": 23400
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.941660028505237,
+      "learning_rate": 4.977613712374582e-06,
+      "loss": 0.259,
+      "step": 23425
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.006340050777933,
+      "learning_rate": 4.977571906354515e-06,
+      "loss": 0.2489,
+      "step": 23450
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.444718325419201,
+      "learning_rate": 4.977530100334448e-06,
+      "loss": 0.2252,
+      "step": 23475
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.315731137511242,
+      "learning_rate": 4.977488294314382e-06,
+      "loss": 0.2444,
+      "step": 23500
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.147536052111137,
+      "learning_rate": 4.977446488294314e-06,
+      "loss": 0.253,
+      "step": 23525
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.637894947771309,
+      "learning_rate": 4.977404682274248e-06,
+      "loss": 0.2411,
+      "step": 23550
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.3547145173015,
+      "learning_rate": 4.977362876254181e-06,
+      "loss": 0.249,
+      "step": 23575
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.191389830163696,
+      "learning_rate": 4.977321070234114e-06,
+      "loss": 0.2977,
+      "step": 23600
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.747541667430566,
+      "learning_rate": 4.977279264214047e-06,
+      "loss": 0.2449,
+      "step": 23625
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.520805102077198,
+      "learning_rate": 4.97723745819398e-06,
+      "loss": 0.2308,
+      "step": 23650
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.417279655484847,
+      "learning_rate": 4.977195652173913e-06,
+      "loss": 0.2282,
+      "step": 23675
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.814135551224237,
+      "learning_rate": 4.977153846153847e-06,
+      "loss": 0.2407,
+      "step": 23700
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.799363124129896,
+      "learning_rate": 4.977112040133779e-06,
+      "loss": 0.2464,
+      "step": 23725
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.926212906223406,
+      "learning_rate": 4.977070234113713e-06,
+      "loss": 0.2543,
+      "step": 23750
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.757663142452678,
+      "learning_rate": 4.977028428093646e-06,
+      "loss": 0.2452,
+      "step": 23775
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.6489188278172895,
+      "learning_rate": 4.976986622073579e-06,
+      "loss": 0.2517,
+      "step": 23800
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.577979099426105,
+      "learning_rate": 4.976944816053512e-06,
+      "loss": 0.2605,
+      "step": 23825
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.9466236367411005,
+      "learning_rate": 4.976903010033445e-06,
+      "loss": 0.2655,
+      "step": 23850
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.7414326777762685,
+      "learning_rate": 4.976861204013378e-06,
+      "loss": 0.2304,
+      "step": 23875
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.522499469061028,
+      "learning_rate": 4.976819397993312e-06,
+      "loss": 0.2454,
+      "step": 23900
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.343226074167374,
+      "learning_rate": 4.976777591973244e-06,
+      "loss": 0.2555,
+      "step": 23925
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.908837038372152,
+      "learning_rate": 4.976735785953178e-06,
+      "loss": 0.2479,
+      "step": 23950
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.9308077824937016,
+      "learning_rate": 4.976693979933111e-06,
+      "loss": 0.2571,
+      "step": 23975
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.187889109702333,
+      "learning_rate": 4.976652173913044e-06,
+      "loss": 0.2705,
+      "step": 24000
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.591103804137682,
+      "learning_rate": 4.976610367892978e-06,
+      "loss": 0.2448,
+      "step": 24025
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.396768803047713,
+      "learning_rate": 4.97656856187291e-06,
+      "loss": 0.2243,
+      "step": 24050
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.7909268110643275,
+      "learning_rate": 4.976526755852843e-06,
+      "loss": 0.2332,
+      "step": 24075
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.7576206413020232,
+      "learning_rate": 4.976484949832776e-06,
+      "loss": 0.2549,
+      "step": 24100
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.4611210915597415,
+      "learning_rate": 4.976443143812709e-06,
+      "loss": 0.2458,
+      "step": 24125
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.089932740094683,
+      "learning_rate": 4.976401337792642e-06,
+      "loss": 0.2503,
+      "step": 24150
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.268253661077563,
+      "learning_rate": 4.9763595317725755e-06,
+      "loss": 0.225,
+      "step": 24175
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.204555335361072,
+      "learning_rate": 4.976317725752508e-06,
+      "loss": 0.2713,
+      "step": 24200
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.021816391599869,
+      "learning_rate": 4.976275919732442e-06,
+      "loss": 0.2336,
+      "step": 24225
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.8455887000789835,
+      "learning_rate": 4.9762341137123745e-06,
+      "loss": 0.2279,
+      "step": 24250
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.225031289810486,
+      "learning_rate": 4.976192307692308e-06,
+      "loss": 0.2275,
+      "step": 24275
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.159430367632941,
+      "learning_rate": 4.976150501672241e-06,
+      "loss": 0.2492,
+      "step": 24300
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.127979739902747,
+      "learning_rate": 4.976108695652174e-06,
+      "loss": 0.2336,
+      "step": 24325
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.238071839251816,
+      "learning_rate": 4.976066889632107e-06,
+      "loss": 0.2801,
+      "step": 24350
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.157082993930431,
+      "learning_rate": 4.9760250836120405e-06,
+      "loss": 0.2307,
+      "step": 24375
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.751821236126511,
+      "learning_rate": 4.975983277591974e-06,
+      "loss": 0.2466,
+      "step": 24400
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.723512280716872,
+      "learning_rate": 4.975941471571907e-06,
+      "loss": 0.2533,
+      "step": 24425
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.529999483690673,
+      "learning_rate": 4.97589966555184e-06,
+      "loss": 0.2305,
+      "step": 24450
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.640020922858098,
+      "learning_rate": 4.975857859531773e-06,
+      "loss": 0.24,
+      "step": 24475
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.647549375619507,
+      "learning_rate": 4.9758160535117065e-06,
+      "loss": 0.2444,
+      "step": 24500
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.391480590113418,
+      "learning_rate": 4.975774247491639e-06,
+      "loss": 0.2585,
+      "step": 24525
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.489122500542812,
+      "learning_rate": 4.975732441471573e-06,
+      "loss": 0.2475,
+      "step": 24550
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.499688320980486,
+      "learning_rate": 4.9756906354515055e-06,
+      "loss": 0.2402,
+      "step": 24575
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.331955267278932,
+      "learning_rate": 4.975648829431439e-06,
+      "loss": 0.2444,
+      "step": 24600
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.408758125870933,
+      "learning_rate": 4.975607023411372e-06,
+      "loss": 0.227,
+      "step": 24625
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.756801869357502,
+      "learning_rate": 4.975565217391305e-06,
+      "loss": 0.2472,
+      "step": 24650
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.472101388539881,
+      "learning_rate": 4.975523411371238e-06,
+      "loss": 0.2178,
+      "step": 24675
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.244728796693067,
+      "learning_rate": 4.9754816053511715e-06,
+      "loss": 0.2409,
+      "step": 24700
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.422212395250679,
+      "learning_rate": 4.975439799331104e-06,
+      "loss": 0.2553,
+      "step": 24725
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.733100754021741,
+      "learning_rate": 4.975397993311038e-06,
+      "loss": 0.2423,
+      "step": 24750
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.907380286889155,
+      "learning_rate": 4.97535618729097e-06,
+      "loss": 0.246,
+      "step": 24775
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.336386993051637,
+      "learning_rate": 4.975314381270903e-06,
+      "loss": 0.2437,
+      "step": 24800
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.3744020887867645,
+      "learning_rate": 4.975272575250837e-06,
+      "loss": 0.2537,
+      "step": 24825
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.130606979870505,
+      "learning_rate": 4.975230769230769e-06,
+      "loss": 0.2185,
+      "step": 24850
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.393159176636358,
+      "learning_rate": 4.975188963210703e-06,
+      "loss": 0.2447,
+      "step": 24875
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.37010512506881,
+      "learning_rate": 4.975147157190636e-06,
+      "loss": 0.2537,
+      "step": 24900
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.781150540673855,
+      "learning_rate": 4.975105351170569e-06,
+      "loss": 0.2322,
+      "step": 24925
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.670847613893096,
+      "learning_rate": 4.975063545150502e-06,
+      "loss": 0.2464,
+      "step": 24950
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.727990256628477,
+      "learning_rate": 4.9750217391304354e-06,
+      "loss": 0.2581,
+      "step": 24975
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.544263592893811,
+      "learning_rate": 4.974979933110368e-06,
+      "loss": 0.2628,
+      "step": 25000
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.69470940676318,
+      "learning_rate": 4.974938127090302e-06,
+      "loss": 0.2223,
+      "step": 25025
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.792583847102043,
+      "learning_rate": 4.974896321070234e-06,
+      "loss": 0.2343,
+      "step": 25050
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.870490355918843,
+      "learning_rate": 4.974854515050168e-06,
+      "loss": 0.2473,
+      "step": 25075
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.202473423301913,
+      "learning_rate": 4.974812709030101e-06,
+      "loss": 0.2507,
+      "step": 25100
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.7239036684055895,
+      "learning_rate": 4.974770903010034e-06,
+      "loss": 0.2539,
+      "step": 25125
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.222922132018913,
+      "learning_rate": 4.974729096989967e-06,
+      "loss": 0.2239,
+      "step": 25150
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.957670491347016,
+      "learning_rate": 4.9746872909699e-06,
+      "loss": 0.2482,
+      "step": 25175
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.8279149756498105,
+      "learning_rate": 4.974645484949833e-06,
+      "loss": 0.241,
+      "step": 25200
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.200294070704066,
+      "learning_rate": 4.974605351170569e-06,
+      "loss": 0.242,
+      "step": 25225
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.670126129827119,
+      "learning_rate": 4.974563545150502e-06,
+      "loss": 0.2406,
+      "step": 25250
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.544898567700344,
+      "learning_rate": 4.974521739130435e-06,
+      "loss": 0.2338,
+      "step": 25275
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.806206312623487,
+      "learning_rate": 4.974479933110368e-06,
+      "loss": 0.2265,
+      "step": 25300
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.53793647458314,
+      "learning_rate": 4.9744381270903014e-06,
+      "loss": 0.2345,
+      "step": 25325
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.608502420190892,
+      "learning_rate": 4.974396321070234e-06,
+      "loss": 0.2385,
+      "step": 25350
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.1286787349789815,
+      "learning_rate": 4.974354515050168e-06,
+      "loss": 0.2453,
+      "step": 25375
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.469570857307241,
+      "learning_rate": 4.9743127090301e-06,
+      "loss": 0.2294,
+      "step": 25400
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3534162143803563,
+      "learning_rate": 4.974270903010034e-06,
+      "loss": 0.2361,
+      "step": 25425
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.706718780670229,
+      "learning_rate": 4.974229096989967e-06,
+      "loss": 0.2079,
+      "step": 25450
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4059544274203903,
+      "learning_rate": 4.9741872909699e-06,
+      "loss": 0.2031,
+      "step": 25475
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7037901786451766,
+      "learning_rate": 4.974145484949834e-06,
+      "loss": 0.2252,
+      "step": 25500
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3558198308855107,
+      "learning_rate": 4.974103678929766e-06,
+      "loss": 0.2154,
+      "step": 25525
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.6682224121680143,
+      "learning_rate": 4.9740618729097e-06,
+      "loss": 0.2129,
+      "step": 25550
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.4577637750260048,
+      "learning_rate": 4.974020066889633e-06,
+      "loss": 0.2024,
+      "step": 25575
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.5087352378433831,
+      "learning_rate": 4.973978260869566e-06,
+      "loss": 0.2109,
+      "step": 25600
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.6402724021534647,
+      "learning_rate": 4.973936454849499e-06,
+      "loss": 0.212,
+      "step": 25625
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9264263538960484,
+      "learning_rate": 4.9738946488294324e-06,
+      "loss": 0.2143,
+      "step": 25650
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9174315090045257,
+      "learning_rate": 4.973852842809365e-06,
+      "loss": 0.2117,
+      "step": 25675
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8429762901411832,
+      "learning_rate": 4.973811036789299e-06,
+      "loss": 0.2059,
+      "step": 25700
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.5033416778790167,
+      "learning_rate": 4.9737692307692305e-06,
+      "loss": 0.182,
+      "step": 25725
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.330321894521951,
+      "learning_rate": 4.973727424749164e-06,
+      "loss": 0.1881,
+      "step": 25750
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.8292089071205324,
+      "learning_rate": 4.973685618729097e-06,
+      "loss": 0.1975,
+      "step": 25775
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.2598690654621236,
+      "learning_rate": 4.97364381270903e-06,
+      "loss": 0.2053,
+      "step": 25800
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.435229607084148,
+      "learning_rate": 4.973602006688963e-06,
+      "loss": 0.2169,
+      "step": 25825
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.1185880545936135,
+      "learning_rate": 4.973560200668897e-06,
+      "loss": 0.1958,
+      "step": 25850
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.3672188771049028,
+      "learning_rate": 4.973518394648829e-06,
+      "loss": 0.2104,
+      "step": 25875
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.863357164049792,
+      "learning_rate": 4.973476588628763e-06,
+      "loss": 0.1857,
+      "step": 25900
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9990414028836165,
+      "learning_rate": 4.973434782608696e-06,
+      "loss": 0.1921,
+      "step": 25925
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.3154262807639836,
+      "learning_rate": 4.973392976588629e-06,
+      "loss": 0.2266,
+      "step": 25950
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.3950612420595643,
+      "learning_rate": 4.973351170568563e-06,
+      "loss": 0.2064,
+      "step": 25975
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.408288078068823,
+      "learning_rate": 4.973309364548495e-06,
+      "loss": 0.2071,
+      "step": 26000
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.9525039250801453,
+      "learning_rate": 4.973267558528429e-06,
+      "loss": 0.2118,
+      "step": 26025
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.8278710821331015,
+      "learning_rate": 4.9732257525083616e-06,
+      "loss": 0.2262,
+      "step": 26050
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.6148367474628422,
+      "learning_rate": 4.973183946488295e-06,
+      "loss": 0.1902,
+      "step": 26075
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.5038758535892578,
+      "learning_rate": 4.973142140468228e-06,
+      "loss": 0.2051,
+      "step": 26100
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.55344960183899,
+      "learning_rate": 4.973100334448161e-06,
+      "loss": 0.2026,
+      "step": 26125
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.1894008457226986,
+      "learning_rate": 4.973058528428094e-06,
+      "loss": 0.2085,
+      "step": 26150
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.585075493380135,
+      "learning_rate": 4.973016722408028e-06,
+      "loss": 0.2036,
+      "step": 26175
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8233869368121183,
+      "learning_rate": 4.97297491638796e-06,
+      "loss": 0.2235,
+      "step": 26200
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.8237982393202128,
+      "learning_rate": 4.972934782608696e-06,
+      "loss": 0.2225,
+      "step": 26225
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.870096249638426,
+      "learning_rate": 4.972892976588629e-06,
+      "loss": 0.2277,
+      "step": 26250
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.084326368520519,
+      "learning_rate": 4.972851170568562e-06,
+      "loss": 0.2166,
+      "step": 26275
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.786309173531602,
+      "learning_rate": 4.972809364548495e-06,
+      "loss": 0.1952,
+      "step": 26300
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.118377964073037,
+      "learning_rate": 4.972767558528429e-06,
+      "loss": 0.2035,
+      "step": 26325
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.4408933539946762,
+      "learning_rate": 4.972725752508361e-06,
+      "loss": 0.2101,
+      "step": 26350
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.5247018069177714,
+      "learning_rate": 4.972683946488295e-06,
+      "loss": 0.2146,
+      "step": 26375
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.4099242258025098,
+      "learning_rate": 4.9726421404682275e-06,
+      "loss": 0.2125,
+      "step": 26400
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.385534734677316,
+      "learning_rate": 4.972600334448161e-06,
+      "loss": 0.2031,
+      "step": 26425
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.3518252038997889,
+      "learning_rate": 4.972558528428094e-06,
+      "loss": 0.2054,
+      "step": 26450
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.7923556960029041,
+      "learning_rate": 4.972516722408027e-06,
+      "loss": 0.2048,
+      "step": 26475
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.6394896847363243,
+      "learning_rate": 4.97247491638796e-06,
+      "loss": 0.2172,
+      "step": 26500
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.9050366746717975,
+      "learning_rate": 4.972433110367894e-06,
+      "loss": 0.2072,
+      "step": 26525
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.4611510590944903,
+      "learning_rate": 4.972391304347826e-06,
+      "loss": 0.2026,
+      "step": 26550
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.757163035064764,
+      "learning_rate": 4.97234949832776e-06,
+      "loss": 0.2169,
+      "step": 26575
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.2821175785153787,
+      "learning_rate": 4.9723076923076925e-06,
+      "loss": 0.2181,
+      "step": 26600
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.4184262153912761,
+      "learning_rate": 4.972265886287626e-06,
+      "loss": 0.1958,
+      "step": 26625
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.7602753726816016,
+      "learning_rate": 4.97222408026756e-06,
+      "loss": 0.2009,
+      "step": 26650
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.5974760210540015,
+      "learning_rate": 4.9721822742474915e-06,
+      "loss": 0.2243,
+      "step": 26675
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.645642258789146,
+      "learning_rate": 4.972140468227425e-06,
+      "loss": 0.2038,
+      "step": 26700
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.4401284677266035,
+      "learning_rate": 4.972098662207358e-06,
+      "loss": 0.1898,
+      "step": 26725
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.430617922584249,
+      "learning_rate": 4.972056856187291e-06,
+      "loss": 0.2111,
+      "step": 26750
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.7904074848447293,
+      "learning_rate": 4.972015050167224e-06,
+      "loss": 0.2129,
+      "step": 26775
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.406346980994871,
+      "learning_rate": 4.9719732441471575e-06,
+      "loss": 0.2045,
+      "step": 26800
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.2583526864827905,
+      "learning_rate": 4.97193143812709e-06,
+      "loss": 0.2266,
+      "step": 26825
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7655942648716196,
+      "learning_rate": 4.971889632107024e-06,
+      "loss": 0.1951,
+      "step": 26850
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.467109180429437,
+      "learning_rate": 4.9718478260869564e-06,
+      "loss": 0.2284,
+      "step": 26875
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7516265119310985,
+      "learning_rate": 4.97180602006689e-06,
+      "loss": 0.2234,
+      "step": 26900
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.5583524171970238,
+      "learning_rate": 4.971764214046823e-06,
+      "loss": 0.2205,
+      "step": 26925
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.5551307670089498,
+      "learning_rate": 4.971722408026756e-06,
+      "loss": 0.2013,
+      "step": 26950
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.7858242769151977,
+      "learning_rate": 4.971680602006689e-06,
+      "loss": 0.2037,
+      "step": 26975
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6377397935158822,
+      "learning_rate": 4.9716387959866225e-06,
+      "loss": 0.2002,
+      "step": 27000
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9240347144018508,
+      "learning_rate": 4.971596989966555e-06,
+      "loss": 0.2101,
+      "step": 27025
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9628276445796573,
+      "learning_rate": 4.971555183946489e-06,
+      "loss": 0.2132,
+      "step": 27050
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.9744874761939027,
+      "learning_rate": 4.971513377926422e-06,
+      "loss": 0.2101,
+      "step": 27075
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.6399308156848618,
+      "learning_rate": 4.971471571906355e-06,
+      "loss": 0.2181,
+      "step": 27100
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.1026445818397668,
+      "learning_rate": 4.9714297658862885e-06,
+      "loss": 0.197,
+      "step": 27125
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.2558271888800925,
+      "learning_rate": 4.971387959866221e-06,
+      "loss": 0.2128,
+      "step": 27150
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8603882142931765,
+      "learning_rate": 4.971346153846155e-06,
+      "loss": 0.1947,
+      "step": 27175
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7503812536338823,
+      "learning_rate": 4.9713043478260875e-06,
+      "loss": 0.1932,
+      "step": 27200
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.567055034127128,
+      "learning_rate": 4.971264214046823e-06,
+      "loss": 0.1959,
+      "step": 27225
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.133487976541747,
+      "learning_rate": 4.971222408026756e-06,
+      "loss": 0.2218,
+      "step": 27250
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.7226015136938018,
+      "learning_rate": 4.9711806020066895e-06,
+      "loss": 0.2246,
+      "step": 27275
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.5324982024611993,
+      "learning_rate": 4.971138795986622e-06,
+      "loss": 0.2249,
+      "step": 27300
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.2666638439290532,
+      "learning_rate": 4.971096989966556e-06,
+      "loss": 0.2023,
+      "step": 27325
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.7936378450293402,
+      "learning_rate": 4.9710551839464885e-06,
+      "loss": 0.2209,
+      "step": 27350
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.4685172845320271,
+      "learning_rate": 4.971013377926422e-06,
+      "loss": 0.212,
+      "step": 27375
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.3475774524686015,
+      "learning_rate": 4.970971571906355e-06,
+      "loss": 0.2038,
+      "step": 27400
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.1356769724282703,
+      "learning_rate": 4.970929765886288e-06,
+      "loss": 0.2052,
+      "step": 27425
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.3461405690608945,
+      "learning_rate": 4.970887959866221e-06,
+      "loss": 0.2077,
+      "step": 27450
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.2787743115282497,
+      "learning_rate": 4.9708461538461545e-06,
+      "loss": 0.1933,
+      "step": 27475
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.710480020197042,
+      "learning_rate": 4.970804347826087e-06,
+      "loss": 0.1945,
+      "step": 27500
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.74469357398459,
+      "learning_rate": 4.970762541806021e-06,
+      "loss": 0.2212,
+      "step": 27525
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.553720300641038,
+      "learning_rate": 4.9707207357859535e-06,
+      "loss": 0.2071,
+      "step": 27550
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.5925560178851168,
+      "learning_rate": 4.970678929765887e-06,
+      "loss": 0.2082,
+      "step": 27575
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.1786027786262885,
+      "learning_rate": 4.97063712374582e-06,
+      "loss": 0.2185,
+      "step": 27600
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.7290929692337984,
+      "learning_rate": 4.970595317725752e-06,
+      "loss": 0.1983,
+      "step": 27625
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9473623643856743,
+      "learning_rate": 4.970553511705686e-06,
+      "loss": 0.1961,
+      "step": 27650
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.4076384394311243,
+      "learning_rate": 4.970511705685619e-06,
+      "loss": 0.197,
+      "step": 27675
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.9993400148837797,
+      "learning_rate": 4.970469899665552e-06,
+      "loss": 0.1954,
+      "step": 27700
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.271427684128234,
+      "learning_rate": 4.970428093645485e-06,
+      "loss": 0.2048,
+      "step": 27725
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.591987460804486,
+      "learning_rate": 4.9703862876254184e-06,
+      "loss": 0.1951,
+      "step": 27750
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.5791865767097701,
+      "learning_rate": 4.970344481605351e-06,
+      "loss": 0.2057,
+      "step": 27775
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.1896274986314004,
+      "learning_rate": 4.970302675585285e-06,
+      "loss": 0.2098,
+      "step": 27800
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.6903447291808138,
+      "learning_rate": 4.970260869565217e-06,
+      "loss": 0.1987,
+      "step": 27825
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.490242793429607,
+      "learning_rate": 4.970219063545151e-06,
+      "loss": 0.1942,
+      "step": 27850
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.2304358371703357,
+      "learning_rate": 4.970177257525084e-06,
+      "loss": 0.2152,
+      "step": 27875
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.890308908845425,
+      "learning_rate": 4.970135451505017e-06,
+      "loss": 0.2066,
+      "step": 27900
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.5472152303564,
+      "learning_rate": 4.97009364548495e-06,
+      "loss": 0.2295,
+      "step": 27925
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.8402540061996924,
+      "learning_rate": 4.970051839464883e-06,
+      "loss": 0.2103,
+      "step": 27950
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.5521269697737237,
+      "learning_rate": 4.970010033444816e-06,
+      "loss": 0.1894,
+      "step": 27975
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.1789685652154698,
+      "learning_rate": 4.96996822742475e-06,
+      "loss": 0.1998,
+      "step": 28000
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.971887880644852,
+      "learning_rate": 4.969926421404682e-06,
+      "loss": 0.2217,
+      "step": 28025
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.6583518083353077,
+      "learning_rate": 4.969884615384616e-06,
+      "loss": 0.2077,
+      "step": 28050
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.2231691799156768,
+      "learning_rate": 4.969842809364549e-06,
+      "loss": 0.2074,
+      "step": 28075
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.4124682639555552,
+      "learning_rate": 4.969801003344482e-06,
+      "loss": 0.2002,
+      "step": 28100
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.5315963028247968,
+      "learning_rate": 4.969759197324415e-06,
+      "loss": 0.2116,
+      "step": 28125
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.774063539155429,
+      "learning_rate": 4.969717391304348e-06,
+      "loss": 0.2003,
+      "step": 28150
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.3548270043032278,
+      "learning_rate": 4.969675585284282e-06,
+      "loss": 0.2266,
+      "step": 28175
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.5113091523893083,
+      "learning_rate": 4.969633779264215e-06,
+      "loss": 0.2182,
+      "step": 28200
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.9251487332267831,
+      "learning_rate": 4.9695936454849505e-06,
+      "loss": 0.2202,
+      "step": 28225
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.8324301345828689,
+      "learning_rate": 4.969551839464883e-06,
+      "loss": 0.2255,
+      "step": 28250
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.4775932179452143,
+      "learning_rate": 4.969510033444817e-06,
+      "loss": 0.1994,
+      "step": 28275
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.6333919121168157,
+      "learning_rate": 4.969468227424749e-06,
+      "loss": 0.2081,
+      "step": 28300
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.864925684984239,
+      "learning_rate": 4.969426421404683e-06,
+      "loss": 0.1797,
+      "step": 28325
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9096612346707407,
+      "learning_rate": 4.969384615384616e-06,
+      "loss": 0.206,
+      "step": 28350
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9576478394690897,
+      "learning_rate": 4.969342809364549e-06,
+      "loss": 0.2197,
+      "step": 28375
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.5136863296140723,
+      "learning_rate": 4.969301003344482e-06,
+      "loss": 0.199,
+      "step": 28400
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.2010364526097577,
+      "learning_rate": 4.9692591973244154e-06,
+      "loss": 0.2107,
+      "step": 28425
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.407440767715517,
+      "learning_rate": 4.969217391304348e-06,
+      "loss": 0.1996,
+      "step": 28450
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.037763973567924,
+      "learning_rate": 4.969175585284282e-06,
+      "loss": 0.1935,
+      "step": 28475
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.7243218825000752,
+      "learning_rate": 4.969133779264214e-06,
+      "loss": 0.2144,
+      "step": 28500
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.973852579552326,
+      "learning_rate": 4.969091973244148e-06,
+      "loss": 0.1863,
+      "step": 28525
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.5798095723052332,
+      "learning_rate": 4.969050167224081e-06,
+      "loss": 0.2177,
+      "step": 28550
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.4624355487653509,
+      "learning_rate": 4.969008361204013e-06,
+      "loss": 0.2158,
+      "step": 28575
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.2634512898863046,
+      "learning_rate": 4.968966555183947e-06,
+      "loss": 0.1805,
+      "step": 28600
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.292016189175459,
+      "learning_rate": 4.9689247491638796e-06,
+      "loss": 0.1972,
+      "step": 28625
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.887658715804907,
+      "learning_rate": 4.968882943143813e-06,
+      "loss": 0.1994,
+      "step": 28650
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.5728950499609098,
+      "learning_rate": 4.968841137123746e-06,
+      "loss": 0.2084,
+      "step": 28675
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.5170270635605654,
+      "learning_rate": 4.968799331103679e-06,
+      "loss": 0.1836,
+      "step": 28700
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.182392469384069,
+      "learning_rate": 4.968757525083612e-06,
+      "loss": 0.1927,
+      "step": 28725
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.467566207323012,
+      "learning_rate": 4.968715719063546e-06,
+      "loss": 0.211,
+      "step": 28750
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.272554103873533,
+      "learning_rate": 4.968673913043478e-06,
+      "loss": 0.2185,
+      "step": 28775
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.8133824165206656,
+      "learning_rate": 4.968632107023412e-06,
+      "loss": 0.2006,
+      "step": 28800
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.1311871034158731,
+      "learning_rate": 4.9685903010033445e-06,
+      "loss": 0.2058,
+      "step": 28825
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.5883722748526226,
+      "learning_rate": 4.968548494983278e-06,
+      "loss": 0.2064,
+      "step": 28850
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.5904261682283662,
+      "learning_rate": 4.968506688963211e-06,
+      "loss": 0.2024,
+      "step": 28875
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.1478828955712104,
+      "learning_rate": 4.968464882943144e-06,
+      "loss": 0.235,
+      "step": 28900
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.630161077340849,
+      "learning_rate": 4.968423076923077e-06,
+      "loss": 0.1974,
+      "step": 28925
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8662904772136388,
+      "learning_rate": 4.9683812709030106e-06,
+      "loss": 0.2017,
+      "step": 28950
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8506149538880028,
+      "learning_rate": 4.968339464882943e-06,
+      "loss": 0.2079,
+      "step": 28975
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7198221954597992,
+      "learning_rate": 4.968297658862877e-06,
+      "loss": 0.2012,
+      "step": 29000
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.2468499981047616,
+      "learning_rate": 4.9682558528428095e-06,
+      "loss": 0.2241,
+      "step": 29025
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.871449485888433,
+      "learning_rate": 4.968214046822743e-06,
+      "loss": 0.2259,
+      "step": 29050
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.2013415537996155,
+      "learning_rate": 4.968172240802676e-06,
+      "loss": 0.1966,
+      "step": 29075
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8646445635350821,
+      "learning_rate": 4.968130434782609e-06,
+      "loss": 0.2081,
+      "step": 29100
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.7362617133719327,
+      "learning_rate": 4.968088628762542e-06,
+      "loss": 0.2033,
+      "step": 29125
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9583229334368888,
+      "learning_rate": 4.9680468227424756e-06,
+      "loss": 0.2195,
+      "step": 29150
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.9756496260469272,
+      "learning_rate": 4.968005016722408e-06,
+      "loss": 0.2138,
+      "step": 29175
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8136961850820263,
+      "learning_rate": 4.967963210702342e-06,
+      "loss": 0.1989,
+      "step": 29200
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8737358384053303,
+      "learning_rate": 4.9679214046822745e-06,
+      "loss": 0.1914,
+      "step": 29225
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8604986074744494,
+      "learning_rate": 4.96788127090301e-06,
+      "loss": 0.2203,
+      "step": 29250
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.349940979840139,
+      "learning_rate": 4.967839464882944e-06,
+      "loss": 0.2278,
+      "step": 29275
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8282393434449213,
+      "learning_rate": 4.9677976588628766e-06,
+      "loss": 0.2236,
+      "step": 29300
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.5180697118545403,
+      "learning_rate": 4.96775585284281e-06,
+      "loss": 0.2302,
+      "step": 29325
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8114203235294122,
+      "learning_rate": 4.967714046822743e-06,
+      "loss": 0.1866,
+      "step": 29350
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.3440834890531017,
+      "learning_rate": 4.967672240802676e-06,
+      "loss": 0.2201,
+      "step": 29375
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.3107235122075616,
+      "learning_rate": 4.967630434782609e-06,
+      "loss": 0.2028,
+      "step": 29400
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.329479003913603,
+      "learning_rate": 4.967588628762543e-06,
+      "loss": 0.1991,
+      "step": 29425
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.7727406488523492,
+      "learning_rate": 4.967546822742475e-06,
+      "loss": 0.2009,
+      "step": 29450
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.6058323726648749,
+      "learning_rate": 4.967505016722409e-06,
+      "loss": 0.2118,
+      "step": 29475
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.2674018423135895,
+      "learning_rate": 4.9674632107023415e-06,
+      "loss": 0.2054,
+      "step": 29500
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.7825070235459493,
+      "learning_rate": 4.967421404682274e-06,
+      "loss": 0.2254,
+      "step": 29525
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.043836906677806,
+      "learning_rate": 4.967379598662208e-06,
+      "loss": 0.2188,
+      "step": 29550
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.0876415569608375,
+      "learning_rate": 4.9673377926421405e-06,
+      "loss": 0.2041,
+      "step": 29575
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.804802164781188,
+      "learning_rate": 4.967295986622074e-06,
+      "loss": 0.1919,
+      "step": 29600
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.183654886861711,
+      "learning_rate": 4.967254180602007e-06,
+      "loss": 0.1988,
+      "step": 29625
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.8097074308078962,
+      "learning_rate": 4.96721237458194e-06,
+      "loss": 0.2106,
+      "step": 29650
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.8124071979254552,
+      "learning_rate": 4.967170568561873e-06,
+      "loss": 0.2138,
+      "step": 29675
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.4488631236080394,
+      "learning_rate": 4.9671287625418065e-06,
+      "loss": 0.2091,
+      "step": 29700
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.3220655136855233,
+      "learning_rate": 4.967086956521739e-06,
+      "loss": 0.2245,
+      "step": 29725
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.5358184563094297,
+      "learning_rate": 4.967045150501673e-06,
+      "loss": 0.1831,
+      "step": 29750
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.5721622318412813,
+      "learning_rate": 4.9670033444816055e-06,
+      "loss": 0.2164,
+      "step": 29775
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.3569247099530573,
+      "learning_rate": 4.966961538461539e-06,
+      "loss": 0.2083,
+      "step": 29800
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.919524305810182,
+      "learning_rate": 4.966919732441472e-06,
+      "loss": 0.2012,
+      "step": 29825
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 1.6943687241906875,
+      "learning_rate": 4.966877926421405e-06,
+      "loss": 0.223,
+      "step": 29850
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.6313740896270725,
+      "learning_rate": 4.966836120401338e-06,
+      "loss": 0.2194,
+      "step": 29875
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.6895490084997424,
+      "learning_rate": 4.9667943143812715e-06,
+      "loss": 0.2077,
+      "step": 29900
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.601228861172109,
+      "learning_rate": 4.966752508361204e-06,
+      "loss": 0.1814,
+      "step": 29925
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.219366909148872,
+      "learning_rate": 4.966710702341138e-06,
+      "loss": 0.2117,
+      "step": 29950
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.2499897794948742,
+      "learning_rate": 4.9666688963210704e-06,
+      "loss": 0.1791,
+      "step": 29975
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.3071238159646315,
+      "learning_rate": 4.966627090301004e-06,
+      "loss": 0.2244,
+      "step": 30000
+    },
+    {
+      "epoch": 1.18,
+      "eval_loss": 0.5361328125,
+      "eval_runtime": 11518.7802,
+      "eval_samples_per_second": 0.822,
+      "eval_steps_per_second": 0.051,
+      "eval_wer": 0.12534787574602255,
+      "step": 30000
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.7823343242326686,
+      "learning_rate": 4.966585284280937e-06,
+      "loss": 0.2065,
+      "step": 30025
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.5221486291926747,
+      "learning_rate": 4.96654347826087e-06,
+      "loss": 0.1987,
+      "step": 30050
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.319001734849566,
+      "learning_rate": 4.966501672240803e-06,
+      "loss": 0.1983,
+      "step": 30075
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.7989713632094604,
+      "learning_rate": 4.9664598662207365e-06,
+      "loss": 0.2139,
+      "step": 30100
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.6123524206023738,
+      "learning_rate": 4.966418060200669e-06,
+      "loss": 0.2362,
+      "step": 30125
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.1978865463050377,
+      "learning_rate": 4.966376254180603e-06,
+      "loss": 0.2037,
+      "step": 30150
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.3585992261357787,
+      "learning_rate": 4.966334448160535e-06,
+      "loss": 0.2092,
+      "step": 30175
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.5811291808324892,
+      "learning_rate": 4.966292642140468e-06,
+      "loss": 0.195,
+      "step": 30200
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.6853682024289205,
+      "learning_rate": 4.966250836120402e-06,
+      "loss": 0.2238,
+      "step": 30225
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7722744670231796,
+      "learning_rate": 4.9662107023411375e-06,
+      "loss": 0.2075,
+      "step": 30250
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.881184080308395,
+      "learning_rate": 4.966168896321071e-06,
+      "loss": 0.1813,
+      "step": 30275
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.0946512002383884,
+      "learning_rate": 4.966127090301004e-06,
+      "loss": 0.2055,
+      "step": 30300
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.3313241329178456,
+      "learning_rate": 4.966085284280937e-06,
+      "loss": 0.2141,
+      "step": 30325
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.3531092149433521,
+      "learning_rate": 4.96604347826087e-06,
+      "loss": 0.1969,
+      "step": 30350
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.9444146528281916,
+      "learning_rate": 4.9660016722408035e-06,
+      "loss": 0.2217,
+      "step": 30375
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.1850621451492553,
+      "learning_rate": 4.965959866220736e-06,
+      "loss": 0.2144,
+      "step": 30400
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.8003297169728758,
+      "learning_rate": 4.96591806020067e-06,
+      "loss": 0.1884,
+      "step": 30425
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.301931614004386,
+      "learning_rate": 4.9658762541806025e-06,
+      "loss": 0.2166,
+      "step": 30450
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.4598479583328665,
+      "learning_rate": 4.965834448160535e-06,
+      "loss": 0.2008,
+      "step": 30475
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.4011597406764777,
+      "learning_rate": 4.965792642140469e-06,
+      "loss": 0.2057,
+      "step": 30500
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.0890831441737023,
+      "learning_rate": 4.965750836120401e-06,
+      "loss": 0.2002,
+      "step": 30525
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.0317922862792932,
+      "learning_rate": 4.965709030100335e-06,
+      "loss": 0.2098,
+      "step": 30550
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.3339129809340755,
+      "learning_rate": 4.965667224080268e-06,
+      "loss": 0.2073,
+      "step": 30575
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.3910047992276036,
+      "learning_rate": 4.965625418060201e-06,
+      "loss": 0.2241,
+      "step": 30600
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.937110026502965,
+      "learning_rate": 4.965583612040134e-06,
+      "loss": 0.2048,
+      "step": 30625
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.4355058481059215,
+      "learning_rate": 4.9655418060200674e-06,
+      "loss": 0.2099,
+      "step": 30650
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.5529814044707928,
+      "learning_rate": 4.9655e-06,
+      "loss": 0.1749,
+      "step": 30675
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.247459234736527,
+      "learning_rate": 4.965458193979934e-06,
+      "loss": 0.2111,
+      "step": 30700
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.839503804605756,
+      "learning_rate": 4.965416387959866e-06,
+      "loss": 0.2074,
+      "step": 30725
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8945095656682283,
+      "learning_rate": 4.9653745819398e-06,
+      "loss": 0.1872,
+      "step": 30750
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8914319288406296,
+      "learning_rate": 4.965332775919733e-06,
+      "loss": 0.2194,
+      "step": 30775
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.8360889606904163,
+      "learning_rate": 4.965290969899666e-06,
+      "loss": 0.2243,
+      "step": 30800
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.7850880114912815,
+      "learning_rate": 4.965249163879599e-06,
+      "loss": 0.2205,
+      "step": 30825
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.596884049166259,
+      "learning_rate": 4.9652073578595324e-06,
+      "loss": 0.221,
+      "step": 30850
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.9903428122299165,
+      "learning_rate": 4.965165551839465e-06,
+      "loss": 0.2093,
+      "step": 30875
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.5969796027518697,
+      "learning_rate": 4.965123745819399e-06,
+      "loss": 0.2066,
+      "step": 30900
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.4232465281818925,
+      "learning_rate": 4.965081939799331e-06,
+      "loss": 0.1918,
+      "step": 30925
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.6069161329849821,
+      "learning_rate": 4.965040133779265e-06,
+      "loss": 0.2285,
+      "step": 30950
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.2697223955884902,
+      "learning_rate": 4.964998327759198e-06,
+      "loss": 0.2016,
+      "step": 30975
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.2930314056749161,
+      "learning_rate": 4.964956521739131e-06,
+      "loss": 0.2227,
+      "step": 31000
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.1212062680858963,
+      "learning_rate": 4.964914715719064e-06,
+      "loss": 0.2011,
+      "step": 31025
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9658032132564844,
+      "learning_rate": 4.964872909698997e-06,
+      "loss": 0.2285,
+      "step": 31050
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.3693017625531883,
+      "learning_rate": 4.96483110367893e-06,
+      "loss": 0.216,
+      "step": 31075
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9956647662239217,
+      "learning_rate": 4.964789297658864e-06,
+      "loss": 0.2185,
+      "step": 31100
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.202282939519141,
+      "learning_rate": 4.964747491638796e-06,
+      "loss": 0.2048,
+      "step": 31125
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.6666987073505568,
+      "learning_rate": 4.964705685618729e-06,
+      "loss": 0.2308,
+      "step": 31150
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9958955477076095,
+      "learning_rate": 4.964663879598663e-06,
+      "loss": 0.2031,
+      "step": 31175
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.4079118493249374,
+      "learning_rate": 4.964622073578595e-06,
+      "loss": 0.207,
+      "step": 31200
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.6437168762178533,
+      "learning_rate": 4.964580267558529e-06,
+      "loss": 0.2088,
+      "step": 31225
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.0817719152582503,
+      "learning_rate": 4.964540133779265e-06,
+      "loss": 0.1928,
+      "step": 31250
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.9167781534885544,
+      "learning_rate": 4.964498327759198e-06,
+      "loss": 0.1914,
+      "step": 31275
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9066328276379372,
+      "learning_rate": 4.964456521739131e-06,
+      "loss": 0.2069,
+      "step": 31300
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.9776175873759618,
+      "learning_rate": 4.9644147157190645e-06,
+      "loss": 0.2121,
+      "step": 31325
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.3144155588840976,
+      "learning_rate": 4.964372909698997e-06,
+      "loss": 0.1927,
+      "step": 31350
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.4071128683786893,
+      "learning_rate": 4.96433110367893e-06,
+      "loss": 0.1953,
+      "step": 31375
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.0973251877720371,
+      "learning_rate": 4.964289297658863e-06,
+      "loss": 0.2166,
+      "step": 31400
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.6240742135179362,
+      "learning_rate": 4.964247491638796e-06,
+      "loss": 0.2036,
+      "step": 31425
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9782177053431977,
+      "learning_rate": 4.96420568561873e-06,
+      "loss": 0.2265,
+      "step": 31450
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.5819617657187213,
+      "learning_rate": 4.964163879598662e-06,
+      "loss": 0.1893,
+      "step": 31475
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.0167586079498867,
+      "learning_rate": 4.964122073578596e-06,
+      "loss": 0.2175,
+      "step": 31500
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.6880449215025803,
+      "learning_rate": 4.964080267558529e-06,
+      "loss": 0.2047,
+      "step": 31525
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.3911746414902157,
+      "learning_rate": 4.964038461538462e-06,
+      "loss": 0.2083,
+      "step": 31550
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.6077306397848325,
+      "learning_rate": 4.963996655518395e-06,
+      "loss": 0.2038,
+      "step": 31575
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8547104842544427,
+      "learning_rate": 4.963954849498328e-06,
+      "loss": 0.1909,
+      "step": 31600
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.2230790698786718,
+      "learning_rate": 4.963913043478261e-06,
+      "loss": 0.2144,
+      "step": 31625
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.6684024560293576,
+      "learning_rate": 4.963871237458195e-06,
+      "loss": 0.2259,
+      "step": 31650
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.4888511188162146,
+      "learning_rate": 4.963829431438127e-06,
+      "loss": 0.1831,
+      "step": 31675
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.078967899199761,
+      "learning_rate": 4.963787625418061e-06,
+      "loss": 0.2105,
+      "step": 31700
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.510441172072911,
+      "learning_rate": 4.9637458193979936e-06,
+      "loss": 0.1981,
+      "step": 31725
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.646878944946446,
+      "learning_rate": 4.963704013377927e-06,
+      "loss": 0.2097,
+      "step": 31750
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.8328540352865563,
+      "learning_rate": 4.96366220735786e-06,
+      "loss": 0.1886,
+      "step": 31775
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.1374340269320808,
+      "learning_rate": 4.963620401337793e-06,
+      "loss": 0.191,
+      "step": 31800
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9988391709765514,
+      "learning_rate": 4.963578595317726e-06,
+      "loss": 0.2059,
+      "step": 31825
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.1328657514428446,
+      "learning_rate": 4.96353678929766e-06,
+      "loss": 0.2285,
+      "step": 31850
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9394059197287883,
+      "learning_rate": 4.963494983277592e-06,
+      "loss": 0.2108,
+      "step": 31875
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.0461897291306157,
+      "learning_rate": 4.963453177257526e-06,
+      "loss": 0.2116,
+      "step": 31900
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.3526856502795168,
+      "learning_rate": 4.9634113712374585e-06,
+      "loss": 0.2267,
+      "step": 31925
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8917535020913774,
+      "learning_rate": 4.963369565217392e-06,
+      "loss": 0.2163,
+      "step": 31950
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.4229763247970235,
+      "learning_rate": 4.963327759197325e-06,
+      "loss": 0.1974,
+      "step": 31975
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.091442768892304,
+      "learning_rate": 4.963285953177258e-06,
+      "loss": 0.2062,
+      "step": 32000
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.513372953462432,
+      "learning_rate": 4.963244147157191e-06,
+      "loss": 0.2136,
+      "step": 32025
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.43008542750968,
+      "learning_rate": 4.9632023411371246e-06,
+      "loss": 0.2091,
+      "step": 32050
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.199026863323161,
+      "learning_rate": 4.963160535117057e-06,
+      "loss": 0.2209,
+      "step": 32075
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.5908105703435922,
+      "learning_rate": 4.96311872909699e-06,
+      "loss": 0.2005,
+      "step": 32100
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.4953338699146124,
+      "learning_rate": 4.9630769230769235e-06,
+      "loss": 0.2449,
+      "step": 32125
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.502883147552933,
+      "learning_rate": 4.963035117056856e-06,
+      "loss": 0.2131,
+      "step": 32150
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.751577299753483,
+      "learning_rate": 4.96299331103679e-06,
+      "loss": 0.217,
+      "step": 32175
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.394809862880557,
+      "learning_rate": 4.9629515050167225e-06,
+      "loss": 0.2091,
+      "step": 32200
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.2060001394612345,
+      "learning_rate": 4.962909698996656e-06,
+      "loss": 0.2069,
+      "step": 32225
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9708415738072285,
+      "learning_rate": 4.962869565217392e-06,
+      "loss": 0.2026,
+      "step": 32250
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.001205087461328,
+      "learning_rate": 4.962827759197325e-06,
+      "loss": 0.2032,
+      "step": 32275
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.266515112142646,
+      "learning_rate": 4.962785953177258e-06,
+      "loss": 0.2161,
+      "step": 32300
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.5658738110557742,
+      "learning_rate": 4.962744147157191e-06,
+      "loss": 0.196,
+      "step": 32325
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9588945387829928,
+      "learning_rate": 4.962702341137124e-06,
+      "loss": 0.2175,
+      "step": 32350
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.5897518353484381,
+      "learning_rate": 4.962660535117057e-06,
+      "loss": 0.2134,
+      "step": 32375
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.2728706316819134,
+      "learning_rate": 4.9626187290969906e-06,
+      "loss": 0.1953,
+      "step": 32400
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.391470189830352,
+      "learning_rate": 4.962576923076923e-06,
+      "loss": 0.21,
+      "step": 32425
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.5455063030077856,
+      "learning_rate": 4.962535117056857e-06,
+      "loss": 0.1955,
+      "step": 32450
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.8416205159711543,
+      "learning_rate": 4.9624933110367895e-06,
+      "loss": 0.1971,
+      "step": 32475
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.6448546575874654,
+      "learning_rate": 4.962451505016723e-06,
+      "loss": 0.2074,
+      "step": 32500
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 3.3357850095647246,
+      "learning_rate": 4.962409698996656e-06,
+      "loss": 0.2314,
+      "step": 32525
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.9861129354509446,
+      "learning_rate": 4.962367892976589e-06,
+      "loss": 0.2014,
+      "step": 32550
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.3127887409214314,
+      "learning_rate": 4.962326086956522e-06,
+      "loss": 0.2087,
+      "step": 32575
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.5428255216234126,
+      "learning_rate": 4.9622842809364555e-06,
+      "loss": 0.1927,
+      "step": 32600
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.265547190895961,
+      "learning_rate": 4.962242474916388e-06,
+      "loss": 0.2127,
+      "step": 32625
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.1891946036863335,
+      "learning_rate": 4.962200668896322e-06,
+      "loss": 0.1993,
+      "step": 32650
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.7958715126981217,
+      "learning_rate": 4.9621588628762545e-06,
+      "loss": 0.214,
+      "step": 32675
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.6405593238103398,
+      "learning_rate": 4.962117056856188e-06,
+      "loss": 0.2097,
+      "step": 32700
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.5373448233299254,
+      "learning_rate": 4.962075250836121e-06,
+      "loss": 0.2151,
+      "step": 32725
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0068111663697987,
+      "learning_rate": 4.962033444816054e-06,
+      "loss": 0.1893,
+      "step": 32750
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.0679584686902626,
+      "learning_rate": 4.961991638795987e-06,
+      "loss": 0.2236,
+      "step": 32775
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.2454742075576373,
+      "learning_rate": 4.9619498327759205e-06,
+      "loss": 0.1778,
+      "step": 32800
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.494891691231001,
+      "learning_rate": 4.961908026755853e-06,
+      "loss": 0.2111,
+      "step": 32825
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.972901323601928,
+      "learning_rate": 4.961866220735787e-06,
+      "loss": 0.2329,
+      "step": 32850
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.4863004691478272,
+      "learning_rate": 4.9618244147157195e-06,
+      "loss": 0.2036,
+      "step": 32875
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.2875445717253988,
+      "learning_rate": 4.961782608695653e-06,
+      "loss": 0.2026,
+      "step": 32900
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.5038045805937086,
+      "learning_rate": 4.961740802675586e-06,
+      "loss": 0.207,
+      "step": 32925
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.649781568126332,
+      "learning_rate": 4.961698996655519e-06,
+      "loss": 0.1957,
+      "step": 32950
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.451803928656282,
+      "learning_rate": 4.961657190635452e-06,
+      "loss": 0.2139,
+      "step": 32975
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.096939736947867,
+      "learning_rate": 4.9616153846153855e-06,
+      "loss": 0.2167,
+      "step": 33000
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.5804173291083985,
+      "learning_rate": 4.961573578595318e-06,
+      "loss": 0.2084,
+      "step": 33025
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.900713897980342,
+      "learning_rate": 4.961531772575251e-06,
+      "loss": 0.2094,
+      "step": 33050
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.8047019720202782,
+      "learning_rate": 4.9614899665551844e-06,
+      "loss": 0.2045,
+      "step": 33075
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.663316027429542,
+      "learning_rate": 4.961448160535117e-06,
+      "loss": 0.2075,
+      "step": 33100
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9090718281643422,
+      "learning_rate": 4.961406354515051e-06,
+      "loss": 0.1984,
+      "step": 33125
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.8038054052477845,
+      "learning_rate": 4.961364548494983e-06,
+      "loss": 0.2071,
+      "step": 33150
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9119519029509469,
+      "learning_rate": 4.961322742474917e-06,
+      "loss": 0.2132,
+      "step": 33175
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.9349267181687344,
+      "learning_rate": 4.96128093645485e-06,
+      "loss": 0.2159,
+      "step": 33200
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.550384787253934,
+      "learning_rate": 4.961239130434783e-06,
+      "loss": 0.1922,
+      "step": 33225
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.1539016786904788,
+      "learning_rate": 4.961198996655519e-06,
+      "loss": 0.2158,
+      "step": 33250
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.1496086242086543,
+      "learning_rate": 4.961157190635452e-06,
+      "loss": 0.2176,
+      "step": 33275
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.6905461372752362,
+      "learning_rate": 4.961115384615384e-06,
+      "loss": 0.1866,
+      "step": 33300
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.1509723343168803,
+      "learning_rate": 4.961073578595318e-06,
+      "loss": 0.2218,
+      "step": 33325
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.0867727056995953,
+      "learning_rate": 4.961031772575251e-06,
+      "loss": 0.2073,
+      "step": 33350
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.5010059727175604,
+      "learning_rate": 4.960989966555184e-06,
+      "loss": 0.2009,
+      "step": 33375
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.3178677208541214,
+      "learning_rate": 4.960948160535117e-06,
+      "loss": 0.2053,
+      "step": 33400
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.5525133283022474,
+      "learning_rate": 4.9609063545150504e-06,
+      "loss": 0.1934,
+      "step": 33425
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.5592614964414115,
+      "learning_rate": 4.960864548494984e-06,
+      "loss": 0.2024,
+      "step": 33450
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.1553832312538863,
+      "learning_rate": 4.960822742474917e-06,
+      "loss": 0.2109,
+      "step": 33475
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8607289492966907,
+      "learning_rate": 4.96078093645485e-06,
+      "loss": 0.1962,
+      "step": 33500
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.503789861400716,
+      "learning_rate": 4.960739130434783e-06,
+      "loss": 0.2088,
+      "step": 33525
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8929282998591728,
+      "learning_rate": 4.9606973244147165e-06,
+      "loss": 0.2209,
+      "step": 33550
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.0496277495832924,
+      "learning_rate": 4.960655518394649e-06,
+      "loss": 0.1954,
+      "step": 33575
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.4165638236647178,
+      "learning_rate": 4.960613712374583e-06,
+      "loss": 0.184,
+      "step": 33600
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.3598210314527397,
+      "learning_rate": 4.960571906354515e-06,
+      "loss": 0.2002,
+      "step": 33625
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.4783079603213292,
+      "learning_rate": 4.960530100334449e-06,
+      "loss": 0.2033,
+      "step": 33650
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.9540334215512212,
+      "learning_rate": 4.960488294314382e-06,
+      "loss": 0.2249,
+      "step": 33675
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.1034996218987867,
+      "learning_rate": 4.960446488294315e-06,
+      "loss": 0.1903,
+      "step": 33700
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.3139815085334834,
+      "learning_rate": 4.960404682274248e-06,
+      "loss": 0.1939,
+      "step": 33725
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.9968667345143176,
+      "learning_rate": 4.9603628762541814e-06,
+      "loss": 0.2057,
+      "step": 33750
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.5320555945517205,
+      "learning_rate": 4.960321070234114e-06,
+      "loss": 0.2073,
+      "step": 33775
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.282025684245721,
+      "learning_rate": 4.960279264214048e-06,
+      "loss": 0.2133,
+      "step": 33800
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.777269893458473,
+      "learning_rate": 4.96023745819398e-06,
+      "loss": 0.2243,
+      "step": 33825
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.5943182163866918,
+      "learning_rate": 4.960195652173914e-06,
+      "loss": 0.2153,
+      "step": 33850
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.379151340617046,
+      "learning_rate": 4.960153846153847e-06,
+      "loss": 0.2043,
+      "step": 33875
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.1521103544665638,
+      "learning_rate": 4.96011204013378e-06,
+      "loss": 0.2238,
+      "step": 33900
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.2182907874261988,
+      "learning_rate": 4.960070234113713e-06,
+      "loss": 0.2097,
+      "step": 33925
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.387589834089496,
+      "learning_rate": 4.960028428093646e-06,
+      "loss": 0.1899,
+      "step": 33950
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.5191996191680612,
+      "learning_rate": 4.959986622073579e-06,
+      "loss": 0.2107,
+      "step": 33975
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.4337854691253606,
+      "learning_rate": 4.959944816053512e-06,
+      "loss": 0.2131,
+      "step": 34000
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.8589806989817563,
+      "learning_rate": 4.959903010033445e-06,
+      "loss": 0.2162,
+      "step": 34025
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.303077233635094,
+      "learning_rate": 4.959861204013378e-06,
+      "loss": 0.203,
+      "step": 34050
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.5672615001886725,
+      "learning_rate": 4.959819397993312e-06,
+      "loss": 0.2073,
+      "step": 34075
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.163916859493643,
+      "learning_rate": 4.959777591973244e-06,
+      "loss": 0.2,
+      "step": 34100
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.6051761264160214,
+      "learning_rate": 4.959735785953178e-06,
+      "loss": 0.2052,
+      "step": 34125
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.3556812408093033,
+      "learning_rate": 4.9596939799331106e-06,
+      "loss": 0.2308,
+      "step": 34150
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.3137216096393576,
+      "learning_rate": 4.959652173913044e-06,
+      "loss": 0.2077,
+      "step": 34175
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.5157679496546062,
+      "learning_rate": 4.959610367892977e-06,
+      "loss": 0.2038,
+      "step": 34200
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.236311175171304,
+      "learning_rate": 4.95956856187291e-06,
+      "loss": 0.1943,
+      "step": 34225
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.4890581893083967,
+      "learning_rate": 4.959528428093645e-06,
+      "loss": 0.2026,
+      "step": 34250
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.0697802575341915,
+      "learning_rate": 4.959486622073579e-06,
+      "loss": 0.2175,
+      "step": 34275
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.0491147263632876,
+      "learning_rate": 4.9594448160535116e-06,
+      "loss": 0.2117,
+      "step": 34300
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.46097970439395,
+      "learning_rate": 4.959403010033445e-06,
+      "loss": 0.1839,
+      "step": 34325
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.014542005118715,
+      "learning_rate": 4.959361204013378e-06,
+      "loss": 0.2171,
+      "step": 34350
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.3888128070932415,
+      "learning_rate": 4.959319397993311e-06,
+      "loss": 0.1863,
+      "step": 34375
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.7106357207309615,
+      "learning_rate": 4.959277591973244e-06,
+      "loss": 0.1865,
+      "step": 34400
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.7455015346010385,
+      "learning_rate": 4.959235785953178e-06,
+      "loss": 0.2059,
+      "step": 34425
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.9415841930783881,
+      "learning_rate": 4.95919397993311e-06,
+      "loss": 0.2034,
+      "step": 34450
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.3452049148270158,
+      "learning_rate": 4.959152173913044e-06,
+      "loss": 0.2055,
+      "step": 34475
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.4039121644682209,
+      "learning_rate": 4.9591103678929765e-06,
+      "loss": 0.1914,
+      "step": 34500
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.4796081378428405,
+      "learning_rate": 4.95906856187291e-06,
+      "loss": 0.2172,
+      "step": 34525
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9045711516273565,
+      "learning_rate": 4.959026755852843e-06,
+      "loss": 0.1927,
+      "step": 34550
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.7547093039225692,
+      "learning_rate": 4.958984949832776e-06,
+      "loss": 0.228,
+      "step": 34575
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9785998947274708,
+      "learning_rate": 4.95894314381271e-06,
+      "loss": 0.2026,
+      "step": 34600
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.2354236137870933,
+      "learning_rate": 4.958901337792643e-06,
+      "loss": 0.214,
+      "step": 34625
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.3622120646886287,
+      "learning_rate": 4.958859531772576e-06,
+      "loss": 0.1868,
+      "step": 34650
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.573347646006675,
+      "learning_rate": 4.958817725752509e-06,
+      "loss": 0.1997,
+      "step": 34675
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8322249477184158,
+      "learning_rate": 4.958775919732442e-06,
+      "loss": 0.2188,
+      "step": 34700
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.4524561976880421,
+      "learning_rate": 4.958734113712375e-06,
+      "loss": 0.2097,
+      "step": 34725
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.259548939098752,
+      "learning_rate": 4.958692307692309e-06,
+      "loss": 0.2152,
+      "step": 34750
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.477510113801705,
+      "learning_rate": 4.958650501672241e-06,
+      "loss": 0.1898,
+      "step": 34775
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.2690260150220294,
+      "learning_rate": 4.958608695652175e-06,
+      "loss": 0.2182,
+      "step": 34800
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.5844701742556588,
+      "learning_rate": 4.9585668896321076e-06,
+      "loss": 0.1961,
+      "step": 34825
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.273853459237293,
+      "learning_rate": 4.958525083612041e-06,
+      "loss": 0.1922,
+      "step": 34850
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.8175047287317434,
+      "learning_rate": 4.958483277591974e-06,
+      "loss": 0.2038,
+      "step": 34875
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.836042614757162,
+      "learning_rate": 4.958441471571907e-06,
+      "loss": 0.2168,
+      "step": 34900
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.3796229553862953,
+      "learning_rate": 4.958399665551839e-06,
+      "loss": 0.2062,
+      "step": 34925
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.3102116543430355,
+      "learning_rate": 4.958357859531773e-06,
+      "loss": 0.2078,
+      "step": 34950
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.360851412118792,
+      "learning_rate": 4.9583160535117054e-06,
+      "loss": 0.1949,
+      "step": 34975
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.2283119602827541,
+      "learning_rate": 4.958274247491639e-06,
+      "loss": 0.2074,
+      "step": 35000
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.13008899942473,
+      "learning_rate": 4.9582324414715725e-06,
+      "loss": 0.2098,
+      "step": 35025
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.033370788229089,
+      "learning_rate": 4.958190635451505e-06,
+      "loss": 0.2019,
+      "step": 35050
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.2650220644179653,
+      "learning_rate": 4.958148829431439e-06,
+      "loss": 0.2119,
+      "step": 35075
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.6882292388102216,
+      "learning_rate": 4.9581070234113715e-06,
+      "loss": 0.204,
+      "step": 35100
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.5380929836431863,
+      "learning_rate": 4.958065217391305e-06,
+      "loss": 0.2127,
+      "step": 35125
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.1509889226964374,
+      "learning_rate": 4.958023411371238e-06,
+      "loss": 0.1943,
+      "step": 35150
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.4326488451072428,
+      "learning_rate": 4.957981605351171e-06,
+      "loss": 0.1855,
+      "step": 35175
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.383235843774178,
+      "learning_rate": 4.957939799331104e-06,
+      "loss": 0.2132,
+      "step": 35200
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.5270733379843522,
+      "learning_rate": 4.9578979933110375e-06,
+      "loss": 0.1998,
+      "step": 35225
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.4047334942033511,
+      "learning_rate": 4.9578578595317725e-06,
+      "loss": 0.2344,
+      "step": 35250
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.3056747609702148,
+      "learning_rate": 4.957816053511706e-06,
+      "loss": 0.1981,
+      "step": 35275
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8432393756629464,
+      "learning_rate": 4.957774247491639e-06,
+      "loss": 0.1912,
+      "step": 35300
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.255197218187179,
+      "learning_rate": 4.957732441471572e-06,
+      "loss": 0.1947,
+      "step": 35325
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.873243680140684,
+      "learning_rate": 4.957690635451505e-06,
+      "loss": 0.2078,
+      "step": 35350
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.048000455141224,
+      "learning_rate": 4.9576488294314385e-06,
+      "loss": 0.2059,
+      "step": 35375
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 3.4680651759429137,
+      "learning_rate": 4.957607023411371e-06,
+      "loss": 0.22,
+      "step": 35400
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.53131994025086,
+      "learning_rate": 4.957565217391305e-06,
+      "loss": 0.2256,
+      "step": 35425
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.8548841688370339,
+      "learning_rate": 4.9575234113712375e-06,
+      "loss": 0.2143,
+      "step": 35450
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.6403964177995343,
+      "learning_rate": 4.957481605351171e-06,
+      "loss": 0.2128,
+      "step": 35475
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.6156960137162244,
+      "learning_rate": 4.957439799331104e-06,
+      "loss": 0.2203,
+      "step": 35500
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.5593538805691032,
+      "learning_rate": 4.957397993311037e-06,
+      "loss": 0.2111,
+      "step": 35525
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.3940280246709056,
+      "learning_rate": 4.95735618729097e-06,
+      "loss": 0.2053,
+      "step": 35550
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.2706106350468436,
+      "learning_rate": 4.9573143812709035e-06,
+      "loss": 0.2074,
+      "step": 35575
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.9178553441649687,
+      "learning_rate": 4.957272575250836e-06,
+      "loss": 0.1981,
+      "step": 35600
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.4261776676677842,
+      "learning_rate": 4.95723076923077e-06,
+      "loss": 0.1853,
+      "step": 35625
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.1696659611524045,
+      "learning_rate": 4.9571889632107024e-06,
+      "loss": 0.2134,
+      "step": 35650
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7602557743841436,
+      "learning_rate": 4.957147157190636e-06,
+      "loss": 0.2126,
+      "step": 35675
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.492907454571274,
+      "learning_rate": 4.9571053511705695e-06,
+      "loss": 0.216,
+      "step": 35700
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.367209562020745,
+      "learning_rate": 4.957063545150502e-06,
+      "loss": 0.2195,
+      "step": 35725
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8227019149029202,
+      "learning_rate": 4.957021739130436e-06,
+      "loss": 0.1937,
+      "step": 35750
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.2682518351707983,
+      "learning_rate": 4.9569799331103685e-06,
+      "loss": 0.2019,
+      "step": 35775
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.262927556471107,
+      "learning_rate": 4.956938127090302e-06,
+      "loss": 0.1928,
+      "step": 35800
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.42063661058105,
+      "learning_rate": 4.956896321070235e-06,
+      "loss": 0.2019,
+      "step": 35825
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.4089099655471107,
+      "learning_rate": 4.956854515050168e-06,
+      "loss": 0.2014,
+      "step": 35850
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.0092033649001106,
+      "learning_rate": 4.9568127090301e-06,
+      "loss": 0.2002,
+      "step": 35875
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.9427722122886302,
+      "learning_rate": 4.956770903010034e-06,
+      "loss": 0.2208,
+      "step": 35900
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.6812779714137775,
+      "learning_rate": 4.956729096989966e-06,
+      "loss": 0.1991,
+      "step": 35925
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.3752830801753009,
+      "learning_rate": 4.9566872909699e-06,
+      "loss": 0.2051,
+      "step": 35950
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.6857256233320776,
+      "learning_rate": 4.956645484949833e-06,
+      "loss": 0.206,
+      "step": 35975
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.84815296780552,
+      "learning_rate": 4.956603678929766e-06,
+      "loss": 0.2146,
+      "step": 36000
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.2132768250520185,
+      "learning_rate": 4.956561872909699e-06,
+      "loss": 0.1858,
+      "step": 36025
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.3922235347411074,
+      "learning_rate": 4.956520066889632e-06,
+      "loss": 0.1841,
+      "step": 36050
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.4109203186054706,
+      "learning_rate": 4.956478260869565e-06,
+      "loss": 0.2042,
+      "step": 36075
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.076333380375572,
+      "learning_rate": 4.956436454849499e-06,
+      "loss": 0.2103,
+      "step": 36100
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.28510984363013,
+      "learning_rate": 4.956394648829432e-06,
+      "loss": 0.1989,
+      "step": 36125
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.9777457396941752,
+      "learning_rate": 4.956352842809365e-06,
+      "loss": 0.2305,
+      "step": 36150
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.5988408503297,
+      "learning_rate": 4.9563110367892984e-06,
+      "loss": 0.2053,
+      "step": 36175
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.6674285290812338,
+      "learning_rate": 4.956269230769231e-06,
+      "loss": 0.2098,
+      "step": 36200
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.1275665989695556,
+      "learning_rate": 4.956227424749165e-06,
+      "loss": 0.2157,
+      "step": 36225
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.298315469696719,
+      "learning_rate": 4.9561872909699e-06,
+      "loss": 0.1852,
+      "step": 36250
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.4806491205789225,
+      "learning_rate": 4.956145484949833e-06,
+      "loss": 0.2023,
+      "step": 36275
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7496689306471362,
+      "learning_rate": 4.956103678929766e-06,
+      "loss": 0.2057,
+      "step": 36300
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.723557779898001,
+      "learning_rate": 4.9560618729096995e-06,
+      "loss": 0.217,
+      "step": 36325
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.700269519444899,
+      "learning_rate": 4.956020066889632e-06,
+      "loss": 0.1761,
+      "step": 36350
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.320993952778013,
+      "learning_rate": 4.955978260869566e-06,
+      "loss": 0.1874,
+      "step": 36375
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.5904614343317816,
+      "learning_rate": 4.955936454849498e-06,
+      "loss": 0.2055,
+      "step": 36400
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.1408285586096996,
+      "learning_rate": 4.955894648829432e-06,
+      "loss": 0.2046,
+      "step": 36425
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.7747940283874366,
+      "learning_rate": 4.955852842809365e-06,
+      "loss": 0.1819,
+      "step": 36450
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 1.4295834467887427,
+      "learning_rate": 4.955811036789298e-06,
+      "loss": 0.1995,
+      "step": 36475
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.2445542660386075,
+      "learning_rate": 4.955769230769231e-06,
+      "loss": 0.1994,
+      "step": 36500
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.4077690160385432,
+      "learning_rate": 4.9557274247491644e-06,
+      "loss": 0.2321,
+      "step": 36525
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.0359673240598635,
+      "learning_rate": 4.955685618729097e-06,
+      "loss": 0.2202,
+      "step": 36550
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.4659278382451848,
+      "learning_rate": 4.955643812709031e-06,
+      "loss": 0.2263,
+      "step": 36575
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.3295779187540848,
+      "learning_rate": 4.955602006688963e-06,
+      "loss": 0.2192,
+      "step": 36600
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.1071332705463832,
+      "learning_rate": 4.955560200668897e-06,
+      "loss": 0.2132,
+      "step": 36625
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.265229352325454,
+      "learning_rate": 4.95551839464883e-06,
+      "loss": 0.1988,
+      "step": 36650
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.9800704407638031,
+      "learning_rate": 4.955476588628763e-06,
+      "loss": 0.2032,
+      "step": 36675
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.9925356070357615,
+      "learning_rate": 4.955434782608696e-06,
+      "loss": 0.215,
+      "step": 36700
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.5204227004534854,
+      "learning_rate": 4.955392976588629e-06,
+      "loss": 0.2107,
+      "step": 36725
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.4452324885329624,
+      "learning_rate": 4.955351170568562e-06,
+      "loss": 0.2164,
+      "step": 36750
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.959056822372053,
+      "learning_rate": 4.955309364548496e-06,
+      "loss": 0.217,
+      "step": 36775
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.3043549385517452,
+      "learning_rate": 4.955267558528428e-06,
+      "loss": 0.2211,
+      "step": 36800
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.2806395544186833,
+      "learning_rate": 4.955225752508361e-06,
+      "loss": 0.1887,
+      "step": 36825
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.3016123333945022,
+      "learning_rate": 4.955183946488295e-06,
+      "loss": 0.2025,
+      "step": 36850
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.4292394564351278,
+      "learning_rate": 4.955142140468227e-06,
+      "loss": 0.2076,
+      "step": 36875
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9148503359389857,
+      "learning_rate": 4.955100334448161e-06,
+      "loss": 0.2046,
+      "step": 36900
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.6141787965473307,
+      "learning_rate": 4.9550585284280935e-06,
+      "loss": 0.2014,
+      "step": 36925
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.9101519644967175,
+      "learning_rate": 4.955016722408027e-06,
+      "loss": 0.2111,
+      "step": 36950
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.7403141468024392,
+      "learning_rate": 4.95497491638796e-06,
+      "loss": 0.2077,
+      "step": 36975
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7207730170828357,
+      "learning_rate": 4.954933110367893e-06,
+      "loss": 0.1927,
+      "step": 37000
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8605826806771688,
+      "learning_rate": 4.954891304347826e-06,
+      "loss": 0.2184,
+      "step": 37025
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.6175520560976995,
+      "learning_rate": 4.9548494983277596e-06,
+      "loss": 0.2098,
+      "step": 37050
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.8085251236732929,
+      "learning_rate": 4.954807692307692e-06,
+      "loss": 0.2257,
+      "step": 37075
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.3604634040424115,
+      "learning_rate": 4.954765886287626e-06,
+      "loss": 0.2019,
+      "step": 37100
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.6502918851829131,
+      "learning_rate": 4.9547240802675585e-06,
+      "loss": 0.1919,
+      "step": 37125
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.7617138039619495,
+      "learning_rate": 4.954682274247492e-06,
+      "loss": 0.2058,
+      "step": 37150
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.4810928710156535,
+      "learning_rate": 4.954640468227425e-06,
+      "loss": 0.1996,
+      "step": 37175
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.5350605129741175,
+      "learning_rate": 4.954598662207358e-06,
+      "loss": 0.2055,
+      "step": 37200
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.0090394477173805,
+      "learning_rate": 4.954556856187292e-06,
+      "loss": 0.1838,
+      "step": 37225
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.2115676280987648,
+      "learning_rate": 4.954516722408027e-06,
+      "loss": 0.2182,
+      "step": 37250
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.2809098857963606,
+      "learning_rate": 4.95447491638796e-06,
+      "loss": 0.2116,
+      "step": 37275
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.7286107298277917,
+      "learning_rate": 4.954433110367893e-06,
+      "loss": 0.1862,
+      "step": 37300
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.3745604372273177,
+      "learning_rate": 4.954391304347827e-06,
+      "loss": 0.2003,
+      "step": 37325
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.179529589103856,
+      "learning_rate": 4.954349498327759e-06,
+      "loss": 0.2023,
+      "step": 37350
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.379480802423559,
+      "learning_rate": 4.954307692307693e-06,
+      "loss": 0.2044,
+      "step": 37375
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.704416220877015,
+      "learning_rate": 4.9542658862876256e-06,
+      "loss": 0.2179,
+      "step": 37400
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.2195378522173297,
+      "learning_rate": 4.954224080267559e-06,
+      "loss": 0.2084,
+      "step": 37425
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.548215462518736,
+      "learning_rate": 4.954182274247492e-06,
+      "loss": 0.2095,
+      "step": 37450
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.488083163244097,
+      "learning_rate": 4.954140468227425e-06,
+      "loss": 0.1998,
+      "step": 37475
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.735700092430531,
+      "learning_rate": 4.954098662207358e-06,
+      "loss": 0.2143,
+      "step": 37500
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.3001115190622823,
+      "learning_rate": 4.954056856187292e-06,
+      "loss": 0.2107,
+      "step": 37525
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.724639356179923,
+      "learning_rate": 4.954015050167224e-06,
+      "loss": 0.2267,
+      "step": 37550
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.836013236024464,
+      "learning_rate": 4.953973244147158e-06,
+      "loss": 0.223,
+      "step": 37575
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.63422094494171,
+      "learning_rate": 4.9539314381270905e-06,
+      "loss": 0.1909,
+      "step": 37600
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.6384726135540142,
+      "learning_rate": 4.953889632107024e-06,
+      "loss": 0.2025,
+      "step": 37625
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.9438112258099434,
+      "learning_rate": 4.953847826086957e-06,
+      "loss": 0.1905,
+      "step": 37650
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.0403963480844216,
+      "learning_rate": 4.95380602006689e-06,
+      "loss": 0.2039,
+      "step": 37675
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.5742246476102364,
+      "learning_rate": 4.953764214046823e-06,
+      "loss": 0.2102,
+      "step": 37700
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.485529188887855,
+      "learning_rate": 4.953722408026757e-06,
+      "loss": 0.2192,
+      "step": 37725
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.761172543764633,
+      "learning_rate": 4.953680602006689e-06,
+      "loss": 0.2076,
+      "step": 37750
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.9982263034997709,
+      "learning_rate": 4.953638795986622e-06,
+      "loss": 0.2182,
+      "step": 37775
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.2204246293357652,
+      "learning_rate": 4.9535969899665555e-06,
+      "loss": 0.2172,
+      "step": 37800
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.3041545197815172,
+      "learning_rate": 4.953555183946488e-06,
+      "loss": 0.21,
+      "step": 37825
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.6384467244877707,
+      "learning_rate": 4.953513377926422e-06,
+      "loss": 0.2086,
+      "step": 37850
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.8368676858921458,
+      "learning_rate": 4.9534715719063545e-06,
+      "loss": 0.2193,
+      "step": 37875
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.68655526333874,
+      "learning_rate": 4.953429765886288e-06,
+      "loss": 0.2122,
+      "step": 37900
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.283888377497305,
+      "learning_rate": 4.953387959866221e-06,
+      "loss": 0.2013,
+      "step": 37925
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.1147280818501166,
+      "learning_rate": 4.953346153846154e-06,
+      "loss": 0.2119,
+      "step": 37950
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.131251084192741,
+      "learning_rate": 4.953304347826087e-06,
+      "loss": 0.1857,
+      "step": 37975
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.879560812566554,
+      "learning_rate": 4.9532625418060205e-06,
+      "loss": 0.2138,
+      "step": 38000
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.1426136680022196,
+      "learning_rate": 4.953220735785953e-06,
+      "loss": 0.1844,
+      "step": 38025
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.5973757490635936,
+      "learning_rate": 4.953178929765887e-06,
+      "loss": 0.2084,
+      "step": 38050
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.8939082265559706,
+      "learning_rate": 4.9531371237458194e-06,
+      "loss": 0.2129,
+      "step": 38075
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.496482622396757,
+      "learning_rate": 4.953095317725753e-06,
+      "loss": 0.2145,
+      "step": 38100
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.593350227368928,
+      "learning_rate": 4.953053511705686e-06,
+      "loss": 0.2064,
+      "step": 38125
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.4347207475008577,
+      "learning_rate": 4.953011705685619e-06,
+      "loss": 0.2101,
+      "step": 38150
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.5414470961110491,
+      "learning_rate": 4.952969899665552e-06,
+      "loss": 0.1994,
+      "step": 38175
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.2999288110788667,
+      "learning_rate": 4.9529280936454855e-06,
+      "loss": 0.19,
+      "step": 38200
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.3896836661728793,
+      "learning_rate": 4.952886287625418e-06,
+      "loss": 0.2014,
+      "step": 38225
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.395012428279921,
+      "learning_rate": 4.952846153846154e-06,
+      "loss": 0.1805,
+      "step": 38250
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.4366727872711265,
+      "learning_rate": 4.9528043478260875e-06,
+      "loss": 0.19,
+      "step": 38275
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.2645834734452694,
+      "learning_rate": 4.95276254180602e-06,
+      "loss": 0.1943,
+      "step": 38300
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.2963361249917345,
+      "learning_rate": 4.952720735785954e-06,
+      "loss": 0.1991,
+      "step": 38325
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.4866651771754051,
+      "learning_rate": 4.9526789297658865e-06,
+      "loss": 0.2245,
+      "step": 38350
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.9546225113518195,
+      "learning_rate": 4.95263712374582e-06,
+      "loss": 0.222,
+      "step": 38375
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.7859246566220357,
+      "learning_rate": 4.952595317725753e-06,
+      "loss": 0.2001,
+      "step": 38400
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.919164444138617,
+      "learning_rate": 4.952553511705686e-06,
+      "loss": 0.1928,
+      "step": 38425
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.3516142246793494,
+      "learning_rate": 4.952511705685619e-06,
+      "loss": 0.2054,
+      "step": 38450
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.8640375790937023,
+      "learning_rate": 4.9524698996655525e-06,
+      "loss": 0.1886,
+      "step": 38475
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.8124716196702333,
+      "learning_rate": 4.952428093645485e-06,
+      "loss": 0.2149,
+      "step": 38500
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.5193585343880305,
+      "learning_rate": 4.952386287625419e-06,
+      "loss": 0.1966,
+      "step": 38525
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.832736275735568,
+      "learning_rate": 4.9523444816053515e-06,
+      "loss": 0.2038,
+      "step": 38550
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.736944723639645,
+      "learning_rate": 4.952302675585285e-06,
+      "loss": 0.1933,
+      "step": 38575
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.630971173062863,
+      "learning_rate": 4.952260869565218e-06,
+      "loss": 0.2004,
+      "step": 38600
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.8873355888793588,
+      "learning_rate": 4.952219063545151e-06,
+      "loss": 0.2176,
+      "step": 38625
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.3335092351991724,
+      "learning_rate": 4.952177257525084e-06,
+      "loss": 0.2189,
+      "step": 38650
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.1938959308869856,
+      "learning_rate": 4.9521354515050175e-06,
+      "loss": 0.2158,
+      "step": 38675
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.1451216064786816,
+      "learning_rate": 4.95209364548495e-06,
+      "loss": 0.1996,
+      "step": 38700
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.1085029403514461,
+      "learning_rate": 4.952051839464883e-06,
+      "loss": 0.2116,
+      "step": 38725
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.4562454164261576,
+      "learning_rate": 4.9520100334448164e-06,
+      "loss": 0.2174,
+      "step": 38750
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.9786520420055989,
+      "learning_rate": 4.951968227424749e-06,
+      "loss": 0.1918,
+      "step": 38775
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.5097439606342058,
+      "learning_rate": 4.951926421404683e-06,
+      "loss": 0.229,
+      "step": 38800
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.7567248665862696,
+      "learning_rate": 4.951884615384615e-06,
+      "loss": 0.2072,
+      "step": 38825
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.7139284656959257,
+      "learning_rate": 4.951842809364549e-06,
+      "loss": 0.2102,
+      "step": 38850
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.7839798240238915,
+      "learning_rate": 4.951801003344482e-06,
+      "loss": 0.2117,
+      "step": 38875
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.3293739028446008,
+      "learning_rate": 4.951759197324415e-06,
+      "loss": 0.1878,
+      "step": 38900
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.6080849310086829,
+      "learning_rate": 4.951717391304348e-06,
+      "loss": 0.2101,
+      "step": 38925
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.250962569095591,
+      "learning_rate": 4.951675585284281e-06,
+      "loss": 0.1938,
+      "step": 38950
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.1190166838811497,
+      "learning_rate": 4.951633779264214e-06,
+      "loss": 0.2159,
+      "step": 38975
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.4326007393935587,
+      "learning_rate": 4.951591973244148e-06,
+      "loss": 0.1985,
+      "step": 39000
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.6414580007592534,
+      "learning_rate": 4.95155016722408e-06,
+      "loss": 0.2048,
+      "step": 39025
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.9216076329480545,
+      "learning_rate": 4.951508361204014e-06,
+      "loss": 0.2124,
+      "step": 39050
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.2728391425359964,
+      "learning_rate": 4.951466555183947e-06,
+      "loss": 0.2193,
+      "step": 39075
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.5009755698819605,
+      "learning_rate": 4.95142474916388e-06,
+      "loss": 0.2223,
+      "step": 39100
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.9770053209281309,
+      "learning_rate": 4.951382943143813e-06,
+      "loss": 0.2027,
+      "step": 39125
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.6300988064698791,
+      "learning_rate": 4.951341137123746e-06,
+      "loss": 0.2077,
+      "step": 39150
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.561710138010411,
+      "learning_rate": 4.951299331103679e-06,
+      "loss": 0.1915,
+      "step": 39175
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.705460436343843,
+      "learning_rate": 4.951257525083613e-06,
+      "loss": 0.192,
+      "step": 39200
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.5991161773874567,
+      "learning_rate": 4.951215719063545e-06,
+      "loss": 0.2001,
+      "step": 39225
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.2368222120398074,
+      "learning_rate": 4.951175585284281e-06,
+      "loss": 0.1984,
+      "step": 39250
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.787832230883491,
+      "learning_rate": 4.951133779264215e-06,
+      "loss": 0.2138,
+      "step": 39275
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.7658617749710668,
+      "learning_rate": 4.951091973244147e-06,
+      "loss": 0.2084,
+      "step": 39300
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.4069418293510692,
+      "learning_rate": 4.951050167224081e-06,
+      "loss": 0.209,
+      "step": 39325
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.167347269613874,
+      "learning_rate": 4.951008361204014e-06,
+      "loss": 0.2259,
+      "step": 39350
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.4967667199328691,
+      "learning_rate": 4.950966555183947e-06,
+      "loss": 0.1938,
+      "step": 39375
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.1203864757858373,
+      "learning_rate": 4.95092474916388e-06,
+      "loss": 0.1996,
+      "step": 39400
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.8801469194282086,
+      "learning_rate": 4.9508829431438135e-06,
+      "loss": 0.2039,
+      "step": 39425
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.327003950516436,
+      "learning_rate": 4.950841137123746e-06,
+      "loss": 0.2206,
+      "step": 39450
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.41131914892425,
+      "learning_rate": 4.95079933110368e-06,
+      "loss": 0.2014,
+      "step": 39475
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.0819564181136587,
+      "learning_rate": 4.950757525083612e-06,
+      "loss": 0.213,
+      "step": 39500
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.6673113392240486,
+      "learning_rate": 4.950715719063546e-06,
+      "loss": 0.2184,
+      "step": 39525
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.380558230210356,
+      "learning_rate": 4.950673913043479e-06,
+      "loss": 0.1943,
+      "step": 39550
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.1697122354109033,
+      "learning_rate": 4.950632107023412e-06,
+      "loss": 0.1906,
+      "step": 39575
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.9418778704353494,
+      "learning_rate": 4.950590301003345e-06,
+      "loss": 0.189,
+      "step": 39600
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.4589182357598085,
+      "learning_rate": 4.950548494983278e-06,
+      "loss": 0.1947,
+      "step": 39625
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.8193230995535195,
+      "learning_rate": 4.950506688963211e-06,
+      "loss": 0.2207,
+      "step": 39650
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.3338701052042354,
+      "learning_rate": 4.950464882943144e-06,
+      "loss": 0.2067,
+      "step": 39675
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.7410998436865661,
+      "learning_rate": 4.950423076923077e-06,
+      "loss": 0.1927,
+      "step": 39700
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.3042207681102427,
+      "learning_rate": 4.95038127090301e-06,
+      "loss": 0.2183,
+      "step": 39725
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.7924918618547276,
+      "learning_rate": 4.950339464882944e-06,
+      "loss": 0.2164,
+      "step": 39750
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.9474636138675756,
+      "learning_rate": 4.950297658862876e-06,
+      "loss": 0.205,
+      "step": 39775
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.6557124540827683,
+      "learning_rate": 4.95025585284281e-06,
+      "loss": 0.2181,
+      "step": 39800
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.2497598066214064,
+      "learning_rate": 4.9502140468227426e-06,
+      "loss": 0.2372,
+      "step": 39825
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.6087866416925465,
+      "learning_rate": 4.950172240802676e-06,
+      "loss": 0.2086,
+      "step": 39850
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.4725315047956657,
+      "learning_rate": 4.950130434782609e-06,
+      "loss": 0.2029,
+      "step": 39875
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.4054406114906426,
+      "learning_rate": 4.950088628762542e-06,
+      "loss": 0.1899,
+      "step": 39900
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.705726548553069,
+      "learning_rate": 4.950046822742475e-06,
+      "loss": 0.1959,
+      "step": 39925
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.3147943647253022,
+      "learning_rate": 4.950005016722409e-06,
+      "loss": 0.2099,
+      "step": 39950
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.199900310959841,
+      "learning_rate": 4.949963210702341e-06,
+      "loss": 0.1991,
+      "step": 39975
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.5347734934426658,
+      "learning_rate": 4.949921404682275e-06,
+      "loss": 0.2056,
+      "step": 40000
+    },
+    {
+      "epoch": 1.57,
+      "eval_loss": 0.471435546875,
+      "eval_runtime": 11603.9924,
+      "eval_samples_per_second": 0.816,
+      "eval_steps_per_second": 0.051,
+      "eval_wer": 0.12227530846323376,
+      "step": 40000
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.2625514121666555,
+      "learning_rate": 4.9498795986622075e-06,
+      "loss": 0.2043,
+      "step": 40025
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.598268805422133,
+      "learning_rate": 4.949837792642141e-06,
+      "loss": 0.2222,
+      "step": 40050
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.8626188809803597,
+      "learning_rate": 4.949795986622074e-06,
+      "loss": 0.2062,
+      "step": 40075
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.2772100379671945,
+      "learning_rate": 4.949754180602007e-06,
+      "loss": 0.1935,
+      "step": 40100
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.623164374874199,
+      "learning_rate": 4.94971237458194e-06,
+      "loss": 0.2072,
+      "step": 40125
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.8510882784804008,
+      "learning_rate": 4.9496705685618736e-06,
+      "loss": 0.2103,
+      "step": 40150
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.1111748213696224,
+      "learning_rate": 4.949628762541806e-06,
+      "loss": 0.2119,
+      "step": 40175
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.8817631319158592,
+      "learning_rate": 4.94958695652174e-06,
+      "loss": 0.2284,
+      "step": 40200
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.7672470719864912,
+      "learning_rate": 4.9495451505016725e-06,
+      "loss": 0.2094,
+      "step": 40225
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.5564127286984561,
+      "learning_rate": 4.949505016722408e-06,
+      "loss": 0.1975,
+      "step": 40250
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.6106298300859476,
+      "learning_rate": 4.949463210702342e-06,
+      "loss": 0.2036,
+      "step": 40275
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.9484881833986716,
+      "learning_rate": 4.949421404682275e-06,
+      "loss": 0.2186,
+      "step": 40300
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.9152229538502286,
+      "learning_rate": 4.949379598662208e-06,
+      "loss": 0.195,
+      "step": 40325
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.014374946325913,
+      "learning_rate": 4.949337792642141e-06,
+      "loss": 0.2268,
+      "step": 40350
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.4380015003183522,
+      "learning_rate": 4.949295986622074e-06,
+      "loss": 0.2022,
+      "step": 40375
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.8807119242703392,
+      "learning_rate": 4.949254180602007e-06,
+      "loss": 0.2071,
+      "step": 40400
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.2257784971118444,
+      "learning_rate": 4.949212374581941e-06,
+      "loss": 0.2172,
+      "step": 40425
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.3513506180545665,
+      "learning_rate": 4.949170568561873e-06,
+      "loss": 0.2005,
+      "step": 40450
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.7195640433114703,
+      "learning_rate": 4.949128762541807e-06,
+      "loss": 0.2044,
+      "step": 40475
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.8785362354699997,
+      "learning_rate": 4.9490869565217396e-06,
+      "loss": 0.2008,
+      "step": 40500
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.4693582668589034,
+      "learning_rate": 4.949045150501673e-06,
+      "loss": 0.1974,
+      "step": 40525
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.126779893482259,
+      "learning_rate": 4.949003344481606e-06,
+      "loss": 0.2031,
+      "step": 40550
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.0166603572842985,
+      "learning_rate": 4.9489615384615385e-06,
+      "loss": 0.203,
+      "step": 40575
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.530039181431065,
+      "learning_rate": 4.948919732441472e-06,
+      "loss": 0.205,
+      "step": 40600
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.2765710499831335,
+      "learning_rate": 4.948877926421405e-06,
+      "loss": 0.1917,
+      "step": 40625
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.361058220611067,
+      "learning_rate": 4.948836120401338e-06,
+      "loss": 0.1953,
+      "step": 40650
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.077832653814155,
+      "learning_rate": 4.948794314381271e-06,
+      "loss": 0.2052,
+      "step": 40675
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.4160926776241496,
+      "learning_rate": 4.9487525083612045e-06,
+      "loss": 0.2133,
+      "step": 40700
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.0655600758672423,
+      "learning_rate": 4.948710702341137e-06,
+      "loss": 0.1919,
+      "step": 40725
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.6603254897395308,
+      "learning_rate": 4.948668896321071e-06,
+      "loss": 0.2079,
+      "step": 40750
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.768524529622312,
+      "learning_rate": 4.9486270903010035e-06,
+      "loss": 0.2002,
+      "step": 40775
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.7012943976613921,
+      "learning_rate": 4.948585284280937e-06,
+      "loss": 0.2011,
+      "step": 40800
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.7082794174545288,
+      "learning_rate": 4.94854347826087e-06,
+      "loss": 0.2062,
+      "step": 40825
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.928279567494617,
+      "learning_rate": 4.948501672240803e-06,
+      "loss": 0.1939,
+      "step": 40850
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.9975469181079872,
+      "learning_rate": 4.948459866220736e-06,
+      "loss": 0.195,
+      "step": 40875
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.9802397774014037,
+      "learning_rate": 4.9484180602006695e-06,
+      "loss": 0.2166,
+      "step": 40900
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.443205417331069,
+      "learning_rate": 4.948376254180602e-06,
+      "loss": 0.219,
+      "step": 40925
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.8443352124627206,
+      "learning_rate": 4.948334448160536e-06,
+      "loss": 0.1943,
+      "step": 40950
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.193610360908578,
+      "learning_rate": 4.9482926421404685e-06,
+      "loss": 0.197,
+      "step": 40975
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.4623521124140761,
+      "learning_rate": 4.948250836120402e-06,
+      "loss": 0.2081,
+      "step": 41000
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.3374220657812355,
+      "learning_rate": 4.948209030100335e-06,
+      "loss": 0.1975,
+      "step": 41025
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.1336669393881718,
+      "learning_rate": 4.948167224080268e-06,
+      "loss": 0.203,
+      "step": 41050
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.2493898272913384,
+      "learning_rate": 4.948125418060201e-06,
+      "loss": 0.1948,
+      "step": 41075
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.2972621198738996,
+      "learning_rate": 4.9480836120401345e-06,
+      "loss": 0.2192,
+      "step": 41100
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.388290835184551,
+      "learning_rate": 4.948041806020067e-06,
+      "loss": 0.2166,
+      "step": 41125
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.5916665380726456,
+      "learning_rate": 4.948000000000001e-06,
+      "loss": 0.2136,
+      "step": 41150
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.8625346417846564,
+      "learning_rate": 4.9479581939799334e-06,
+      "loss": 0.2046,
+      "step": 41175
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.2016947650644971,
+      "learning_rate": 4.947916387959867e-06,
+      "loss": 0.1867,
+      "step": 41200
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.5740502812375239,
+      "learning_rate": 4.9478745819398e-06,
+      "loss": 0.1927,
+      "step": 41225
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.3675578577012184,
+      "learning_rate": 4.9478344481605355e-06,
+      "loss": 0.1967,
+      "step": 41250
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.3133012100842971,
+      "learning_rate": 4.947792642140469e-06,
+      "loss": 0.2053,
+      "step": 41275
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.8309449231056476,
+      "learning_rate": 4.947750836120402e-06,
+      "loss": 0.2045,
+      "step": 41300
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.000483291522688,
+      "learning_rate": 4.947709030100335e-06,
+      "loss": 0.2107,
+      "step": 41325
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.8991039424577518,
+      "learning_rate": 4.947667224080268e-06,
+      "loss": 0.1881,
+      "step": 41350
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.7183900225861886,
+      "learning_rate": 4.9476254180602015e-06,
+      "loss": 0.1982,
+      "step": 41375
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.5200229992542378,
+      "learning_rate": 4.947583612040134e-06,
+      "loss": 0.1985,
+      "step": 41400
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.420445600987573,
+      "learning_rate": 4.947541806020068e-06,
+      "loss": 0.2126,
+      "step": 41425
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.1764873951692207,
+      "learning_rate": 4.9475000000000005e-06,
+      "loss": 0.21,
+      "step": 41450
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.2261555787798102,
+      "learning_rate": 4.947458193979934e-06,
+      "loss": 0.21,
+      "step": 41475
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.7188400172880676,
+      "learning_rate": 4.947416387959867e-06,
+      "loss": 0.1842,
+      "step": 41500
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.0411835673772099,
+      "learning_rate": 4.9473745819397994e-06,
+      "loss": 0.1884,
+      "step": 41525
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.8882216836071882,
+      "learning_rate": 4.947332775919733e-06,
+      "loss": 0.1994,
+      "step": 41550
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.1029055988375176,
+      "learning_rate": 4.947290969899666e-06,
+      "loss": 0.1998,
+      "step": 41575
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.3852789220034836,
+      "learning_rate": 4.947249163879599e-06,
+      "loss": 0.2331,
+      "step": 41600
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.324731226538153,
+      "learning_rate": 4.947207357859532e-06,
+      "loss": 0.2266,
+      "step": 41625
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.4657036693086165,
+      "learning_rate": 4.9471655518394655e-06,
+      "loss": 0.215,
+      "step": 41650
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.527084020270202,
+      "learning_rate": 4.947123745819398e-06,
+      "loss": 0.1884,
+      "step": 41675
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.6846948970525009,
+      "learning_rate": 4.947081939799332e-06,
+      "loss": 0.2197,
+      "step": 41700
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.3851303237662287,
+      "learning_rate": 4.947040133779264e-06,
+      "loss": 0.2075,
+      "step": 41725
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.734382175640918,
+      "learning_rate": 4.946998327759198e-06,
+      "loss": 0.1901,
+      "step": 41750
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.8727000776687595,
+      "learning_rate": 4.946956521739131e-06,
+      "loss": 0.2099,
+      "step": 41775
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.3410587525533615,
+      "learning_rate": 4.946914715719064e-06,
+      "loss": 0.2144,
+      "step": 41800
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.232771457346267,
+      "learning_rate": 4.946872909698997e-06,
+      "loss": 0.2167,
+      "step": 41825
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.5134483112849184,
+      "learning_rate": 4.9468311036789304e-06,
+      "loss": 0.2034,
+      "step": 41850
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.2186748135314416,
+      "learning_rate": 4.946789297658863e-06,
+      "loss": 0.2016,
+      "step": 41875
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.5576513070546023,
+      "learning_rate": 4.946747491638797e-06,
+      "loss": 0.2147,
+      "step": 41900
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.350090298252575,
+      "learning_rate": 4.946705685618729e-06,
+      "loss": 0.2087,
+      "step": 41925
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.5693305584665407,
+      "learning_rate": 4.946663879598663e-06,
+      "loss": 0.2105,
+      "step": 41950
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.664226564042428,
+      "learning_rate": 4.946622073578596e-06,
+      "loss": 0.2092,
+      "step": 41975
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.5203247472068897,
+      "learning_rate": 4.946580267558529e-06,
+      "loss": 0.21,
+      "step": 42000
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.6732043500656888,
+      "learning_rate": 4.946538461538462e-06,
+      "loss": 0.2172,
+      "step": 42025
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.1170712647253107,
+      "learning_rate": 4.946496655518395e-06,
+      "loss": 0.2095,
+      "step": 42050
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.27085798758864,
+      "learning_rate": 4.946454849498328e-06,
+      "loss": 0.1996,
+      "step": 42075
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.121975214434564,
+      "learning_rate": 4.946413043478262e-06,
+      "loss": 0.1915,
+      "step": 42100
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.8454823137618726,
+      "learning_rate": 4.946371237458194e-06,
+      "loss": 0.2185,
+      "step": 42125
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.7165066242154234,
+      "learning_rate": 4.946329431438128e-06,
+      "loss": 0.2285,
+      "step": 42150
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.1935629947391357,
+      "learning_rate": 4.946287625418061e-06,
+      "loss": 0.2055,
+      "step": 42175
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.6263960825451074,
+      "learning_rate": 4.946245819397994e-06,
+      "loss": 0.1963,
+      "step": 42200
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.5830643088692502,
+      "learning_rate": 4.946204013377927e-06,
+      "loss": 0.1938,
+      "step": 42225
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.188351267211218,
+      "learning_rate": 4.946163879598663e-06,
+      "loss": 0.1972,
+      "step": 42250
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.5228388266516044,
+      "learning_rate": 4.946122073578596e-06,
+      "loss": 0.2146,
+      "step": 42275
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.2417331494377257,
+      "learning_rate": 4.946080267558529e-06,
+      "loss": 0.1901,
+      "step": 42300
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.2050316831091845,
+      "learning_rate": 4.9460384615384625e-06,
+      "loss": 0.1964,
+      "step": 42325
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7456359383137927,
+      "learning_rate": 4.945996655518395e-06,
+      "loss": 0.2114,
+      "step": 42350
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.1280620070130793,
+      "learning_rate": 4.945954849498329e-06,
+      "loss": 0.204,
+      "step": 42375
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.8154667054830138,
+      "learning_rate": 4.945913043478261e-06,
+      "loss": 0.2017,
+      "step": 42400
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.6384097778731235,
+      "learning_rate": 4.945871237458195e-06,
+      "loss": 0.2066,
+      "step": 42425
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.8156792744274872,
+      "learning_rate": 4.945829431438128e-06,
+      "loss": 0.1958,
+      "step": 42450
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.1599616876765457,
+      "learning_rate": 4.94578762541806e-06,
+      "loss": 0.2014,
+      "step": 42475
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.4247687110866212,
+      "learning_rate": 4.945745819397993e-06,
+      "loss": 0.2277,
+      "step": 42500
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.8355672996659873,
+      "learning_rate": 4.945704013377927e-06,
+      "loss": 0.2099,
+      "step": 42525
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.425217424300148,
+      "learning_rate": 4.94566220735786e-06,
+      "loss": 0.2016,
+      "step": 42550
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.7198208972430582,
+      "learning_rate": 4.945620401337793e-06,
+      "loss": 0.202,
+      "step": 42575
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.1552545949551052,
+      "learning_rate": 4.945578595317726e-06,
+      "loss": 0.221,
+      "step": 42600
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.4029219339538905,
+      "learning_rate": 4.945536789297659e-06,
+      "loss": 0.1987,
+      "step": 42625
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.3882183271546116,
+      "learning_rate": 4.945494983277593e-06,
+      "loss": 0.2081,
+      "step": 42650
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.9199663729648839,
+      "learning_rate": 4.945453177257525e-06,
+      "loss": 0.1973,
+      "step": 42675
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8901908136062112,
+      "learning_rate": 4.945411371237459e-06,
+      "loss": 0.2012,
+      "step": 42700
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.3827221258424418,
+      "learning_rate": 4.945369565217392e-06,
+      "loss": 0.2079,
+      "step": 42725
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.7149862183278763,
+      "learning_rate": 4.945327759197325e-06,
+      "loss": 0.1983,
+      "step": 42750
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.9186732501542165,
+      "learning_rate": 4.945285953177258e-06,
+      "loss": 0.2022,
+      "step": 42775
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.633325078705575,
+      "learning_rate": 4.945244147157191e-06,
+      "loss": 0.2094,
+      "step": 42800
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.397331371465462,
+      "learning_rate": 4.945202341137124e-06,
+      "loss": 0.2227,
+      "step": 42825
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.4918868776955483,
+      "learning_rate": 4.945160535117058e-06,
+      "loss": 0.214,
+      "step": 42850
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.595591918880962,
+      "learning_rate": 4.94511872909699e-06,
+      "loss": 0.2151,
+      "step": 42875
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.7222991867626707,
+      "learning_rate": 4.945076923076924e-06,
+      "loss": 0.2128,
+      "step": 42900
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.1691471473743391,
+      "learning_rate": 4.9450351170568566e-06,
+      "loss": 0.2075,
+      "step": 42925
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.4662345851129484,
+      "learning_rate": 4.94499331103679e-06,
+      "loss": 0.2125,
+      "step": 42950
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.5717643240108634,
+      "learning_rate": 4.944951505016723e-06,
+      "loss": 0.1802,
+      "step": 42975
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.3287909014275803,
+      "learning_rate": 4.944909698996656e-06,
+      "loss": 0.1952,
+      "step": 43000
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.097159089815716,
+      "learning_rate": 4.944867892976589e-06,
+      "loss": 0.202,
+      "step": 43025
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.9895232385883712,
+      "learning_rate": 4.944826086956523e-06,
+      "loss": 0.2109,
+      "step": 43050
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.5103690475667175,
+      "learning_rate": 4.944784280936455e-06,
+      "loss": 0.189,
+      "step": 43075
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.882631524404233,
+      "learning_rate": 4.944742474916389e-06,
+      "loss": 0.1967,
+      "step": 43100
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.3046423330005568,
+      "learning_rate": 4.9447006688963215e-06,
+      "loss": 0.1788,
+      "step": 43125
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.693461277845761,
+      "learning_rate": 4.944658862876255e-06,
+      "loss": 0.1978,
+      "step": 43150
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.453539084307599,
+      "learning_rate": 4.944617056856188e-06,
+      "loss": 0.2291,
+      "step": 43175
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.7794757369908873,
+      "learning_rate": 4.9445752508361205e-06,
+      "loss": 0.2148,
+      "step": 43200
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.7997181495063956,
+      "learning_rate": 4.944533444816054e-06,
+      "loss": 0.1912,
+      "step": 43225
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.103310345158487,
+      "learning_rate": 4.94449331103679e-06,
+      "loss": 0.2029,
+      "step": 43250
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.3283148483923,
+      "learning_rate": 4.944451505016723e-06,
+      "loss": 0.2016,
+      "step": 43275
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.5231191363195111,
+      "learning_rate": 4.944409698996656e-06,
+      "loss": 0.194,
+      "step": 43300
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.727077984512469,
+      "learning_rate": 4.94436789297659e-06,
+      "loss": 0.2031,
+      "step": 43325
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.453989318654602,
+      "learning_rate": 4.944326086956522e-06,
+      "loss": 0.2083,
+      "step": 43350
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.5469548856275384,
+      "learning_rate": 4.944284280936456e-06,
+      "loss": 0.2176,
+      "step": 43375
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.160825934896527,
+      "learning_rate": 4.944242474916388e-06,
+      "loss": 0.2022,
+      "step": 43400
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.5719087440594064,
+      "learning_rate": 4.944200668896321e-06,
+      "loss": 0.2125,
+      "step": 43425
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.3656354570889466,
+      "learning_rate": 4.944158862876254e-06,
+      "loss": 0.2117,
+      "step": 43450
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.5367941234175355,
+      "learning_rate": 4.9441170568561875e-06,
+      "loss": 0.2176,
+      "step": 43475
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.4417364583297174,
+      "learning_rate": 4.94407525083612e-06,
+      "loss": 0.1879,
+      "step": 43500
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.2630478840085972,
+      "learning_rate": 4.944033444816054e-06,
+      "loss": 0.2,
+      "step": 43525
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.112072451582671,
+      "learning_rate": 4.9439916387959865e-06,
+      "loss": 0.1968,
+      "step": 43550
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.7713642185306702,
+      "learning_rate": 4.94394983277592e-06,
+      "loss": 0.2267,
+      "step": 43575
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.2056972754693054,
+      "learning_rate": 4.943908026755853e-06,
+      "loss": 0.2014,
+      "step": 43600
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.092655557657814,
+      "learning_rate": 4.943866220735786e-06,
+      "loss": 0.2009,
+      "step": 43625
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.7225505908943808,
+      "learning_rate": 4.94382441471572e-06,
+      "loss": 0.2268,
+      "step": 43650
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.5368215742705937,
+      "learning_rate": 4.9437826086956525e-06,
+      "loss": 0.2055,
+      "step": 43675
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.3197247012408562,
+      "learning_rate": 4.943740802675586e-06,
+      "loss": 0.2343,
+      "step": 43700
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.3772135831744492,
+      "learning_rate": 4.943698996655519e-06,
+      "loss": 0.1976,
+      "step": 43725
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.4326058066713432,
+      "learning_rate": 4.943657190635452e-06,
+      "loss": 0.2031,
+      "step": 43750
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.5888807750342064,
+      "learning_rate": 4.943615384615385e-06,
+      "loss": 0.2074,
+      "step": 43775
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.5396007104076448,
+      "learning_rate": 4.9435735785953185e-06,
+      "loss": 0.2067,
+      "step": 43800
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.4011127551195726,
+      "learning_rate": 4.943531772575251e-06,
+      "loss": 0.1787,
+      "step": 43825
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.833780754470428,
+      "learning_rate": 4.943489966555185e-06,
+      "loss": 0.1943,
+      "step": 43850
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.4268589028719112,
+      "learning_rate": 4.9434481605351175e-06,
+      "loss": 0.184,
+      "step": 43875
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.1879663087707617,
+      "learning_rate": 4.943406354515051e-06,
+      "loss": 0.2,
+      "step": 43900
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.4837138515357116,
+      "learning_rate": 4.943364548494984e-06,
+      "loss": 0.1938,
+      "step": 43925
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.302090890735732,
+      "learning_rate": 4.943322742474917e-06,
+      "loss": 0.2179,
+      "step": 43950
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.861136682417232,
+      "learning_rate": 4.94328093645485e-06,
+      "loss": 0.2182,
+      "step": 43975
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.335286311587073,
+      "learning_rate": 4.9432391304347835e-06,
+      "loss": 0.2214,
+      "step": 44000
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.6769155432405607,
+      "learning_rate": 4.943197324414716e-06,
+      "loss": 0.2042,
+      "step": 44025
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.200106902410679,
+      "learning_rate": 4.94315551839465e-06,
+      "loss": 0.1987,
+      "step": 44050
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.7563502007021174,
+      "learning_rate": 4.9431137123745825e-06,
+      "loss": 0.1951,
+      "step": 44075
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.5530867625540488,
+      "learning_rate": 4.943071906354516e-06,
+      "loss": 0.2139,
+      "step": 44100
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.3534943489075693,
+      "learning_rate": 4.943030100334449e-06,
+      "loss": 0.1773,
+      "step": 44125
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.4957898910507397,
+      "learning_rate": 4.942988294314381e-06,
+      "loss": 0.228,
+      "step": 44150
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.8785662557550045,
+      "learning_rate": 4.942946488294315e-06,
+      "loss": 0.2081,
+      "step": 44175
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.712164070763034,
+      "learning_rate": 4.942904682274248e-06,
+      "loss": 0.2042,
+      "step": 44200
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.52945812067655,
+      "learning_rate": 4.942862876254181e-06,
+      "loss": 0.2277,
+      "step": 44225
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.3722433996508723,
+      "learning_rate": 4.942822742474917e-06,
+      "loss": 0.2299,
+      "step": 44250
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.5393189925564164,
+      "learning_rate": 4.94278093645485e-06,
+      "loss": 0.2175,
+      "step": 44275
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.5822179394649376,
+      "learning_rate": 4.942739130434783e-06,
+      "loss": 0.2038,
+      "step": 44300
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.3792366307810666,
+      "learning_rate": 4.942697324414717e-06,
+      "loss": 0.2047,
+      "step": 44325
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.70816553270609,
+      "learning_rate": 4.942655518394649e-06,
+      "loss": 0.2217,
+      "step": 44350
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.2792536160078478,
+      "learning_rate": 4.942613712374582e-06,
+      "loss": 0.2131,
+      "step": 44375
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.4575673021789084,
+      "learning_rate": 4.942571906354515e-06,
+      "loss": 0.2131,
+      "step": 44400
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.8400518563065749,
+      "learning_rate": 4.9425301003344485e-06,
+      "loss": 0.2037,
+      "step": 44425
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.5635883512841116,
+      "learning_rate": 4.942488294314381e-06,
+      "loss": 0.2091,
+      "step": 44450
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.0007327369281773,
+      "learning_rate": 4.942446488294315e-06,
+      "loss": 0.2127,
+      "step": 44475
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.119742302055427,
+      "learning_rate": 4.942404682274247e-06,
+      "loss": 0.2012,
+      "step": 44500
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.9344147417555144,
+      "learning_rate": 4.942362876254181e-06,
+      "loss": 0.2013,
+      "step": 44525
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.328539603876774,
+      "learning_rate": 4.942321070234114e-06,
+      "loss": 0.1986,
+      "step": 44550
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.7040707858619548,
+      "learning_rate": 4.942279264214047e-06,
+      "loss": 0.1954,
+      "step": 44575
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.4008583236352887,
+      "learning_rate": 4.94223745819398e-06,
+      "loss": 0.2226,
+      "step": 44600
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.148723618960234,
+      "learning_rate": 4.9421956521739134e-06,
+      "loss": 0.2067,
+      "step": 44625
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.4833158950358802,
+      "learning_rate": 4.942153846153846e-06,
+      "loss": 0.2012,
+      "step": 44650
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.8801569799377404,
+      "learning_rate": 4.94211204013378e-06,
+      "loss": 0.204,
+      "step": 44675
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.6972529337434588,
+      "learning_rate": 4.942070234113712e-06,
+      "loss": 0.2183,
+      "step": 44700
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.7424024925861386,
+      "learning_rate": 4.942028428093646e-06,
+      "loss": 0.196,
+      "step": 44725
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.419744820585979,
+      "learning_rate": 4.9419866220735795e-06,
+      "loss": 0.2205,
+      "step": 44750
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.773705854022055,
+      "learning_rate": 4.941944816053512e-06,
+      "loss": 0.1951,
+      "step": 44775
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.6575512331545263,
+      "learning_rate": 4.941903010033446e-06,
+      "loss": 0.1998,
+      "step": 44800
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.8632058992042098,
+      "learning_rate": 4.941861204013378e-06,
+      "loss": 0.2272,
+      "step": 44825
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.8833068240202187,
+      "learning_rate": 4.941819397993312e-06,
+      "loss": 0.23,
+      "step": 44850
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.5386804945912553,
+      "learning_rate": 4.941777591973245e-06,
+      "loss": 0.2026,
+      "step": 44875
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.1855200203299385,
+      "learning_rate": 4.941735785953178e-06,
+      "loss": 0.2161,
+      "step": 44900
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.9564640886071363,
+      "learning_rate": 4.941693979933111e-06,
+      "loss": 0.1789,
+      "step": 44925
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.340358116387867,
+      "learning_rate": 4.9416521739130444e-06,
+      "loss": 0.1937,
+      "step": 44950
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.1872850428854087,
+      "learning_rate": 4.941610367892977e-06,
+      "loss": 0.2088,
+      "step": 44975
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.3941237415424022,
+      "learning_rate": 4.941568561872911e-06,
+      "loss": 0.215,
+      "step": 45000
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.2422150166796995,
+      "learning_rate": 4.941526755852843e-06,
+      "loss": 0.1978,
+      "step": 45025
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.2834289450493046,
+      "learning_rate": 4.941484949832777e-06,
+      "loss": 0.2019,
+      "step": 45050
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.3887225318643923,
+      "learning_rate": 4.941443143812709e-06,
+      "loss": 0.2005,
+      "step": 45075
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.5957356966876162,
+      "learning_rate": 4.941401337792642e-06,
+      "loss": 0.2116,
+      "step": 45100
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.7998947945044745,
+      "learning_rate": 4.941359531772575e-06,
+      "loss": 0.2113,
+      "step": 45125
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.2863221238150597,
+      "learning_rate": 4.9413177257525086e-06,
+      "loss": 0.1909,
+      "step": 45150
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.291509095965881,
+      "learning_rate": 4.941275919732442e-06,
+      "loss": 0.1993,
+      "step": 45175
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.1958530909710876,
+      "learning_rate": 4.941234113712375e-06,
+      "loss": 0.2115,
+      "step": 45200
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.5439304960130933,
+      "learning_rate": 4.941192307692308e-06,
+      "loss": 0.1959,
+      "step": 45225
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.9055756132780866,
+      "learning_rate": 4.941152173913044e-06,
+      "loss": 0.2022,
+      "step": 45250
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.2942694636126806,
+      "learning_rate": 4.941110367892977e-06,
+      "loss": 0.1945,
+      "step": 45275
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.6378120736690176,
+      "learning_rate": 4.94106856187291e-06,
+      "loss": 0.2001,
+      "step": 45300
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.7309152118782125,
+      "learning_rate": 4.941026755852843e-06,
+      "loss": 0.189,
+      "step": 45325
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.5926592940812094,
+      "learning_rate": 4.940984949832776e-06,
+      "loss": 0.1886,
+      "step": 45350
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.4814435621208106,
+      "learning_rate": 4.940943143812709e-06,
+      "loss": 0.2175,
+      "step": 45375
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.3340113593410514,
+      "learning_rate": 4.940901337792642e-06,
+      "loss": 0.2064,
+      "step": 45400
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.7179296631796945,
+      "learning_rate": 4.940859531772576e-06,
+      "loss": 0.2027,
+      "step": 45425
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.3525854738210719,
+      "learning_rate": 4.940817725752508e-06,
+      "loss": 0.2069,
+      "step": 45450
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.1243049994361427,
+      "learning_rate": 4.940775919732442e-06,
+      "loss": 0.1942,
+      "step": 45475
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.2883069022850577,
+      "learning_rate": 4.9407341137123746e-06,
+      "loss": 0.1776,
+      "step": 45500
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.3987378560793549,
+      "learning_rate": 4.940692307692308e-06,
+      "loss": 0.2085,
+      "step": 45525
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.8141337973143972,
+      "learning_rate": 4.940650501672241e-06,
+      "loss": 0.2077,
+      "step": 45550
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.1709295579225634,
+      "learning_rate": 4.940608695652174e-06,
+      "loss": 0.1949,
+      "step": 45575
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.5535520468286816,
+      "learning_rate": 4.940566889632107e-06,
+      "loss": 0.1988,
+      "step": 45600
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.9291991595641431,
+      "learning_rate": 4.940525083612041e-06,
+      "loss": 0.2142,
+      "step": 45625
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.944908850889749,
+      "learning_rate": 4.940483277591973e-06,
+      "loss": 0.2189,
+      "step": 45650
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.8924849523796625,
+      "learning_rate": 4.940441471571907e-06,
+      "loss": 0.2042,
+      "step": 45675
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.0381586631748612,
+      "learning_rate": 4.9403996655518395e-06,
+      "loss": 0.1973,
+      "step": 45700
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.596229546211035,
+      "learning_rate": 4.940357859531773e-06,
+      "loss": 0.1971,
+      "step": 45725
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.478086237165827,
+      "learning_rate": 4.940316053511706e-06,
+      "loss": 0.2172,
+      "step": 45750
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.559672032174473,
+      "learning_rate": 4.940274247491639e-06,
+      "loss": 0.2215,
+      "step": 45775
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.2609063742885178,
+      "learning_rate": 4.940232441471572e-06,
+      "loss": 0.1815,
+      "step": 45800
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.2924032797525853,
+      "learning_rate": 4.9401906354515056e-06,
+      "loss": 0.2064,
+      "step": 45825
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.4174361862563374,
+      "learning_rate": 4.940148829431439e-06,
+      "loss": 0.217,
+      "step": 45850
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.373477111385322,
+      "learning_rate": 4.940107023411372e-06,
+      "loss": 0.1953,
+      "step": 45875
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.8558451130861622,
+      "learning_rate": 4.940065217391305e-06,
+      "loss": 0.2054,
+      "step": 45900
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.6990765889691408,
+      "learning_rate": 4.940023411371238e-06,
+      "loss": 0.1899,
+      "step": 45925
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.8060030463753445,
+      "learning_rate": 4.939981605351172e-06,
+      "loss": 0.2202,
+      "step": 45950
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.0036006697315363,
+      "learning_rate": 4.939939799331104e-06,
+      "loss": 0.1892,
+      "step": 45975
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.5889893762168747,
+      "learning_rate": 4.939897993311037e-06,
+      "loss": 0.2014,
+      "step": 46000
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.2885407324111298,
+      "learning_rate": 4.93985618729097e-06,
+      "loss": 0.2022,
+      "step": 46025
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.4911043166689495,
+      "learning_rate": 4.939814381270903e-06,
+      "loss": 0.219,
+      "step": 46050
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.2755317856393762,
+      "learning_rate": 4.939772575250836e-06,
+      "loss": 0.2054,
+      "step": 46075
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.994187232620831,
+      "learning_rate": 4.9397307692307695e-06,
+      "loss": 0.1891,
+      "step": 46100
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.5475701584616484,
+      "learning_rate": 4.939688963210702e-06,
+      "loss": 0.2022,
+      "step": 46125
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.4436373558259408,
+      "learning_rate": 4.939647157190636e-06,
+      "loss": 0.197,
+      "step": 46150
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.7292220715855722,
+      "learning_rate": 4.9396053511705684e-06,
+      "loss": 0.2127,
+      "step": 46175
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.00863798888671,
+      "learning_rate": 4.939563545150502e-06,
+      "loss": 0.1874,
+      "step": 46200
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.575347210475048,
+      "learning_rate": 4.939521739130435e-06,
+      "loss": 0.2142,
+      "step": 46225
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.8865538136561686,
+      "learning_rate": 4.9394816053511705e-06,
+      "loss": 0.194,
+      "step": 46250
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.3867064113440684,
+      "learning_rate": 4.939439799331104e-06,
+      "loss": 0.2238,
+      "step": 46275
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.6636489722849195,
+      "learning_rate": 4.939397993311037e-06,
+      "loss": 0.2124,
+      "step": 46300
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.796035846087043,
+      "learning_rate": 4.93935618729097e-06,
+      "loss": 0.2086,
+      "step": 46325
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.6152861037902768,
+      "learning_rate": 4.939314381270903e-06,
+      "loss": 0.2194,
+      "step": 46350
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.872021389069065,
+      "learning_rate": 4.9392725752508365e-06,
+      "loss": 0.1983,
+      "step": 46375
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.3627173247663622,
+      "learning_rate": 4.939230769230769e-06,
+      "loss": 0.2228,
+      "step": 46400
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.314259549099962,
+      "learning_rate": 4.939188963210703e-06,
+      "loss": 0.1979,
+      "step": 46425
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.6032460211094444,
+      "learning_rate": 4.9391471571906355e-06,
+      "loss": 0.2278,
+      "step": 46450
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.775553525039903,
+      "learning_rate": 4.939105351170569e-06,
+      "loss": 0.2044,
+      "step": 46475
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.800363727354035,
+      "learning_rate": 4.939063545150502e-06,
+      "loss": 0.2143,
+      "step": 46500
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.2465956627042085,
+      "learning_rate": 4.939021739130435e-06,
+      "loss": 0.2043,
+      "step": 46525
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.8944500939180655,
+      "learning_rate": 4.938979933110368e-06,
+      "loss": 0.2063,
+      "step": 46550
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.5117363359679743,
+      "learning_rate": 4.9389381270903015e-06,
+      "loss": 0.2122,
+      "step": 46575
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.5887958169703909,
+      "learning_rate": 4.938896321070234e-06,
+      "loss": 0.221,
+      "step": 46600
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.648221177617564,
+      "learning_rate": 4.938854515050168e-06,
+      "loss": 0.1771,
+      "step": 46625
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.55324428745445,
+      "learning_rate": 4.9388127090301005e-06,
+      "loss": 0.1883,
+      "step": 46650
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.7723178618451372,
+      "learning_rate": 4.938770903010034e-06,
+      "loss": 0.1927,
+      "step": 46675
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.153448369196902,
+      "learning_rate": 4.938729096989967e-06,
+      "loss": 0.2044,
+      "step": 46700
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.1268320701701797,
+      "learning_rate": 4.9386872909699e-06,
+      "loss": 0.2428,
+      "step": 46725
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.3631634906977346,
+      "learning_rate": 4.938645484949833e-06,
+      "loss": 0.1857,
+      "step": 46750
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.6168270762112813,
+      "learning_rate": 4.9386036789297665e-06,
+      "loss": 0.2011,
+      "step": 46775
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.6696110343047001,
+      "learning_rate": 4.938561872909699e-06,
+      "loss": 0.2204,
+      "step": 46800
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.192494551736515,
+      "learning_rate": 4.938520066889633e-06,
+      "loss": 0.2038,
+      "step": 46825
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.7077939692501043,
+      "learning_rate": 4.9384782608695654e-06,
+      "loss": 0.1955,
+      "step": 46850
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.8191368453499213,
+      "learning_rate": 4.938436454849499e-06,
+      "loss": 0.208,
+      "step": 46875
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.5432423506062156,
+      "learning_rate": 4.938394648829432e-06,
+      "loss": 0.2058,
+      "step": 46900
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.9468989423937535,
+      "learning_rate": 4.938352842809365e-06,
+      "loss": 0.1768,
+      "step": 46925
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.7161970238767805,
+      "learning_rate": 4.938311036789298e-06,
+      "loss": 0.2131,
+      "step": 46950
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.5485071588534483,
+      "learning_rate": 4.938269230769231e-06,
+      "loss": 0.2218,
+      "step": 46975
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.53263925456021,
+      "learning_rate": 4.938227424749164e-06,
+      "loss": 0.2235,
+      "step": 47000
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.8192841664528814,
+      "learning_rate": 4.938185618729097e-06,
+      "loss": 0.1992,
+      "step": 47025
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.115023350287966,
+      "learning_rate": 4.93814381270903e-06,
+      "loss": 0.2213,
+      "step": 47050
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.5380163293793399,
+      "learning_rate": 4.938102006688963e-06,
+      "loss": 0.1992,
+      "step": 47075
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.4449128177392452,
+      "learning_rate": 4.938060200668897e-06,
+      "loss": 0.2181,
+      "step": 47100
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.8029578697927904,
+      "learning_rate": 4.938018394648829e-06,
+      "loss": 0.1921,
+      "step": 47125
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.0395632871911147,
+      "learning_rate": 4.937976588628763e-06,
+      "loss": 0.1895,
+      "step": 47150
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.3878968532309697,
+      "learning_rate": 4.937934782608696e-06,
+      "loss": 0.1842,
+      "step": 47175
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.3434122603027006,
+      "learning_rate": 4.937892976588629e-06,
+      "loss": 0.2056,
+      "step": 47200
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.3490671768973181,
+      "learning_rate": 4.937851170568562e-06,
+      "loss": 0.1866,
+      "step": 47225
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.613883740225785,
+      "learning_rate": 4.937811036789298e-06,
+      "loss": 0.1979,
+      "step": 47250
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.7926615672094046,
+      "learning_rate": 4.937769230769231e-06,
+      "loss": 0.1977,
+      "step": 47275
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.484911009082714,
+      "learning_rate": 4.937727424749164e-06,
+      "loss": 0.2022,
+      "step": 47300
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.9218949249621484,
+      "learning_rate": 4.9376856187290975e-06,
+      "loss": 0.2221,
+      "step": 47325
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.4316456230383503,
+      "learning_rate": 4.937645484949833e-06,
+      "loss": 0.1902,
+      "step": 47350
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.7052129122552293,
+      "learning_rate": 4.937603678929767e-06,
+      "loss": 0.2221,
+      "step": 47375
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.2637211421190688,
+      "learning_rate": 4.9375618729096995e-06,
+      "loss": 0.2058,
+      "step": 47400
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.8451450491794519,
+      "learning_rate": 4.937520066889632e-06,
+      "loss": 0.1993,
+      "step": 47425
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.9283921636952341,
+      "learning_rate": 4.937478260869566e-06,
+      "loss": 0.2055,
+      "step": 47450
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.8414663077620397,
+      "learning_rate": 4.9374364548494985e-06,
+      "loss": 0.2203,
+      "step": 47475
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.0753580914227365,
+      "learning_rate": 4.937394648829432e-06,
+      "loss": 0.213,
+      "step": 47500
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.2078240932003506,
+      "learning_rate": 4.937352842809365e-06,
+      "loss": 0.2136,
+      "step": 47525
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.7304545694227498,
+      "learning_rate": 4.937311036789298e-06,
+      "loss": 0.215,
+      "step": 47550
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.6773895264603955,
+      "learning_rate": 4.937269230769231e-06,
+      "loss": 0.184,
+      "step": 47575
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.6133216723159176,
+      "learning_rate": 4.9372274247491645e-06,
+      "loss": 0.2233,
+      "step": 47600
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.3083125789789483,
+      "learning_rate": 4.937185618729097e-06,
+      "loss": 0.2224,
+      "step": 47625
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.6499407481703168,
+      "learning_rate": 4.937143812709031e-06,
+      "loss": 0.2013,
+      "step": 47650
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.9096100207893898,
+      "learning_rate": 4.9371020066889635e-06,
+      "loss": 0.1966,
+      "step": 47675
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.0102219012043174,
+      "learning_rate": 4.937060200668897e-06,
+      "loss": 0.2247,
+      "step": 47700
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.8440109791151151,
+      "learning_rate": 4.93701839464883e-06,
+      "loss": 0.1832,
+      "step": 47725
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.6405657863329357,
+      "learning_rate": 4.936976588628763e-06,
+      "loss": 0.2021,
+      "step": 47750
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.5352039277718714,
+      "learning_rate": 4.936934782608696e-06,
+      "loss": 0.2106,
+      "step": 47775
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.218168105886781,
+      "learning_rate": 4.9368929765886295e-06,
+      "loss": 0.2028,
+      "step": 47800
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.9303573799655196,
+      "learning_rate": 4.936851170568562e-06,
+      "loss": 0.208,
+      "step": 47825
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.3720528426394387,
+      "learning_rate": 4.936809364548496e-06,
+      "loss": 0.213,
+      "step": 47850
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.330231745796533,
+      "learning_rate": 4.9367675585284284e-06,
+      "loss": 0.194,
+      "step": 47875
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.6102889821188229,
+      "learning_rate": 4.936725752508362e-06,
+      "loss": 0.2059,
+      "step": 47900
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.7900179616799938,
+      "learning_rate": 4.936683946488295e-06,
+      "loss": 0.2099,
+      "step": 47925
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.876167524982635,
+      "learning_rate": 4.936642140468228e-06,
+      "loss": 0.2108,
+      "step": 47950
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.3715764517785491,
+      "learning_rate": 4.936600334448161e-06,
+      "loss": 0.2118,
+      "step": 47975
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.242091850621187,
+      "learning_rate": 4.9365585284280945e-06,
+      "loss": 0.2091,
+      "step": 48000
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 2.381253912271751,
+      "learning_rate": 4.936516722408027e-06,
+      "loss": 0.1934,
+      "step": 48025
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.412172971145708,
+      "learning_rate": 4.936474916387961e-06,
+      "loss": 0.2054,
+      "step": 48050
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.5116219256655863,
+      "learning_rate": 4.936433110367893e-06,
+      "loss": 0.2155,
+      "step": 48075
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.2887142644122933,
+      "learning_rate": 4.936391304347827e-06,
+      "loss": 0.2091,
+      "step": 48100
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.6820263408725942,
+      "learning_rate": 4.93634949832776e-06,
+      "loss": 0.1912,
+      "step": 48125
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.5815843596491972,
+      "learning_rate": 4.936307692307692e-06,
+      "loss": 0.2055,
+      "step": 48150
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.6081837772093868,
+      "learning_rate": 4.936265886287626e-06,
+      "loss": 0.2117,
+      "step": 48175
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.4467885553172026,
+      "learning_rate": 4.936224080267559e-06,
+      "loss": 0.216,
+      "step": 48200
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.1519428579352176,
+      "learning_rate": 4.936182274247492e-06,
+      "loss": 0.2017,
+      "step": 48225
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.5300606843300384,
+      "learning_rate": 4.936140468227425e-06,
+      "loss": 0.205,
+      "step": 48250
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.720404121032897,
+      "learning_rate": 4.936098662207358e-06,
+      "loss": 0.2007,
+      "step": 48275
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.6425840844489246,
+      "learning_rate": 4.936056856187291e-06,
+      "loss": 0.1853,
+      "step": 48300
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.459659103020749,
+      "learning_rate": 4.936015050167225e-06,
+      "loss": 0.2116,
+      "step": 48325
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.3712585119025817,
+      "learning_rate": 4.935973244147157e-06,
+      "loss": 0.2106,
+      "step": 48350
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.2759914401336694,
+      "learning_rate": 4.935931438127091e-06,
+      "loss": 0.2099,
+      "step": 48375
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.1003315934242315,
+      "learning_rate": 4.935889632107024e-06,
+      "loss": 0.2006,
+      "step": 48400
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.1531531222871791,
+      "learning_rate": 4.935847826086957e-06,
+      "loss": 0.1993,
+      "step": 48425
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.9522070688750959,
+      "learning_rate": 4.93580602006689e-06,
+      "loss": 0.2029,
+      "step": 48450
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.9647196482698353,
+      "learning_rate": 4.935764214046823e-06,
+      "loss": 0.1777,
+      "step": 48475
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.357874217165264,
+      "learning_rate": 4.935722408026756e-06,
+      "loss": 0.202,
+      "step": 48500
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.706086540465037,
+      "learning_rate": 4.93568060200669e-06,
+      "loss": 0.2099,
+      "step": 48525
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.368230675138682,
+      "learning_rate": 4.935638795986622e-06,
+      "loss": 0.1893,
+      "step": 48550
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.1989144444420323,
+      "learning_rate": 4.935596989966556e-06,
+      "loss": 0.234,
+      "step": 48575
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.3439068076959173,
+      "learning_rate": 4.9355551839464886e-06,
+      "loss": 0.1919,
+      "step": 48600
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.5347802892044207,
+      "learning_rate": 4.935513377926422e-06,
+      "loss": 0.2087,
+      "step": 48625
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.3853060269687627,
+      "learning_rate": 4.935471571906355e-06,
+      "loss": 0.2153,
+      "step": 48650
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.2705654843241438,
+      "learning_rate": 4.935429765886288e-06,
+      "loss": 0.1942,
+      "step": 48675
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.507043182236023,
+      "learning_rate": 4.935387959866221e-06,
+      "loss": 0.2052,
+      "step": 48700
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8852640991515132,
+      "learning_rate": 4.935346153846155e-06,
+      "loss": 0.2126,
+      "step": 48725
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.2880272534231267,
+      "learning_rate": 4.935304347826087e-06,
+      "loss": 0.1823,
+      "step": 48750
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.3419848634582663,
+      "learning_rate": 4.935262541806021e-06,
+      "loss": 0.2085,
+      "step": 48775
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.2454273911509375,
+      "learning_rate": 4.9352207357859535e-06,
+      "loss": 0.198,
+      "step": 48800
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.1964741196519708,
+      "learning_rate": 4.935178929765886e-06,
+      "loss": 0.194,
+      "step": 48825
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.482373609512503,
+      "learning_rate": 4.93513712374582e-06,
+      "loss": 0.1923,
+      "step": 48850
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.4343822110286697,
+      "learning_rate": 4.9350953177257525e-06,
+      "loss": 0.201,
+      "step": 48875
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.9110162286931605,
+      "learning_rate": 4.935053511705686e-06,
+      "loss": 0.2177,
+      "step": 48900
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.4055468589424738,
+      "learning_rate": 4.935011705685619e-06,
+      "loss": 0.1975,
+      "step": 48925
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.1095321871083568,
+      "learning_rate": 4.934969899665552e-06,
+      "loss": 0.2038,
+      "step": 48950
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.6489177988410992,
+      "learning_rate": 4.934928093645485e-06,
+      "loss": 0.2045,
+      "step": 48975
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.0969848442231984,
+      "learning_rate": 4.9348862876254185e-06,
+      "loss": 0.1798,
+      "step": 49000
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.431163706761367,
+      "learning_rate": 4.934844481605351e-06,
+      "loss": 0.2148,
+      "step": 49025
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.1269539448904287,
+      "learning_rate": 4.934802675585285e-06,
+      "loss": 0.1913,
+      "step": 49050
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 2.009600750564663,
+      "learning_rate": 4.9347608695652175e-06,
+      "loss": 0.1902,
+      "step": 49075
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.6341443792861832,
+      "learning_rate": 4.934719063545151e-06,
+      "loss": 0.2069,
+      "step": 49100
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.4131602535107552,
+      "learning_rate": 4.934677257525084e-06,
+      "loss": 0.1899,
+      "step": 49125
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.8377130239703243,
+      "learning_rate": 4.934635451505017e-06,
+      "loss": 0.2038,
+      "step": 49150
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.9368805923358994,
+      "learning_rate": 4.93459364548495e-06,
+      "loss": 0.2089,
+      "step": 49175
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.6406415730986539,
+      "learning_rate": 4.9345518394648835e-06,
+      "loss": 0.2023,
+      "step": 49200
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.5142583337522033,
+      "learning_rate": 4.934510033444816e-06,
+      "loss": 0.2124,
+      "step": 49225
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.4175403686258068,
+      "learning_rate": 4.93446822742475e-06,
+      "loss": 0.1889,
+      "step": 49250
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.680924590381479,
+      "learning_rate": 4.9344264214046824e-06,
+      "loss": 0.18,
+      "step": 49275
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.4182165052837803,
+      "learning_rate": 4.934384615384616e-06,
+      "loss": 0.199,
+      "step": 49300
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.5324006988373318,
+      "learning_rate": 4.934342809364549e-06,
+      "loss": 0.2107,
+      "step": 49325
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.8984134230654337,
+      "learning_rate": 4.9343026755852845e-06,
+      "loss": 0.2195,
+      "step": 49350
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.8449500063553705,
+      "learning_rate": 4.934260869565218e-06,
+      "loss": 0.1778,
+      "step": 49375
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.9070885133107418,
+      "learning_rate": 4.934219063545151e-06,
+      "loss": 0.2199,
+      "step": 49400
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.528503320386569,
+      "learning_rate": 4.934177257525084e-06,
+      "loss": 0.2003,
+      "step": 49425
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.6667344856090824,
+      "learning_rate": 4.934135451505017e-06,
+      "loss": 0.2131,
+      "step": 49450
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.4152939504446955,
+      "learning_rate": 4.9340936454849505e-06,
+      "loss": 0.1976,
+      "step": 49475
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.2226774657729198,
+      "learning_rate": 4.934051839464883e-06,
+      "loss": 0.194,
+      "step": 49500
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.1414769595580594,
+      "learning_rate": 4.934010033444817e-06,
+      "loss": 0.2142,
+      "step": 49525
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.3683179118744992,
+      "learning_rate": 4.9339682274247495e-06,
+      "loss": 0.209,
+      "step": 49550
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8797811276296436,
+      "learning_rate": 4.933926421404683e-06,
+      "loss": 0.1757,
+      "step": 49575
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.6240088021552774,
+      "learning_rate": 4.933884615384616e-06,
+      "loss": 0.2065,
+      "step": 49600
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.3496244482020072,
+      "learning_rate": 4.933842809364549e-06,
+      "loss": 0.1884,
+      "step": 49625
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.9065579862207658,
+      "learning_rate": 4.933801003344482e-06,
+      "loss": 0.2219,
+      "step": 49650
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.6823942734846893,
+      "learning_rate": 4.9337591973244155e-06,
+      "loss": 0.1922,
+      "step": 49675
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.610332780860586,
+      "learning_rate": 4.933717391304348e-06,
+      "loss": 0.2257,
+      "step": 49700
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.4749993288326266,
+      "learning_rate": 4.933675585284282e-06,
+      "loss": 0.2057,
+      "step": 49725
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.1293208475215297,
+      "learning_rate": 4.9336337792642145e-06,
+      "loss": 0.2238,
+      "step": 49750
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.9704102752018713,
+      "learning_rate": 4.933591973244147e-06,
+      "loss": 0.1962,
+      "step": 49775
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.3546928779201595,
+      "learning_rate": 4.933550167224081e-06,
+      "loss": 0.1999,
+      "step": 49800
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.647795435094045,
+      "learning_rate": 4.933508361204013e-06,
+      "loss": 0.187,
+      "step": 49825
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.1719791408575768,
+      "learning_rate": 4.933466555183947e-06,
+      "loss": 0.1982,
+      "step": 49850
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.410677691655115,
+      "learning_rate": 4.93342474916388e-06,
+      "loss": 0.2037,
+      "step": 49875
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.9028353705603924,
+      "learning_rate": 4.933382943143813e-06,
+      "loss": 0.1955,
+      "step": 49900
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.7600379755892537,
+      "learning_rate": 4.933341137123746e-06,
+      "loss": 0.2016,
+      "step": 49925
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.340426249993996,
+      "learning_rate": 4.9332993311036794e-06,
+      "loss": 0.213,
+      "step": 49950
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.7078496902487155,
+      "learning_rate": 4.933257525083612e-06,
+      "loss": 0.2066,
+      "step": 49975
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 2.3783688063483632,
+      "learning_rate": 4.933215719063546e-06,
+      "loss": 0.2034,
+      "step": 50000
+    },
+    {
+      "epoch": 1.97,
+      "eval_loss": 0.49365234375,
+      "eval_runtime": 11498.4763,
+      "eval_samples_per_second": 0.823,
+      "eval_steps_per_second": 0.051,
+      "eval_wer": 0.11952447597445426,
+      "step": 50000
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.9977781030077004,
+      "learning_rate": 4.933173913043478e-06,
+      "loss": 0.1802,
+      "step": 50025
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.5811010265422427,
+      "learning_rate": 4.933132107023412e-06,
+      "loss": 0.2066,
+      "step": 50050
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.8367235353607692,
+      "learning_rate": 4.933090301003345e-06,
+      "loss": 0.2188,
+      "step": 50075
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.4262908481666672,
+      "learning_rate": 4.933048494983278e-06,
+      "loss": 0.1738,
+      "step": 50100
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.8182130900532735,
+      "learning_rate": 4.933006688963211e-06,
+      "loss": 0.2088,
+      "step": 50125
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.0473337182584366,
+      "learning_rate": 4.932964882943144e-06,
+      "loss": 0.2347,
+      "step": 50150
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.2163618200194586,
+      "learning_rate": 4.932923076923077e-06,
+      "loss": 0.2046,
+      "step": 50175
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.7969452660629082,
+      "learning_rate": 4.932881270903011e-06,
+      "loss": 0.1909,
+      "step": 50200
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.8467324578026039,
+      "learning_rate": 4.932839464882943e-06,
+      "loss": 0.2108,
+      "step": 50225
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.217318536045435,
+      "learning_rate": 4.932797658862877e-06,
+      "loss": 0.2038,
+      "step": 50250
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.4419104306585004,
+      "learning_rate": 4.93275585284281e-06,
+      "loss": 0.2081,
+      "step": 50275
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.6288965207748185,
+      "learning_rate": 4.932714046822743e-06,
+      "loss": 0.2051,
+      "step": 50300
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.3830015306405978,
+      "learning_rate": 4.932672240802676e-06,
+      "loss": 0.1817,
+      "step": 50325
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.85804110304336,
+      "learning_rate": 4.932632107023412e-06,
+      "loss": 0.2122,
+      "step": 50350
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.268213957087889,
+      "learning_rate": 4.932590301003345e-06,
+      "loss": 0.2063,
+      "step": 50375
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.985988233626404,
+      "learning_rate": 4.932548494983278e-06,
+      "loss": 0.2142,
+      "step": 50400
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 2.0433158965290197,
+      "learning_rate": 4.9325066889632115e-06,
+      "loss": 0.195,
+      "step": 50425
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.619354197211885,
+      "learning_rate": 4.932464882943144e-06,
+      "loss": 0.1964,
+      "step": 50450
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.4976821988213502,
+      "learning_rate": 4.932423076923078e-06,
+      "loss": 0.2149,
+      "step": 50475
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 2.420182183432332,
+      "learning_rate": 4.93238127090301e-06,
+      "loss": 0.2245,
+      "step": 50500
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.9459403489592213,
+      "learning_rate": 4.932339464882944e-06,
+      "loss": 0.2198,
+      "step": 50525
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.5861503958567087,
+      "learning_rate": 4.932297658862877e-06,
+      "loss": 0.2204,
+      "step": 50550
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.8998163673028098,
+      "learning_rate": 4.93225585284281e-06,
+      "loss": 0.2106,
+      "step": 50575
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.331165993976876,
+      "learning_rate": 4.932214046822743e-06,
+      "loss": 0.1854,
+      "step": 50600
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.733723317933618,
+      "learning_rate": 4.9321722408026764e-06,
+      "loss": 0.1989,
+      "step": 50625
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.78159021767067,
+      "learning_rate": 4.932130434782609e-06,
+      "loss": 0.2156,
+      "step": 50650
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.8765020453068406,
+      "learning_rate": 4.932088628762543e-06,
+      "loss": 0.1994,
+      "step": 50675
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.2239774542089124,
+      "learning_rate": 4.932046822742475e-06,
+      "loss": 0.1955,
+      "step": 50700
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.182462271207217,
+      "learning_rate": 4.932005016722408e-06,
+      "loss": 0.1945,
+      "step": 50725
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.8092880077254656,
+      "learning_rate": 4.931963210702342e-06,
+      "loss": 0.1952,
+      "step": 50750
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.6754850868214068,
+      "learning_rate": 4.931921404682274e-06,
+      "loss": 0.2224,
+      "step": 50775
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.0066490763989873,
+      "learning_rate": 4.931879598662208e-06,
+      "loss": 0.1988,
+      "step": 50800
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.3175579753366817,
+      "learning_rate": 4.9318377926421406e-06,
+      "loss": 0.2053,
+      "step": 50825
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.3893938429753403,
+      "learning_rate": 4.931795986622074e-06,
+      "loss": 0.1887,
+      "step": 50850
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.6195076889793714,
+      "learning_rate": 4.931754180602007e-06,
+      "loss": 0.1549,
+      "step": 50875
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.5831100199621515,
+      "learning_rate": 4.93171237458194e-06,
+      "loss": 0.1559,
+      "step": 50900
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.280031099022124,
+      "learning_rate": 4.931670568561873e-06,
+      "loss": 0.1634,
+      "step": 50925
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.0793384529410677,
+      "learning_rate": 4.931628762541807e-06,
+      "loss": 0.1563,
+      "step": 50950
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.668781784899802,
+      "learning_rate": 4.931586956521739e-06,
+      "loss": 0.1525,
+      "step": 50975
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.532539607395984,
+      "learning_rate": 4.931545150501673e-06,
+      "loss": 0.1557,
+      "step": 51000
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.748971811774227,
+      "learning_rate": 4.9315033444816056e-06,
+      "loss": 0.1626,
+      "step": 51025
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.759658061175033,
+      "learning_rate": 4.931461538461539e-06,
+      "loss": 0.1551,
+      "step": 51050
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.899714433160585,
+      "learning_rate": 4.931419732441472e-06,
+      "loss": 0.1474,
+      "step": 51075
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 3.3506408589981036,
+      "learning_rate": 4.931377926421405e-06,
+      "loss": 0.1537,
+      "step": 51100
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.445193092719799,
+      "learning_rate": 4.931336120401338e-06,
+      "loss": 0.1553,
+      "step": 51125
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.018152535354776,
+      "learning_rate": 4.931294314381272e-06,
+      "loss": 0.1452,
+      "step": 51150
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.2715640087113764,
+      "learning_rate": 4.931252508361204e-06,
+      "loss": 0.159,
+      "step": 51175
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 2.6259858096583195,
+      "learning_rate": 4.931210702341138e-06,
+      "loss": 0.1432,
+      "step": 51200
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.0134734462630113,
+      "learning_rate": 4.9311688963210705e-06,
+      "loss": 0.1746,
+      "step": 51225
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.53530897566775,
+      "learning_rate": 4.931127090301004e-06,
+      "loss": 0.1572,
+      "step": 51250
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.319671554271341,
+      "learning_rate": 4.931085284280937e-06,
+      "loss": 0.1702,
+      "step": 51275
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.1657373621605718,
+      "learning_rate": 4.93104347826087e-06,
+      "loss": 0.1778,
+      "step": 51300
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.8207220923214296,
+      "learning_rate": 4.931001672240803e-06,
+      "loss": 0.1752,
+      "step": 51325
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.6051555436818306,
+      "learning_rate": 4.930961538461539e-06,
+      "loss": 0.1463,
+      "step": 51350
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.2631241428045374,
+      "learning_rate": 4.930919732441472e-06,
+      "loss": 0.1562,
+      "step": 51375
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 2.733310564982773,
+      "learning_rate": 4.930877926421405e-06,
+      "loss": 0.1436,
+      "step": 51400
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 3.0421003211716813,
+      "learning_rate": 4.930836120401339e-06,
+      "loss": 0.1439,
+      "step": 51425
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.8208998532386054,
+      "learning_rate": 4.930794314381271e-06,
+      "loss": 0.1587,
+      "step": 51450
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.619487202524097,
+      "learning_rate": 4.930752508361205e-06,
+      "loss": 0.1508,
+      "step": 51475
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.4972195849639807,
+      "learning_rate": 4.930710702341138e-06,
+      "loss": 0.1439,
+      "step": 51500
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.0416854490311533,
+      "learning_rate": 4.930668896321071e-06,
+      "loss": 0.1682,
+      "step": 51525
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.565121011610061,
+      "learning_rate": 4.930627090301004e-06,
+      "loss": 0.1571,
+      "step": 51550
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.9390353127611322,
+      "learning_rate": 4.930585284280937e-06,
+      "loss": 0.144,
+      "step": 51575
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 3.0758813481142617,
+      "learning_rate": 4.93054347826087e-06,
+      "loss": 0.1611,
+      "step": 51600
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.59428448676563,
+      "learning_rate": 4.930501672240804e-06,
+      "loss": 0.1634,
+      "step": 51625
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.038301554771562,
+      "learning_rate": 4.930459866220736e-06,
+      "loss": 0.1527,
+      "step": 51650
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.0376499179826846,
+      "learning_rate": 4.930418060200669e-06,
+      "loss": 0.165,
+      "step": 51675
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.6312057183566488,
+      "learning_rate": 4.9303762541806026e-06,
+      "loss": 0.1456,
+      "step": 51700
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.310992199986981,
+      "learning_rate": 4.930334448160535e-06,
+      "loss": 0.1473,
+      "step": 51725
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.3033985635608203,
+      "learning_rate": 4.930292642140469e-06,
+      "loss": 0.155,
+      "step": 51750
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.832532339873309,
+      "learning_rate": 4.9302508361204015e-06,
+      "loss": 0.1594,
+      "step": 51775
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.8899396483286703,
+      "learning_rate": 4.930209030100335e-06,
+      "loss": 0.1534,
+      "step": 51800
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.5216428673701485,
+      "learning_rate": 4.930167224080268e-06,
+      "loss": 0.1608,
+      "step": 51825
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 3.8041744241802538,
+      "learning_rate": 4.930125418060201e-06,
+      "loss": 0.1779,
+      "step": 51850
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.479253142386971,
+      "learning_rate": 4.930083612040134e-06,
+      "loss": 0.1623,
+      "step": 51875
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.0835171942963253,
+      "learning_rate": 4.9300418060200675e-06,
+      "loss": 0.1685,
+      "step": 51900
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.4988154900909576,
+      "learning_rate": 4.93e-06,
+      "loss": 0.1615,
+      "step": 51925
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.8221176766564573,
+      "learning_rate": 4.929958193979934e-06,
+      "loss": 0.1502,
+      "step": 51950
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.7415285030938414,
+      "learning_rate": 4.9299163879598665e-06,
+      "loss": 0.1559,
+      "step": 51975
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.704013115530803,
+      "learning_rate": 4.9298745819398e-06,
+      "loss": 0.1685,
+      "step": 52000
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 3.4942317648547885,
+      "learning_rate": 4.929832775919733e-06,
+      "loss": 0.1544,
+      "step": 52025
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.4905741598961249,
+      "learning_rate": 4.929790969899666e-06,
+      "loss": 0.1526,
+      "step": 52050
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.6196280886358214,
+      "learning_rate": 4.929749163879599e-06,
+      "loss": 0.161,
+      "step": 52075
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.8163085825589482,
+      "learning_rate": 4.9297073578595325e-06,
+      "loss": 0.1769,
+      "step": 52100
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.929047133011546,
+      "learning_rate": 4.929665551839465e-06,
+      "loss": 0.1672,
+      "step": 52125
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.483643296525586,
+      "learning_rate": 4.929623745819399e-06,
+      "loss": 0.1441,
+      "step": 52150
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.844006335610527,
+      "learning_rate": 4.929583612040134e-06,
+      "loss": 0.1718,
+      "step": 52175
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.3894318119446356,
+      "learning_rate": 4.929541806020067e-06,
+      "loss": 0.1538,
+      "step": 52200
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 2.965395513590732,
+      "learning_rate": 4.9295e-06,
+      "loss": 0.1496,
+      "step": 52225
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.304504000730068,
+      "learning_rate": 4.9294581939799335e-06,
+      "loss": 0.1669,
+      "step": 52250
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.0721485878168964,
+      "learning_rate": 4.929416387959867e-06,
+      "loss": 0.1425,
+      "step": 52275
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.8325103490337078,
+      "learning_rate": 4.9293745819398e-06,
+      "loss": 0.1522,
+      "step": 52300
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.9327528432617735,
+      "learning_rate": 4.929332775919733e-06,
+      "loss": 0.162,
+      "step": 52325
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.3305621320480137,
+      "learning_rate": 4.929290969899666e-06,
+      "loss": 0.1493,
+      "step": 52350
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.944031143750634,
+      "learning_rate": 4.9292491638795996e-06,
+      "loss": 0.1526,
+      "step": 52375
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.3706222006628677,
+      "learning_rate": 4.929207357859532e-06,
+      "loss": 0.1537,
+      "step": 52400
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.3458044853455924,
+      "learning_rate": 4.929165551839466e-06,
+      "loss": 0.1633,
+      "step": 52425
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.74893596241187,
+      "learning_rate": 4.9291237458193985e-06,
+      "loss": 0.1573,
+      "step": 52450
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 2.422402658813327,
+      "learning_rate": 4.929081939799332e-06,
+      "loss": 0.1769,
+      "step": 52475
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.1581876952824732,
+      "learning_rate": 4.929040133779265e-06,
+      "loss": 0.1667,
+      "step": 52500
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.549994934584253,
+      "learning_rate": 4.928998327759198e-06,
+      "loss": 0.1634,
+      "step": 52525
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.9136228137791405,
+      "learning_rate": 4.928956521739131e-06,
+      "loss": 0.1531,
+      "step": 52550
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.710298426938268,
+      "learning_rate": 4.9289147157190645e-06,
+      "loss": 0.1536,
+      "step": 52575
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.5543413568354856,
+      "learning_rate": 4.928872909698996e-06,
+      "loss": 0.1705,
+      "step": 52600
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.7127212550512338,
+      "learning_rate": 4.92883110367893e-06,
+      "loss": 0.1708,
+      "step": 52625
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.3037309062234965,
+      "learning_rate": 4.928789297658863e-06,
+      "loss": 0.1741,
+      "step": 52650
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.9495570743735675,
+      "learning_rate": 4.928747491638796e-06,
+      "loss": 0.1771,
+      "step": 52675
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.238430814642686,
+      "learning_rate": 4.92870568561873e-06,
+      "loss": 0.1583,
+      "step": 52700
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 2.320952142850099,
+      "learning_rate": 4.9286638795986624e-06,
+      "loss": 0.1593,
+      "step": 52725
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.1755627909046047,
+      "learning_rate": 4.928622073578596e-06,
+      "loss": 0.1469,
+      "step": 52750
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.4189873372395727,
+      "learning_rate": 4.928580267558529e-06,
+      "loss": 0.1456,
+      "step": 52775
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.2511565610144335,
+      "learning_rate": 4.928538461538462e-06,
+      "loss": 0.1436,
+      "step": 52800
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.1374086292631,
+      "learning_rate": 4.928496655518395e-06,
+      "loss": 0.143,
+      "step": 52825
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.461021396003331,
+      "learning_rate": 4.9284548494983285e-06,
+      "loss": 0.1702,
+      "step": 52850
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.9960814769649182,
+      "learning_rate": 4.928413043478261e-06,
+      "loss": 0.1594,
+      "step": 52875
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.568571889804149,
+      "learning_rate": 4.928371237458195e-06,
+      "loss": 0.1537,
+      "step": 52900
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.8880582494240996,
+      "learning_rate": 4.928329431438127e-06,
+      "loss": 0.1487,
+      "step": 52925
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.4868852996183224,
+      "learning_rate": 4.928287625418061e-06,
+      "loss": 0.1567,
+      "step": 52950
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.6979293077536246,
+      "learning_rate": 4.928245819397994e-06,
+      "loss": 0.1661,
+      "step": 52975
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.0399515236372667,
+      "learning_rate": 4.928204013377927e-06,
+      "loss": 0.1573,
+      "step": 53000
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.2471778177012527,
+      "learning_rate": 4.92816220735786e-06,
+      "loss": 0.15,
+      "step": 53025
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.8607165071886655,
+      "learning_rate": 4.9281204013377934e-06,
+      "loss": 0.158,
+      "step": 53050
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.68403495926998,
+      "learning_rate": 4.928078595317726e-06,
+      "loss": 0.1455,
+      "step": 53075
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.1127210702982535,
+      "learning_rate": 4.92803678929766e-06,
+      "loss": 0.1405,
+      "step": 53100
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.843643320066027,
+      "learning_rate": 4.927994983277592e-06,
+      "loss": 0.1716,
+      "step": 53125
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.8317086013921706,
+      "learning_rate": 4.927953177257526e-06,
+      "loss": 0.157,
+      "step": 53150
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.8083401347718393,
+      "learning_rate": 4.927911371237459e-06,
+      "loss": 0.1612,
+      "step": 53175
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.8263439878473045,
+      "learning_rate": 4.927869565217392e-06,
+      "loss": 0.1766,
+      "step": 53200
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 2.485098578437788,
+      "learning_rate": 4.927827759197325e-06,
+      "loss": 0.1884,
+      "step": 53225
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.5075459930953996,
+      "learning_rate": 4.927785953177258e-06,
+      "loss": 0.1412,
+      "step": 53250
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.395353082662295,
+      "learning_rate": 4.927744147157191e-06,
+      "loss": 0.1559,
+      "step": 53275
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.1892457117001767,
+      "learning_rate": 4.927702341137125e-06,
+      "loss": 0.1704,
+      "step": 53300
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.220899724188775,
+      "learning_rate": 4.927660535117057e-06,
+      "loss": 0.1717,
+      "step": 53325
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.894092777332042,
+      "learning_rate": 4.92761872909699e-06,
+      "loss": 0.1421,
+      "step": 53350
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.6879891749223022,
+      "learning_rate": 4.927576923076924e-06,
+      "loss": 0.1481,
+      "step": 53375
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.849208892249495,
+      "learning_rate": 4.927535117056856e-06,
+      "loss": 0.1536,
+      "step": 53400
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 3.234496990961037,
+      "learning_rate": 4.92749331103679e-06,
+      "loss": 0.1578,
+      "step": 53425
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.4492358664142355,
+      "learning_rate": 4.9274515050167225e-06,
+      "loss": 0.1511,
+      "step": 53450
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.685397894450621,
+      "learning_rate": 4.927409698996656e-06,
+      "loss": 0.174,
+      "step": 53475
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 2.3684425125628716,
+      "learning_rate": 4.927367892976589e-06,
+      "loss": 0.1614,
+      "step": 53500
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.11753748476298,
+      "learning_rate": 4.927326086956522e-06,
+      "loss": 0.1559,
+      "step": 53525
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.9354680446303794,
+      "learning_rate": 4.927284280936455e-06,
+      "loss": 0.1663,
+      "step": 53550
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.227156794826095,
+      "learning_rate": 4.9272424749163886e-06,
+      "loss": 0.1799,
+      "step": 53575
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.620633749751813,
+      "learning_rate": 4.927200668896321e-06,
+      "loss": 0.1648,
+      "step": 53600
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.9582576590163816,
+      "learning_rate": 4.927158862876255e-06,
+      "loss": 0.1611,
+      "step": 53625
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.888954284267945,
+      "learning_rate": 4.9271170568561875e-06,
+      "loss": 0.1493,
+      "step": 53650
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.5525653783661335,
+      "learning_rate": 4.927075250836121e-06,
+      "loss": 0.1539,
+      "step": 53675
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.3535390824131723,
+      "learning_rate": 4.927033444816054e-06,
+      "loss": 0.1635,
+      "step": 53700
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 2.773296192539899,
+      "learning_rate": 4.926991638795987e-06,
+      "loss": 0.1585,
+      "step": 53725
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.784025818674246,
+      "learning_rate": 4.92694983277592e-06,
+      "loss": 0.1533,
+      "step": 53750
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.671456960230062,
+      "learning_rate": 4.9269080267558536e-06,
+      "loss": 0.1546,
+      "step": 53775
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.523737848859212,
+      "learning_rate": 4.926866220735786e-06,
+      "loss": 0.1707,
+      "step": 53800
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.530100074434391,
+      "learning_rate": 4.92682441471572e-06,
+      "loss": 0.1684,
+      "step": 53825
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.3561056846922916,
+      "learning_rate": 4.9267826086956525e-06,
+      "loss": 0.1888,
+      "step": 53850
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.0777323154601883,
+      "learning_rate": 4.926740802675586e-06,
+      "loss": 0.1646,
+      "step": 53875
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.493446217332173,
+      "learning_rate": 4.926698996655519e-06,
+      "loss": 0.1517,
+      "step": 53900
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.1136296419506113,
+      "learning_rate": 4.926657190635452e-06,
+      "loss": 0.1633,
+      "step": 53925
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.3989158546724845,
+      "learning_rate": 4.926615384615385e-06,
+      "loss": 0.1561,
+      "step": 53950
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 3.0172378893049374,
+      "learning_rate": 4.9265735785953185e-06,
+      "loss": 0.1728,
+      "step": 53975
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.9134652124787792,
+      "learning_rate": 4.926531772575251e-06,
+      "loss": 0.1601,
+      "step": 54000
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.279368931878358,
+      "learning_rate": 4.926489966555184e-06,
+      "loss": 0.1619,
+      "step": 54025
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.4996559021361437,
+      "learning_rate": 4.9264481605351175e-06,
+      "loss": 0.1535,
+      "step": 54050
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.7199074418315037,
+      "learning_rate": 4.92640635451505e-06,
+      "loss": 0.1526,
+      "step": 54075
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.1648899905797956,
+      "learning_rate": 4.926364548494984e-06,
+      "loss": 0.1653,
+      "step": 54100
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.7286400940104754,
+      "learning_rate": 4.926322742474916e-06,
+      "loss": 0.1557,
+      "step": 54125
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.5889299691023648,
+      "learning_rate": 4.92628093645485e-06,
+      "loss": 0.1623,
+      "step": 54150
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.194237907627962,
+      "learning_rate": 4.926240802675586e-06,
+      "loss": 0.1694,
+      "step": 54175
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.8557622264070135,
+      "learning_rate": 4.926198996655519e-06,
+      "loss": 0.1708,
+      "step": 54200
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.6786879794483356,
+      "learning_rate": 4.926157190635452e-06,
+      "loss": 0.151,
+      "step": 54225
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 2.9939730018473427,
+      "learning_rate": 4.926115384615385e-06,
+      "loss": 0.1538,
+      "step": 54250
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.9875443858741775,
+      "learning_rate": 4.926073578595318e-06,
+      "loss": 0.176,
+      "step": 54275
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.7973867233787022,
+      "learning_rate": 4.926031772575251e-06,
+      "loss": 0.145,
+      "step": 54300
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.216337336553059,
+      "learning_rate": 4.9259899665551845e-06,
+      "loss": 0.1767,
+      "step": 54325
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 3.4129098940982208,
+      "learning_rate": 4.925948160535117e-06,
+      "loss": 0.1485,
+      "step": 54350
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.0007981016491767,
+      "learning_rate": 4.925906354515051e-06,
+      "loss": 0.1441,
+      "step": 54375
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.8583478535737,
+      "learning_rate": 4.9258645484949835e-06,
+      "loss": 0.1594,
+      "step": 54400
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.314179244280809,
+      "learning_rate": 4.925822742474917e-06,
+      "loss": 0.155,
+      "step": 54425
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.0132476893206763,
+      "learning_rate": 4.92578093645485e-06,
+      "loss": 0.1656,
+      "step": 54450
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.8635957755225794,
+      "learning_rate": 4.925739130434783e-06,
+      "loss": 0.1602,
+      "step": 54475
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.834088448909386,
+      "learning_rate": 4.925697324414716e-06,
+      "loss": 0.1691,
+      "step": 54500
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.2049287035923615,
+      "learning_rate": 4.9256555183946495e-06,
+      "loss": 0.1581,
+      "step": 54525
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 3.2612117086630983,
+      "learning_rate": 4.925613712374582e-06,
+      "loss": 0.1581,
+      "step": 54550
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.7241251316995652,
+      "learning_rate": 4.925571906354516e-06,
+      "loss": 0.1623,
+      "step": 54575
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.2161375845886737,
+      "learning_rate": 4.9255301003344484e-06,
+      "loss": 0.173,
+      "step": 54600
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.6279708477585182,
+      "learning_rate": 4.925488294314382e-06,
+      "loss": 0.1622,
+      "step": 54625
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.6368572644697492,
+      "learning_rate": 4.925446488294315e-06,
+      "loss": 0.1665,
+      "step": 54650
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.2374445674653936,
+      "learning_rate": 4.925404682274248e-06,
+      "loss": 0.1645,
+      "step": 54675
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.777041275353947,
+      "learning_rate": 4.925362876254181e-06,
+      "loss": 0.1631,
+      "step": 54700
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.4610825173728323,
+      "learning_rate": 4.9253210702341145e-06,
+      "loss": 0.1594,
+      "step": 54725
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 2.0784666529738263,
+      "learning_rate": 4.925279264214047e-06,
+      "loss": 0.1563,
+      "step": 54750
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.620773172947401,
+      "learning_rate": 4.925237458193981e-06,
+      "loss": 0.1534,
+      "step": 54775
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.5041303743520913,
+      "learning_rate": 4.925195652173913e-06,
+      "loss": 0.1823,
+      "step": 54800
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.3136541774000023,
+      "learning_rate": 4.925153846153847e-06,
+      "loss": 0.1643,
+      "step": 54825
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.2238847850927437,
+      "learning_rate": 4.92511204013378e-06,
+      "loss": 0.1761,
+      "step": 54850
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.713279366369079,
+      "learning_rate": 4.925070234113713e-06,
+      "loss": 0.1657,
+      "step": 54875
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.6333303154824503,
+      "learning_rate": 4.925028428093646e-06,
+      "loss": 0.1627,
+      "step": 54900
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 2.863859374260147,
+      "learning_rate": 4.9249866220735795e-06,
+      "loss": 0.1524,
+      "step": 54925
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.7876622608964243,
+      "learning_rate": 4.924944816053512e-06,
+      "loss": 0.1603,
+      "step": 54950
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.7726951565012345,
+      "learning_rate": 4.924903010033445e-06,
+      "loss": 0.1471,
+      "step": 54975
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.823226550047555,
+      "learning_rate": 4.924861204013378e-06,
+      "loss": 0.1562,
+      "step": 55000
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.6079458847263772,
+      "learning_rate": 4.924819397993311e-06,
+      "loss": 0.1687,
+      "step": 55025
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.505583643161287,
+      "learning_rate": 4.924777591973245e-06,
+      "loss": 0.1666,
+      "step": 55050
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.2835391606286155,
+      "learning_rate": 4.924735785953177e-06,
+      "loss": 0.1619,
+      "step": 55075
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.0288694878473654,
+      "learning_rate": 4.924693979933111e-06,
+      "loss": 0.1612,
+      "step": 55100
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.8297162400569662,
+      "learning_rate": 4.924652173913044e-06,
+      "loss": 0.1679,
+      "step": 55125
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.5215046066698381,
+      "learning_rate": 4.924610367892977e-06,
+      "loss": 0.1636,
+      "step": 55150
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 2.0415805165337577,
+      "learning_rate": 4.924570234113713e-06,
+      "loss": 0.1617,
+      "step": 55175
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.885372598413116,
+      "learning_rate": 4.924528428093646e-06,
+      "loss": 0.1503,
+      "step": 55200
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.693220968582674,
+      "learning_rate": 4.924486622073578e-06,
+      "loss": 0.174,
+      "step": 55225
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.6448831869675682,
+      "learning_rate": 4.924444816053512e-06,
+      "loss": 0.1568,
+      "step": 55250
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.6427824071243027,
+      "learning_rate": 4.924403010033445e-06,
+      "loss": 0.1409,
+      "step": 55275
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 3.5051468060927804,
+      "learning_rate": 4.924361204013378e-06,
+      "loss": 0.16,
+      "step": 55300
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.3554161547534496,
+      "learning_rate": 4.924319397993311e-06,
+      "loss": 0.1635,
+      "step": 55325
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.2438531362592653,
+      "learning_rate": 4.924277591973244e-06,
+      "loss": 0.1696,
+      "step": 55350
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.993729955191998,
+      "learning_rate": 4.924235785953178e-06,
+      "loss": 0.1414,
+      "step": 55375
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.812931969711888,
+      "learning_rate": 4.924193979933111e-06,
+      "loss": 0.1535,
+      "step": 55400
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.2778307356395966,
+      "learning_rate": 4.924152173913044e-06,
+      "loss": 0.1682,
+      "step": 55425
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.1191913216519205,
+      "learning_rate": 4.924110367892977e-06,
+      "loss": 0.1556,
+      "step": 55450
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 2.300298131748246,
+      "learning_rate": 4.9240685618729104e-06,
+      "loss": 0.1451,
+      "step": 55475
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.9976266390195618,
+      "learning_rate": 4.924026755852843e-06,
+      "loss": 0.1544,
+      "step": 55500
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.9317993781076177,
+      "learning_rate": 4.923984949832777e-06,
+      "loss": 0.1729,
+      "step": 55525
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.718753249528282,
+      "learning_rate": 4.923943143812709e-06,
+      "loss": 0.1441,
+      "step": 55550
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.2565840107062485,
+      "learning_rate": 4.923901337792643e-06,
+      "loss": 0.1674,
+      "step": 55575
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.383013144810345,
+      "learning_rate": 4.923859531772576e-06,
+      "loss": 0.1605,
+      "step": 55600
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 3.3092883642771405,
+      "learning_rate": 4.923817725752509e-06,
+      "loss": 0.1669,
+      "step": 55625
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.211414188184337,
+      "learning_rate": 4.923775919732442e-06,
+      "loss": 0.1389,
+      "step": 55650
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.488338988031535,
+      "learning_rate": 4.923734113712375e-06,
+      "loss": 0.168,
+      "step": 55675
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.300045920710584,
+      "learning_rate": 4.923692307692308e-06,
+      "loss": 0.1383,
+      "step": 55700
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.6881951652660336,
+      "learning_rate": 4.923650501672242e-06,
+      "loss": 0.1754,
+      "step": 55725
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 2.930605332694203,
+      "learning_rate": 4.923608695652174e-06,
+      "loss": 0.1585,
+      "step": 55750
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.7962333691225325,
+      "learning_rate": 4.923566889632108e-06,
+      "loss": 0.1539,
+      "step": 55775
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.6517198727704034,
+      "learning_rate": 4.923525083612041e-06,
+      "loss": 0.1631,
+      "step": 55800
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.2519972863357274,
+      "learning_rate": 4.923483277591974e-06,
+      "loss": 0.1734,
+      "step": 55825
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.322960523681727,
+      "learning_rate": 4.923441471571907e-06,
+      "loss": 0.1626,
+      "step": 55850
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.3228889974625515,
+      "learning_rate": 4.92339966555184e-06,
+      "loss": 0.1629,
+      "step": 55875
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.0280038892158694,
+      "learning_rate": 4.923357859531773e-06,
+      "loss": 0.1828,
+      "step": 55900
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.8535550307672954,
+      "learning_rate": 4.923316053511706e-06,
+      "loss": 0.1631,
+      "step": 55925
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 3.309157550953824,
+      "learning_rate": 4.923274247491639e-06,
+      "loss": 0.1722,
+      "step": 55950
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.351125286313886,
+      "learning_rate": 4.923232441471572e-06,
+      "loss": 0.1695,
+      "step": 55975
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.297372457039874,
+      "learning_rate": 4.9231906354515056e-06,
+      "loss": 0.1553,
+      "step": 56000
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 2.0925449099759983,
+      "learning_rate": 4.923148829431438e-06,
+      "loss": 0.1732,
+      "step": 56025
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.6960095376972097,
+      "learning_rate": 4.923107023411372e-06,
+      "loss": 0.1474,
+      "step": 56050
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.9693615031084075,
+      "learning_rate": 4.9230652173913045e-06,
+      "loss": 0.1656,
+      "step": 56075
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.754391153522097,
+      "learning_rate": 4.923023411371238e-06,
+      "loss": 0.1621,
+      "step": 56100
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.6063940292761676,
+      "learning_rate": 4.922981605351171e-06,
+      "loss": 0.142,
+      "step": 56125
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.000603817581123,
+      "learning_rate": 4.922939799331104e-06,
+      "loss": 0.1588,
+      "step": 56150
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.8577242463400019,
+      "learning_rate": 4.922899665551839e-06,
+      "loss": 0.1448,
+      "step": 56175
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 2.3453900092605062,
+      "learning_rate": 4.922857859531773e-06,
+      "loss": 0.1489,
+      "step": 56200
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.7724438857218991,
+      "learning_rate": 4.9228160535117055e-06,
+      "loss": 0.1619,
+      "step": 56225
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.3524805703480405,
+      "learning_rate": 4.922774247491639e-06,
+      "loss": 0.1687,
+      "step": 56250
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 3.3408597681628738,
+      "learning_rate": 4.922732441471572e-06,
+      "loss": 0.1643,
+      "step": 56275
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.7170094914128877,
+      "learning_rate": 4.922690635451505e-06,
+      "loss": 0.1667,
+      "step": 56300
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.9939337796903036,
+      "learning_rate": 4.922648829431438e-06,
+      "loss": 0.1436,
+      "step": 56325
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.413924259574225,
+      "learning_rate": 4.9226070234113716e-06,
+      "loss": 0.1715,
+      "step": 56350
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.322985702949605,
+      "learning_rate": 4.922565217391304e-06,
+      "loss": 0.1658,
+      "step": 56375
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.370524435039194,
+      "learning_rate": 4.922523411371238e-06,
+      "loss": 0.1577,
+      "step": 56400
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.983387072213947,
+      "learning_rate": 4.9224816053511705e-06,
+      "loss": 0.1768,
+      "step": 56425
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 3.188374061799889,
+      "learning_rate": 4.922439799331104e-06,
+      "loss": 0.1555,
+      "step": 56450
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.5451448494834452,
+      "learning_rate": 4.922397993311038e-06,
+      "loss": 0.1627,
+      "step": 56475
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.444626924036314,
+      "learning_rate": 4.92235618729097e-06,
+      "loss": 0.1518,
+      "step": 56500
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.0333573434866046,
+      "learning_rate": 4.922314381270904e-06,
+      "loss": 0.1682,
+      "step": 56525
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 2.404693100907314,
+      "learning_rate": 4.9222725752508365e-06,
+      "loss": 0.1605,
+      "step": 56550
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.8865735811248825,
+      "learning_rate": 4.92223076923077e-06,
+      "loss": 0.17,
+      "step": 56575
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.9171365472921278,
+      "learning_rate": 4.922188963210703e-06,
+      "loss": 0.1693,
+      "step": 56600
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.4763194923869865,
+      "learning_rate": 4.922147157190636e-06,
+      "loss": 0.1562,
+      "step": 56625
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.764068022978239,
+      "learning_rate": 4.922105351170569e-06,
+      "loss": 0.1527,
+      "step": 56650
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.1369551320651,
+      "learning_rate": 4.9220635451505026e-06,
+      "loss": 0.1488,
+      "step": 56675
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.325422901668469,
+      "learning_rate": 4.922021739130435e-06,
+      "loss": 0.1719,
+      "step": 56700
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.1041649166600105,
+      "learning_rate": 4.921979933110369e-06,
+      "loss": 0.1768,
+      "step": 56725
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.7655589341104052,
+      "learning_rate": 4.9219381270903015e-06,
+      "loss": 0.1777,
+      "step": 56750
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.8380477181947645,
+      "learning_rate": 4.921896321070235e-06,
+      "loss": 0.1644,
+      "step": 56775
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 2.3126435559314795,
+      "learning_rate": 4.921854515050168e-06,
+      "loss": 0.1731,
+      "step": 56800
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.56461416141031,
+      "learning_rate": 4.921812709030101e-06,
+      "loss": 0.159,
+      "step": 56825
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.1602569083850414,
+      "learning_rate": 4.921770903010033e-06,
+      "loss": 0.1584,
+      "step": 56850
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.8962818131934016,
+      "learning_rate": 4.921729096989967e-06,
+      "loss": 0.1576,
+      "step": 56875
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.8524281706098542,
+      "learning_rate": 4.9216872909699e-06,
+      "loss": 0.1561,
+      "step": 56900
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.9143396372190553,
+      "learning_rate": 4.921645484949833e-06,
+      "loss": 0.1692,
+      "step": 56925
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.822879321951725,
+      "learning_rate": 4.9216036789297665e-06,
+      "loss": 0.1613,
+      "step": 56950
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.894258326831287,
+      "learning_rate": 4.921561872909699e-06,
+      "loss": 0.1541,
+      "step": 56975
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.7074729590893851,
+      "learning_rate": 4.921520066889633e-06,
+      "loss": 0.1555,
+      "step": 57000
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.612601267245271,
+      "learning_rate": 4.9214782608695654e-06,
+      "loss": 0.145,
+      "step": 57025
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.3502778459294635,
+      "learning_rate": 4.921436454849499e-06,
+      "loss": 0.1602,
+      "step": 57050
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.6450546458329638,
+      "learning_rate": 4.921394648829432e-06,
+      "loss": 0.1717,
+      "step": 57075
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.6578146368646642,
+      "learning_rate": 4.921352842809365e-06,
+      "loss": 0.156,
+      "step": 57100
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.2414245528366807,
+      "learning_rate": 4.921311036789298e-06,
+      "loss": 0.1353,
+      "step": 57125
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.8998348774283498,
+      "learning_rate": 4.9212692307692315e-06,
+      "loss": 0.171,
+      "step": 57150
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.9985448953876657,
+      "learning_rate": 4.9212290969899665e-06,
+      "loss": 0.1467,
+      "step": 57175
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.399374920196816,
+      "learning_rate": 4.9211872909699e-06,
+      "loss": 0.1614,
+      "step": 57200
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.430911112893449,
+      "learning_rate": 4.921145484949833e-06,
+      "loss": 0.1678,
+      "step": 57225
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.119242897772184,
+      "learning_rate": 4.921103678929766e-06,
+      "loss": 0.1608,
+      "step": 57250
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.4857313645191823,
+      "learning_rate": 4.921061872909699e-06,
+      "loss": 0.1827,
+      "step": 57275
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 2.904099021212478,
+      "learning_rate": 4.9210200668896325e-06,
+      "loss": 0.182,
+      "step": 57300
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.548156179946889,
+      "learning_rate": 4.920978260869565e-06,
+      "loss": 0.1628,
+      "step": 57325
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.3494390107702494,
+      "learning_rate": 4.920936454849499e-06,
+      "loss": 0.1479,
+      "step": 57350
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.8864858832812406,
+      "learning_rate": 4.9208946488294314e-06,
+      "loss": 0.159,
+      "step": 57375
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.3675349817558708,
+      "learning_rate": 4.920852842809365e-06,
+      "loss": 0.1698,
+      "step": 57400
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.5225907794607534,
+      "learning_rate": 4.920811036789298e-06,
+      "loss": 0.1651,
+      "step": 57425
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.1154683368095637,
+      "learning_rate": 4.920769230769231e-06,
+      "loss": 0.1577,
+      "step": 57450
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.8981664952230868,
+      "learning_rate": 4.920727424749164e-06,
+      "loss": 0.1719,
+      "step": 57475
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.53863923398963,
+      "learning_rate": 4.9206856187290975e-06,
+      "loss": 0.1644,
+      "step": 57500
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.105776922935975,
+      "learning_rate": 4.92064381270903e-06,
+      "loss": 0.1545,
+      "step": 57525
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 2.0750252396438587,
+      "learning_rate": 4.920602006688964e-06,
+      "loss": 0.155,
+      "step": 57550
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.1660226968910576,
+      "learning_rate": 4.920560200668897e-06,
+      "loss": 0.1731,
+      "step": 57575
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.025064147146057,
+      "learning_rate": 4.92051839464883e-06,
+      "loss": 0.1658,
+      "step": 57600
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.4539822980767205,
+      "learning_rate": 4.9204765886287635e-06,
+      "loss": 0.1496,
+      "step": 57625
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.8245524096248626,
+      "learning_rate": 4.920434782608696e-06,
+      "loss": 0.1479,
+      "step": 57650
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.1986903832701565,
+      "learning_rate": 4.92039297658863e-06,
+      "loss": 0.1757,
+      "step": 57675
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.737008195248559,
+      "learning_rate": 4.9203511705685624e-06,
+      "loss": 0.1574,
+      "step": 57700
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.7053823545914026,
+      "learning_rate": 4.920309364548496e-06,
+      "loss": 0.169,
+      "step": 57725
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.4113450770376206,
+      "learning_rate": 4.920267558528429e-06,
+      "loss": 0.1667,
+      "step": 57750
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 1.970065828181729,
+      "learning_rate": 4.920225752508362e-06,
+      "loss": 0.1661,
+      "step": 57775
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 3.0890846295834673,
+      "learning_rate": 4.920183946488294e-06,
+      "loss": 0.1638,
+      "step": 57800
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 2.076681262223452,
+      "learning_rate": 4.920142140468228e-06,
+      "loss": 0.1864,
+      "step": 57825
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.343128244131371,
+      "learning_rate": 4.92010033444816e-06,
+      "loss": 0.1687,
+      "step": 57850
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.9276428639103855,
+      "learning_rate": 4.920058528428094e-06,
+      "loss": 0.1614,
+      "step": 57875
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.761415206373535,
+      "learning_rate": 4.9200167224080266e-06,
+      "loss": 0.1636,
+      "step": 57900
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.2563054496648425,
+      "learning_rate": 4.91997491638796e-06,
+      "loss": 0.1586,
+      "step": 57925
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.639280854740686,
+      "learning_rate": 4.919933110367893e-06,
+      "loss": 0.1663,
+      "step": 57950
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.4109976776005013,
+      "learning_rate": 4.919891304347826e-06,
+      "loss": 0.1607,
+      "step": 57975
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.6597424343430185,
+      "learning_rate": 4.91984949832776e-06,
+      "loss": 0.142,
+      "step": 58000
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.734114761003077,
+      "learning_rate": 4.919807692307693e-06,
+      "loss": 0.1666,
+      "step": 58025
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.396277658035052,
+      "learning_rate": 4.919765886287626e-06,
+      "loss": 0.1505,
+      "step": 58050
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 2.5980057974065858,
+      "learning_rate": 4.919724080267559e-06,
+      "loss": 0.1425,
+      "step": 58075
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.260543291214527,
+      "learning_rate": 4.919682274247492e-06,
+      "loss": 0.1507,
+      "step": 58100
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.759475488284363,
+      "learning_rate": 4.919640468227425e-06,
+      "loss": 0.1804,
+      "step": 58125
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.8038300402487817,
+      "learning_rate": 4.919598662207359e-06,
+      "loss": 0.1644,
+      "step": 58150
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.260869107549195,
+      "learning_rate": 4.919558528428094e-06,
+      "loss": 0.1508,
+      "step": 58175
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.414584150357879,
+      "learning_rate": 4.919516722408027e-06,
+      "loss": 0.1709,
+      "step": 58200
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.5049432654576578,
+      "learning_rate": 4.91947491638796e-06,
+      "loss": 0.1667,
+      "step": 58225
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.2825260278718655,
+      "learning_rate": 4.919433110367893e-06,
+      "loss": 0.1566,
+      "step": 58250
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.6614178100576562,
+      "learning_rate": 4.919391304347826e-06,
+      "loss": 0.1568,
+      "step": 58275
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 1.7269557876148456,
+      "learning_rate": 4.91934949832776e-06,
+      "loss": 0.1552,
+      "step": 58300
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 2.1426447315621164,
+      "learning_rate": 4.919307692307692e-06,
+      "loss": 0.1544,
+      "step": 58325
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.847512216664601,
+      "learning_rate": 4.919265886287626e-06,
+      "loss": 0.1763,
+      "step": 58350
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.177179572888644,
+      "learning_rate": 4.919224080267559e-06,
+      "loss": 0.1649,
+      "step": 58375
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.3301719269253813,
+      "learning_rate": 4.919182274247492e-06,
+      "loss": 0.1629,
+      "step": 58400
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.9371628068840936,
+      "learning_rate": 4.919140468227425e-06,
+      "loss": 0.1582,
+      "step": 58425
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.1562146801351836,
+      "learning_rate": 4.919098662207358e-06,
+      "loss": 0.1905,
+      "step": 58450
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.191229103225186,
+      "learning_rate": 4.919056856187291e-06,
+      "loss": 0.1694,
+      "step": 58475
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.6064059646004025,
+      "learning_rate": 4.919015050167225e-06,
+      "loss": 0.1488,
+      "step": 58500
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.323530192845432,
+      "learning_rate": 4.918973244147157e-06,
+      "loss": 0.1528,
+      "step": 58525
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.095986879359018,
+      "learning_rate": 4.918931438127091e-06,
+      "loss": 0.1711,
+      "step": 58550
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 2.10375244642294,
+      "learning_rate": 4.918889632107024e-06,
+      "loss": 0.1797,
+      "step": 58575
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.4047401055803066,
+      "learning_rate": 4.918847826086957e-06,
+      "loss": 0.1688,
+      "step": 58600
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.310141666785082,
+      "learning_rate": 4.91880602006689e-06,
+      "loss": 0.1661,
+      "step": 58625
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.86452149655283,
+      "learning_rate": 4.918764214046823e-06,
+      "loss": 0.1591,
+      "step": 58650
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.187797758585452,
+      "learning_rate": 4.918722408026757e-06,
+      "loss": 0.1584,
+      "step": 58675
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.8069341965191519,
+      "learning_rate": 4.91868060200669e-06,
+      "loss": 0.1555,
+      "step": 58700
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.18250105172939,
+      "learning_rate": 4.918638795986623e-06,
+      "loss": 0.1623,
+      "step": 58725
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.6732097415682214,
+      "learning_rate": 4.918596989966555e-06,
+      "loss": 0.1664,
+      "step": 58750
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.2977159672754395,
+      "learning_rate": 4.9185551839464886e-06,
+      "loss": 0.166,
+      "step": 58775
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.007392545610804,
+      "learning_rate": 4.918513377926421e-06,
+      "loss": 0.1634,
+      "step": 58800
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 2.7077236084984597,
+      "learning_rate": 4.918471571906355e-06,
+      "loss": 0.1491,
+      "step": 58825
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.424611894537868,
+      "learning_rate": 4.9184297658862875e-06,
+      "loss": 0.1587,
+      "step": 58850
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.3860933115368321,
+      "learning_rate": 4.918387959866221e-06,
+      "loss": 0.1496,
+      "step": 58875
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.6511674403642846,
+      "learning_rate": 4.918346153846154e-06,
+      "loss": 0.1678,
+      "step": 58900
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.1685696271223653,
+      "learning_rate": 4.918304347826087e-06,
+      "loss": 0.1411,
+      "step": 58925
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.0485582790269428,
+      "learning_rate": 4.91826254180602e-06,
+      "loss": 0.1757,
+      "step": 58950
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.6032184254228543,
+      "learning_rate": 4.9182207357859535e-06,
+      "loss": 0.1625,
+      "step": 58975
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.027124650130354,
+      "learning_rate": 4.918178929765886e-06,
+      "loss": 0.1421,
+      "step": 59000
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.8814506846542645,
+      "learning_rate": 4.91813712374582e-06,
+      "loss": 0.1628,
+      "step": 59025
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.1579894725073154,
+      "learning_rate": 4.9180953177257525e-06,
+      "loss": 0.1654,
+      "step": 59050
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.111540003358208,
+      "learning_rate": 4.918053511705686e-06,
+      "loss": 0.1695,
+      "step": 59075
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.7137729677726354,
+      "learning_rate": 4.9180117056856196e-06,
+      "loss": 0.1486,
+      "step": 59100
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.7812965649458703,
+      "learning_rate": 4.917969899665552e-06,
+      "loss": 0.1638,
+      "step": 59125
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.011544514420453,
+      "learning_rate": 4.917928093645486e-06,
+      "loss": 0.1733,
+      "step": 59150
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.390700805022534,
+      "learning_rate": 4.917887959866221e-06,
+      "loss": 0.1751,
+      "step": 59175
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.969735384499303,
+      "learning_rate": 4.917846153846154e-06,
+      "loss": 0.1729,
+      "step": 59200
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.913785187777396,
+      "learning_rate": 4.917804347826087e-06,
+      "loss": 0.153,
+      "step": 59225
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.8021064858925475,
+      "learning_rate": 4.917762541806021e-06,
+      "loss": 0.1502,
+      "step": 59250
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.833978772258698,
+      "learning_rate": 4.917720735785953e-06,
+      "loss": 0.1724,
+      "step": 59275
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 2.2849194247990177,
+      "learning_rate": 4.917678929765887e-06,
+      "loss": 0.1698,
+      "step": 59300
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.6327536122141721,
+      "learning_rate": 4.9176371237458195e-06,
+      "loss": 0.1479,
+      "step": 59325
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.8262023075338034,
+      "learning_rate": 4.917595317725753e-06,
+      "loss": 0.1642,
+      "step": 59350
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.9784854693898795,
+      "learning_rate": 4.917553511705686e-06,
+      "loss": 0.158,
+      "step": 59375
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.359260864971092,
+      "learning_rate": 4.917511705685619e-06,
+      "loss": 0.1571,
+      "step": 59400
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.2374038047494427,
+      "learning_rate": 4.917469899665552e-06,
+      "loss": 0.1471,
+      "step": 59425
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.0946720369210645,
+      "learning_rate": 4.9174280936454856e-06,
+      "loss": 0.1658,
+      "step": 59450
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.395192692613216,
+      "learning_rate": 4.917386287625418e-06,
+      "loss": 0.1662,
+      "step": 59475
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.107283464209178,
+      "learning_rate": 4.917344481605352e-06,
+      "loss": 0.1646,
+      "step": 59500
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 3.4601379596453037,
+      "learning_rate": 4.9173026755852845e-06,
+      "loss": 0.1577,
+      "step": 59525
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.7885302397106813,
+      "learning_rate": 4.917260869565218e-06,
+      "loss": 0.1491,
+      "step": 59550
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.9441052075639202,
+      "learning_rate": 4.917219063545151e-06,
+      "loss": 0.1518,
+      "step": 59575
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 2.435857624262805,
+      "learning_rate": 4.917177257525084e-06,
+      "loss": 0.1631,
+      "step": 59600
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.4963602725155,
+      "learning_rate": 4.917135451505017e-06,
+      "loss": 0.1588,
+      "step": 59625
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.324784100257762,
+      "learning_rate": 4.9170936454849505e-06,
+      "loss": 0.1539,
+      "step": 59650
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.0187377882547612,
+      "learning_rate": 4.917051839464883e-06,
+      "loss": 0.1695,
+      "step": 59675
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.32940588948083,
+      "learning_rate": 4.917010033444816e-06,
+      "loss": 0.1516,
+      "step": 59700
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.836519988909741,
+      "learning_rate": 4.9169682274247495e-06,
+      "loss": 0.1747,
+      "step": 59725
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.196327753251213,
+      "learning_rate": 4.916926421404682e-06,
+      "loss": 0.1765,
+      "step": 59750
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.2094447994990345,
+      "learning_rate": 4.916884615384616e-06,
+      "loss": 0.1525,
+      "step": 59775
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.5435910280268885,
+      "learning_rate": 4.916842809364548e-06,
+      "loss": 0.1588,
+      "step": 59800
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 2.1170442526101163,
+      "learning_rate": 4.916801003344482e-06,
+      "loss": 0.1652,
+      "step": 59825
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 4.0496458307816825,
+      "learning_rate": 4.916759197324415e-06,
+      "loss": 0.1676,
+      "step": 59850
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.5764839005683617,
+      "learning_rate": 4.916717391304348e-06,
+      "loss": 0.1524,
+      "step": 59875
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.80735502088511,
+      "learning_rate": 4.916675585284281e-06,
+      "loss": 0.1646,
+      "step": 59900
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 3.4931943985698877,
+      "learning_rate": 4.9166337792642145e-06,
+      "loss": 0.1688,
+      "step": 59925
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.620793187205274,
+      "learning_rate": 4.916591973244147e-06,
+      "loss": 0.1645,
+      "step": 59950
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.2352409464681644,
+      "learning_rate": 4.916550167224081e-06,
+      "loss": 0.1485,
+      "step": 59975
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.8848876549630622,
+      "learning_rate": 4.916508361204013e-06,
+      "loss": 0.1717,
+      "step": 60000
+    },
+    {
+      "epoch": 2.36,
+      "eval_loss": 0.5126953125,
+      "eval_runtime": 11570.5057,
+      "eval_samples_per_second": 0.818,
+      "eval_steps_per_second": 0.051,
+      "eval_wer": 0.11778710808680405,
+      "step": 60000
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.5457840080713066,
+      "learning_rate": 4.916466555183947e-06,
+      "loss": 0.163,
+      "step": 60025
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.7137502472167718,
+      "learning_rate": 4.91642474916388e-06,
+      "loss": 0.153,
+      "step": 60050
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.0294624199856943,
+      "learning_rate": 4.916382943143813e-06,
+      "loss": 0.1589,
+      "step": 60075
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 2.3906210473393097,
+      "learning_rate": 4.916341137123746e-06,
+      "loss": 0.1652,
+      "step": 60100
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.2050739597355082,
+      "learning_rate": 4.9162993311036794e-06,
+      "loss": 0.1638,
+      "step": 60125
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 3.2900598636843386,
+      "learning_rate": 4.916257525083612e-06,
+      "loss": 0.1728,
+      "step": 60150
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 2.1524349935755915,
+      "learning_rate": 4.916217391304348e-06,
+      "loss": 0.1489,
+      "step": 60175
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 2.954050934593037,
+      "learning_rate": 4.9161755852842815e-06,
+      "loss": 0.1858,
+      "step": 60200
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.9504239110160009,
+      "learning_rate": 4.916133779264214e-06,
+      "loss": 0.1584,
+      "step": 60225
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 5.834701243563905,
+      "learning_rate": 4.916091973244148e-06,
+      "loss": 0.1653,
+      "step": 60250
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 2.504682679956055,
+      "learning_rate": 4.9160501672240805e-06,
+      "loss": 0.1728,
+      "step": 60275
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 2.6111928108040954,
+      "learning_rate": 4.916008361204014e-06,
+      "loss": 0.1695,
+      "step": 60300
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 2.7115359266814565,
+      "learning_rate": 4.915966555183947e-06,
+      "loss": 0.1659,
+      "step": 60325
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 2.7986136435971476,
+      "learning_rate": 4.91592474916388e-06,
+      "loss": 0.1561,
+      "step": 60350
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.2374505980103256,
+      "learning_rate": 4.915882943143813e-06,
+      "loss": 0.1811,
+      "step": 60375
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.677702691876237,
+      "learning_rate": 4.9158411371237465e-06,
+      "loss": 0.1578,
+      "step": 60400
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.710984414065094,
+      "learning_rate": 4.915799331103679e-06,
+      "loss": 0.1766,
+      "step": 60425
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.8611113380470368,
+      "learning_rate": 4.915757525083613e-06,
+      "loss": 0.1788,
+      "step": 60450
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.6445493558289472,
+      "learning_rate": 4.9157157190635454e-06,
+      "loss": 0.1777,
+      "step": 60475
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.476843097247296,
+      "learning_rate": 4.915673913043479e-06,
+      "loss": 0.1554,
+      "step": 60500
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.009772580934335,
+      "learning_rate": 4.915632107023412e-06,
+      "loss": 0.176,
+      "step": 60525
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.799057734334839,
+      "learning_rate": 4.915590301003345e-06,
+      "loss": 0.1687,
+      "step": 60550
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 2.574514219545785,
+      "learning_rate": 4.915548494983278e-06,
+      "loss": 0.1772,
+      "step": 60575
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.7987424850971203,
+      "learning_rate": 4.9155066889632115e-06,
+      "loss": 0.1662,
+      "step": 60600
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.08168165567613,
+      "learning_rate": 4.915464882943144e-06,
+      "loss": 0.1705,
+      "step": 60625
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.306741935160784,
+      "learning_rate": 4.915423076923077e-06,
+      "loss": 0.1644,
+      "step": 60650
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.9982050594746743,
+      "learning_rate": 4.91538127090301e-06,
+      "loss": 0.1637,
+      "step": 60675
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.3306033444527405,
+      "learning_rate": 4.915339464882943e-06,
+      "loss": 0.166,
+      "step": 60700
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.239829674488769,
+      "learning_rate": 4.915297658862877e-06,
+      "loss": 0.1721,
+      "step": 60725
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.0716748422338145,
+      "learning_rate": 4.915255852842809e-06,
+      "loss": 0.1665,
+      "step": 60750
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.3627489051626163,
+      "learning_rate": 4.915214046822743e-06,
+      "loss": 0.1656,
+      "step": 60775
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.7947561118872883,
+      "learning_rate": 4.915172240802676e-06,
+      "loss": 0.1821,
+      "step": 60800
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.481319736763784,
+      "learning_rate": 4.915130434782609e-06,
+      "loss": 0.1608,
+      "step": 60825
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.5617043799686217,
+      "learning_rate": 4.915088628762542e-06,
+      "loss": 0.1556,
+      "step": 60850
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 2.552707926658039,
+      "learning_rate": 4.915046822742475e-06,
+      "loss": 0.1748,
+      "step": 60875
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.4439378238092067,
+      "learning_rate": 4.915005016722408e-06,
+      "loss": 0.1682,
+      "step": 60900
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.998043095324892,
+      "learning_rate": 4.914963210702342e-06,
+      "loss": 0.1976,
+      "step": 60925
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.2108208593309526,
+      "learning_rate": 4.914921404682274e-06,
+      "loss": 0.1552,
+      "step": 60950
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.9371077350313313,
+      "learning_rate": 4.914879598662208e-06,
+      "loss": 0.1467,
+      "step": 60975
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.7039801639447034,
+      "learning_rate": 4.9148377926421406e-06,
+      "loss": 0.1518,
+      "step": 61000
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.915961676427643,
+      "learning_rate": 4.914795986622074e-06,
+      "loss": 0.1638,
+      "step": 61025
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.9113673117236556,
+      "learning_rate": 4.914754180602007e-06,
+      "loss": 0.1676,
+      "step": 61050
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.4662369168057308,
+      "learning_rate": 4.91471237458194e-06,
+      "loss": 0.1757,
+      "step": 61075
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 2.0060579273255077,
+      "learning_rate": 4.914670568561873e-06,
+      "loss": 0.1525,
+      "step": 61100
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.6452207663404552,
+      "learning_rate": 4.914628762541807e-06,
+      "loss": 0.186,
+      "step": 61125
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.043111214852014,
+      "learning_rate": 4.914586956521739e-06,
+      "loss": 0.1721,
+      "step": 61150
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.2998189089040224,
+      "learning_rate": 4.914546822742475e-06,
+      "loss": 0.173,
+      "step": 61175
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.229036607651321,
+      "learning_rate": 4.914505016722409e-06,
+      "loss": 0.1677,
+      "step": 61200
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.2040697290855484,
+      "learning_rate": 4.914463210702341e-06,
+      "loss": 0.1565,
+      "step": 61225
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.975337279121286,
+      "learning_rate": 4.914421404682275e-06,
+      "loss": 0.1567,
+      "step": 61250
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.1901711072719783,
+      "learning_rate": 4.914379598662208e-06,
+      "loss": 0.1615,
+      "step": 61275
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.410651010100104,
+      "learning_rate": 4.914337792642141e-06,
+      "loss": 0.1596,
+      "step": 61300
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.473836763687653,
+      "learning_rate": 4.914295986622074e-06,
+      "loss": 0.173,
+      "step": 61325
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.0361562862320617,
+      "learning_rate": 4.914254180602007e-06,
+      "loss": 0.1705,
+      "step": 61350
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 2.6879914919584564,
+      "learning_rate": 4.91421237458194e-06,
+      "loss": 0.1607,
+      "step": 61375
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.157305173574173,
+      "learning_rate": 4.914170568561874e-06,
+      "loss": 0.1512,
+      "step": 61400
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.9317567624460439,
+      "learning_rate": 4.914128762541806e-06,
+      "loss": 0.1569,
+      "step": 61425
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.361584589222622,
+      "learning_rate": 4.91408695652174e-06,
+      "loss": 0.1638,
+      "step": 61450
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.8644040646184097,
+      "learning_rate": 4.914045150501673e-06,
+      "loss": 0.1687,
+      "step": 61475
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.198548588498155,
+      "learning_rate": 4.914003344481606e-06,
+      "loss": 0.154,
+      "step": 61500
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.607394816578382,
+      "learning_rate": 4.913961538461539e-06,
+      "loss": 0.1646,
+      "step": 61525
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.7313264618704025,
+      "learning_rate": 4.913919732441472e-06,
+      "loss": 0.1666,
+      "step": 61550
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.8872158667961136,
+      "learning_rate": 4.913877926421405e-06,
+      "loss": 0.1594,
+      "step": 61575
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.005983861142906,
+      "learning_rate": 4.913836120401338e-06,
+      "loss": 0.1671,
+      "step": 61600
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.6811726737767227,
+      "learning_rate": 4.913794314381271e-06,
+      "loss": 0.1773,
+      "step": 61625
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.4459555545243172,
+      "learning_rate": 4.913752508361204e-06,
+      "loss": 0.1666,
+      "step": 61650
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.1227618361315246,
+      "learning_rate": 4.9137107023411376e-06,
+      "loss": 0.1716,
+      "step": 61675
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.1285845172063067,
+      "learning_rate": 4.91366889632107e-06,
+      "loss": 0.1703,
+      "step": 61700
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 3.090357456437784,
+      "learning_rate": 4.913627090301004e-06,
+      "loss": 0.1602,
+      "step": 61725
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.0482938851822245,
+      "learning_rate": 4.9135852842809365e-06,
+      "loss": 0.1505,
+      "step": 61750
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.5875726035509103,
+      "learning_rate": 4.91354347826087e-06,
+      "loss": 0.1724,
+      "step": 61775
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.6512222827222043,
+      "learning_rate": 4.913501672240803e-06,
+      "loss": 0.1617,
+      "step": 61800
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.8674513598205074,
+      "learning_rate": 4.913459866220736e-06,
+      "loss": 0.1602,
+      "step": 61825
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.3103847629985763,
+      "learning_rate": 4.913419732441472e-06,
+      "loss": 0.1747,
+      "step": 61850
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 2.3023261514395763,
+      "learning_rate": 4.913377926421405e-06,
+      "loss": 0.1581,
+      "step": 61875
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.449709317162717,
+      "learning_rate": 4.913336120401338e-06,
+      "loss": 0.1742,
+      "step": 61900
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.2658196543121774,
+      "learning_rate": 4.913294314381271e-06,
+      "loss": 0.1684,
+      "step": 61925
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.2361849806007785,
+      "learning_rate": 4.913252508361205e-06,
+      "loss": 0.1745,
+      "step": 61950
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.8218866051780491,
+      "learning_rate": 4.913210702341137e-06,
+      "loss": 0.1522,
+      "step": 61975
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.1296586951775764,
+      "learning_rate": 4.913168896321071e-06,
+      "loss": 0.1695,
+      "step": 62000
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.7770083377004013,
+      "learning_rate": 4.9131270903010036e-06,
+      "loss": 0.1667,
+      "step": 62025
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.9390687351900326,
+      "learning_rate": 4.913085284280937e-06,
+      "loss": 0.1539,
+      "step": 62050
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.7051591263423855,
+      "learning_rate": 4.91304347826087e-06,
+      "loss": 0.1547,
+      "step": 62075
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 3.1409256857260597,
+      "learning_rate": 4.913001672240803e-06,
+      "loss": 0.176,
+      "step": 62100
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 2.794092126918202,
+      "learning_rate": 4.912959866220736e-06,
+      "loss": 0.1676,
+      "step": 62125
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.165282938697665,
+      "learning_rate": 4.91291806020067e-06,
+      "loss": 0.1711,
+      "step": 62150
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.8794889473406264,
+      "learning_rate": 4.912876254180602e-06,
+      "loss": 0.1683,
+      "step": 62175
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.753755884035197,
+      "learning_rate": 4.912834448160536e-06,
+      "loss": 0.1782,
+      "step": 62200
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.691916719938932,
+      "learning_rate": 4.9127926421404685e-06,
+      "loss": 0.1764,
+      "step": 62225
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.4777575790892183,
+      "learning_rate": 4.912750836120402e-06,
+      "loss": 0.1778,
+      "step": 62250
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.082308380129261,
+      "learning_rate": 4.912709030100335e-06,
+      "loss": 0.1734,
+      "step": 62275
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.5276306526812324,
+      "learning_rate": 4.912667224080268e-06,
+      "loss": 0.166,
+      "step": 62300
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.0447729926081506,
+      "learning_rate": 4.912625418060201e-06,
+      "loss": 0.1519,
+      "step": 62325
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.1623636996029796,
+      "learning_rate": 4.912583612040135e-06,
+      "loss": 0.1607,
+      "step": 62350
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.06902259972889,
+      "learning_rate": 4.912541806020067e-06,
+      "loss": 0.1686,
+      "step": 62375
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 2.4778763891548006,
+      "learning_rate": 4.912500000000001e-06,
+      "loss": 0.1682,
+      "step": 62400
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.9091976161593263,
+      "learning_rate": 4.9124581939799335e-06,
+      "loss": 0.172,
+      "step": 62425
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.913027598613694,
+      "learning_rate": 4.912416387959867e-06,
+      "loss": 0.1601,
+      "step": 62450
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.007501580547803,
+      "learning_rate": 4.9123745819398e-06,
+      "loss": 0.1453,
+      "step": 62475
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.8561817066776358,
+      "learning_rate": 4.9123327759197325e-06,
+      "loss": 0.1705,
+      "step": 62500
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.161749847169862,
+      "learning_rate": 4.912290969899666e-06,
+      "loss": 0.1737,
+      "step": 62525
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.1232595660705242,
+      "learning_rate": 4.912249163879599e-06,
+      "loss": 0.1574,
+      "step": 62550
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.7837035969587884,
+      "learning_rate": 4.912207357859532e-06,
+      "loss": 0.1544,
+      "step": 62575
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.8829737147072323,
+      "learning_rate": 4.912165551839465e-06,
+      "loss": 0.1546,
+      "step": 62600
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.7869648211061147,
+      "learning_rate": 4.9121237458193985e-06,
+      "loss": 0.1605,
+      "step": 62625
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 2.693915151097951,
+      "learning_rate": 4.912081939799331e-06,
+      "loss": 0.1612,
+      "step": 62650
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.8652334580681769,
+      "learning_rate": 4.912040133779265e-06,
+      "loss": 0.1682,
+      "step": 62675
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 3.4003444849061477,
+      "learning_rate": 4.9119983277591974e-06,
+      "loss": 0.1905,
+      "step": 62700
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.3689247940924316,
+      "learning_rate": 4.911956521739131e-06,
+      "loss": 0.1614,
+      "step": 62725
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.08385825279254,
+      "learning_rate": 4.911914715719064e-06,
+      "loss": 0.1698,
+      "step": 62750
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.6497607238941634,
+      "learning_rate": 4.911872909698997e-06,
+      "loss": 0.16,
+      "step": 62775
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.5394626117283912,
+      "learning_rate": 4.91183110367893e-06,
+      "loss": 0.1627,
+      "step": 62800
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.20644693606779,
+      "learning_rate": 4.9117892976588635e-06,
+      "loss": 0.1673,
+      "step": 62825
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.0547792480076508,
+      "learning_rate": 4.911747491638796e-06,
+      "loss": 0.1673,
+      "step": 62850
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.143794378453494,
+      "learning_rate": 4.91170568561873e-06,
+      "loss": 0.154,
+      "step": 62875
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.8088862278469335,
+      "learning_rate": 4.911663879598662e-06,
+      "loss": 0.1577,
+      "step": 62900
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.5161033945231814,
+      "learning_rate": 4.911622073578596e-06,
+      "loss": 0.1738,
+      "step": 62925
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.6053312739319967,
+      "learning_rate": 4.911580267558529e-06,
+      "loss": 0.1561,
+      "step": 62950
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 3.093633337419838,
+      "learning_rate": 4.911538461538462e-06,
+      "loss": 0.1611,
+      "step": 62975
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.5048885696579846,
+      "learning_rate": 4.911496655518395e-06,
+      "loss": 0.1537,
+      "step": 63000
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.4850024515720803,
+      "learning_rate": 4.9114548494983285e-06,
+      "loss": 0.1749,
+      "step": 63025
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.267747673130082,
+      "learning_rate": 4.911413043478261e-06,
+      "loss": 0.1647,
+      "step": 63050
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.013113252947993,
+      "learning_rate": 4.911371237458195e-06,
+      "loss": 0.1516,
+      "step": 63075
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.373805904984799,
+      "learning_rate": 4.911329431438127e-06,
+      "loss": 0.1644,
+      "step": 63100
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.012563965344033,
+      "learning_rate": 4.911287625418061e-06,
+      "loss": 0.1786,
+      "step": 63125
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.510353666710276,
+      "learning_rate": 4.911245819397994e-06,
+      "loss": 0.1585,
+      "step": 63150
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.894923289500306,
+      "learning_rate": 4.911204013377927e-06,
+      "loss": 0.1642,
+      "step": 63175
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.9141467079712533,
+      "learning_rate": 4.91116220735786e-06,
+      "loss": 0.1633,
+      "step": 63200
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.1543105842708106,
+      "learning_rate": 4.911120401337793e-06,
+      "loss": 0.1817,
+      "step": 63225
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.196066091671809,
+      "learning_rate": 4.911078595317726e-06,
+      "loss": 0.1654,
+      "step": 63250
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.574609823964885,
+      "learning_rate": 4.911036789297659e-06,
+      "loss": 0.165,
+      "step": 63275
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.7239441756225606,
+      "learning_rate": 4.910994983277592e-06,
+      "loss": 0.1449,
+      "step": 63300
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.155397841796658,
+      "learning_rate": 4.910953177257525e-06,
+      "loss": 0.1639,
+      "step": 63325
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.0676428516731504,
+      "learning_rate": 4.910911371237459e-06,
+      "loss": 0.1544,
+      "step": 63350
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 2.908000877442016,
+      "learning_rate": 4.910869565217391e-06,
+      "loss": 0.1663,
+      "step": 63375
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.2534265614289273,
+      "learning_rate": 4.910827759197325e-06,
+      "loss": 0.1467,
+      "step": 63400
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.6717733780311184,
+      "learning_rate": 4.9107859531772576e-06,
+      "loss": 0.1734,
+      "step": 63425
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.115600821956127,
+      "learning_rate": 4.910744147157191e-06,
+      "loss": 0.1538,
+      "step": 63450
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.7077889016665388,
+      "learning_rate": 4.910702341137124e-06,
+      "loss": 0.1645,
+      "step": 63475
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.079355530724143,
+      "learning_rate": 4.910660535117057e-06,
+      "loss": 0.1635,
+      "step": 63500
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.6935640568962875,
+      "learning_rate": 4.91061872909699e-06,
+      "loss": 0.1669,
+      "step": 63525
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.501349861253381,
+      "learning_rate": 4.910576923076924e-06,
+      "loss": 0.1618,
+      "step": 63550
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 2.258413469914614,
+      "learning_rate": 4.910535117056856e-06,
+      "loss": 0.1594,
+      "step": 63575
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.048614342796082,
+      "learning_rate": 4.91049331103679e-06,
+      "loss": 0.1712,
+      "step": 63600
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 3.199999060663906,
+      "learning_rate": 4.9104515050167225e-06,
+      "loss": 0.1532,
+      "step": 63625
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.0916681640164947,
+      "learning_rate": 4.910409698996656e-06,
+      "loss": 0.1793,
+      "step": 63650
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.5617092875552199,
+      "learning_rate": 4.910367892976589e-06,
+      "loss": 0.1502,
+      "step": 63675
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.431000733781789,
+      "learning_rate": 4.910326086956522e-06,
+      "loss": 0.1618,
+      "step": 63700
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.03889062924083,
+      "learning_rate": 4.910284280936455e-06,
+      "loss": 0.1408,
+      "step": 63725
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 3.1713570023080253,
+      "learning_rate": 4.9102424749163886e-06,
+      "loss": 0.1626,
+      "step": 63750
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.4983183955799593,
+      "learning_rate": 4.910200668896321e-06,
+      "loss": 0.1668,
+      "step": 63775
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.588454811441986,
+      "learning_rate": 4.910158862876255e-06,
+      "loss": 0.1438,
+      "step": 63800
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.8740606752064477,
+      "learning_rate": 4.9101170568561875e-06,
+      "loss": 0.1613,
+      "step": 63825
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.863894051691388,
+      "learning_rate": 4.910076923076923e-06,
+      "loss": 0.1711,
+      "step": 63850
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.9926581104511647,
+      "learning_rate": 4.910035117056857e-06,
+      "loss": 0.1759,
+      "step": 63875
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.9004004513452737,
+      "learning_rate": 4.90999331103679e-06,
+      "loss": 0.1798,
+      "step": 63900
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 2.7288616511230606,
+      "learning_rate": 4.909951505016723e-06,
+      "loss": 0.1736,
+      "step": 63925
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 2.0606551832809346,
+      "learning_rate": 4.909909698996656e-06,
+      "loss": 0.1632,
+      "step": 63950
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 2.2478370244128305,
+      "learning_rate": 4.909867892976589e-06,
+      "loss": 0.1657,
+      "step": 63975
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.6587952879693575,
+      "learning_rate": 4.909826086956522e-06,
+      "loss": 0.1754,
+      "step": 64000
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 2.3087314952369464,
+      "learning_rate": 4.909784280936456e-06,
+      "loss": 0.1749,
+      "step": 64025
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.886888875975629,
+      "learning_rate": 4.909742474916388e-06,
+      "loss": 0.1704,
+      "step": 64050
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 2.5427882532308765,
+      "learning_rate": 4.909700668896322e-06,
+      "loss": 0.1669,
+      "step": 64075
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 2.6760075354807005,
+      "learning_rate": 4.9096588628762546e-06,
+      "loss": 0.1585,
+      "step": 64100
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 2.1830101383040494,
+      "learning_rate": 4.909617056856188e-06,
+      "loss": 0.1607,
+      "step": 64125
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 2.215983424795264,
+      "learning_rate": 4.909575250836121e-06,
+      "loss": 0.1649,
+      "step": 64150
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 2.471626141069725,
+      "learning_rate": 4.9095334448160535e-06,
+      "loss": 0.1478,
+      "step": 64175
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.65952247452563,
+      "learning_rate": 4.909491638795987e-06,
+      "loss": 0.1804,
+      "step": 64200
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.6232088149469335,
+      "learning_rate": 4.90944983277592e-06,
+      "loss": 0.1594,
+      "step": 64225
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.030118235294256,
+      "learning_rate": 4.909408026755853e-06,
+      "loss": 0.1521,
+      "step": 64250
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.872547185240935,
+      "learning_rate": 4.909366220735786e-06,
+      "loss": 0.1704,
+      "step": 64275
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.5078969587091784,
+      "learning_rate": 4.9093244147157195e-06,
+      "loss": 0.1573,
+      "step": 64300
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.9992021677403815,
+      "learning_rate": 4.909282608695652e-06,
+      "loss": 0.1498,
+      "step": 64325
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.503697200714061,
+      "learning_rate": 4.909240802675586e-06,
+      "loss": 0.1568,
+      "step": 64350
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.9626698849941484,
+      "learning_rate": 4.9091989966555185e-06,
+      "loss": 0.1539,
+      "step": 64375
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.490978407286913,
+      "learning_rate": 4.909157190635452e-06,
+      "loss": 0.1551,
+      "step": 64400
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.783432543381823,
+      "learning_rate": 4.909115384615385e-06,
+      "loss": 0.1824,
+      "step": 64425
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.0354994054583373,
+      "learning_rate": 4.909073578595318e-06,
+      "loss": 0.1626,
+      "step": 64450
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.9963627176197933,
+      "learning_rate": 4.909031772575251e-06,
+      "loss": 0.1672,
+      "step": 64475
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.57838119622678,
+      "learning_rate": 4.9089899665551845e-06,
+      "loss": 0.1816,
+      "step": 64500
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.4899750928360356,
+      "learning_rate": 4.908948160535117e-06,
+      "loss": 0.1634,
+      "step": 64525
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 2.1326207915213065,
+      "learning_rate": 4.908906354515051e-06,
+      "loss": 0.163,
+      "step": 64550
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.7329610880996327,
+      "learning_rate": 4.9088645484949835e-06,
+      "loss": 0.1631,
+      "step": 64575
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.597797070784199,
+      "learning_rate": 4.908822742474917e-06,
+      "loss": 0.1481,
+      "step": 64600
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.593120783537612,
+      "learning_rate": 4.90878093645485e-06,
+      "loss": 0.1619,
+      "step": 64625
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.3340178656573458,
+      "learning_rate": 4.908739130434783e-06,
+      "loss": 0.1619,
+      "step": 64650
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 3.027825908379043,
+      "learning_rate": 4.908697324414716e-06,
+      "loss": 0.1784,
+      "step": 64675
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.3061907558400665,
+      "learning_rate": 4.9086555183946495e-06,
+      "loss": 0.1728,
+      "step": 64700
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.8145354193239998,
+      "learning_rate": 4.908613712374582e-06,
+      "loss": 0.1677,
+      "step": 64725
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 3.2098623467674083,
+      "learning_rate": 4.908571906354516e-06,
+      "loss": 0.1799,
+      "step": 64750
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.9804590850895585,
+      "learning_rate": 4.9085301003344484e-06,
+      "loss": 0.1655,
+      "step": 64775
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.647426146096992,
+      "learning_rate": 4.908488294314382e-06,
+      "loss": 0.1464,
+      "step": 64800
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.9055118736851995,
+      "learning_rate": 4.908446488294315e-06,
+      "loss": 0.1771,
+      "step": 64825
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.8514368286597422,
+      "learning_rate": 4.9084063545150505e-06,
+      "loss": 0.174,
+      "step": 64850
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.9887629605918113,
+      "learning_rate": 4.908364548494984e-06,
+      "loss": 0.1621,
+      "step": 64875
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.4961759409869595,
+      "learning_rate": 4.908322742474917e-06,
+      "loss": 0.1472,
+      "step": 64900
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.4943853813301795,
+      "learning_rate": 4.90828093645485e-06,
+      "loss": 0.159,
+      "step": 64925
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.1198505134386583,
+      "learning_rate": 4.908239130434783e-06,
+      "loss": 0.1427,
+      "step": 64950
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.9455591466937527,
+      "learning_rate": 4.9081973244147165e-06,
+      "loss": 0.1655,
+      "step": 64975
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.276710046339653,
+      "learning_rate": 4.908155518394649e-06,
+      "loss": 0.1622,
+      "step": 65000
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.112276112652573,
+      "learning_rate": 4.908113712374583e-06,
+      "loss": 0.1697,
+      "step": 65025
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.465633305808334,
+      "learning_rate": 4.9080719063545155e-06,
+      "loss": 0.154,
+      "step": 65050
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.393310551281838,
+      "learning_rate": 4.908030100334449e-06,
+      "loss": 0.1471,
+      "step": 65075
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.4821237816549835,
+      "learning_rate": 4.907988294314382e-06,
+      "loss": 0.1585,
+      "step": 65100
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.3884124180340605,
+      "learning_rate": 4.9079464882943144e-06,
+      "loss": 0.1573,
+      "step": 65125
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.8117247832441796,
+      "learning_rate": 4.907904682274248e-06,
+      "loss": 0.1629,
+      "step": 65150
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.8371308882763624,
+      "learning_rate": 4.907862876254181e-06,
+      "loss": 0.164,
+      "step": 65175
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.677839446744097,
+      "learning_rate": 4.907821070234114e-06,
+      "loss": 0.1646,
+      "step": 65200
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.1656191975374735,
+      "learning_rate": 4.907779264214047e-06,
+      "loss": 0.1666,
+      "step": 65225
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.9170574359950836,
+      "learning_rate": 4.9077374581939805e-06,
+      "loss": 0.1752,
+      "step": 65250
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.2651721052092806,
+      "learning_rate": 4.907695652173913e-06,
+      "loss": 0.1652,
+      "step": 65275
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.848088945866107,
+      "learning_rate": 4.907653846153847e-06,
+      "loss": 0.1677,
+      "step": 65300
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.3172246339739884,
+      "learning_rate": 4.907612040133779e-06,
+      "loss": 0.1718,
+      "step": 65325
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.8405472169057213,
+      "learning_rate": 4.907570234113713e-06,
+      "loss": 0.1616,
+      "step": 65350
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.917029554946475,
+      "learning_rate": 4.907528428093646e-06,
+      "loss": 0.1669,
+      "step": 65375
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.2302027109018785,
+      "learning_rate": 4.907486622073579e-06,
+      "loss": 0.1707,
+      "step": 65400
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.1756400588268625,
+      "learning_rate": 4.907444816053512e-06,
+      "loss": 0.1797,
+      "step": 65425
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.441329759624097,
+      "learning_rate": 4.9074030100334454e-06,
+      "loss": 0.1632,
+      "step": 65450
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 2.6632498992499993,
+      "learning_rate": 4.907361204013378e-06,
+      "loss": 0.185,
+      "step": 65475
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.7352876275637894,
+      "learning_rate": 4.907319397993312e-06,
+      "loss": 0.1627,
+      "step": 65500
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 2.4227494238068776,
+      "learning_rate": 4.907277591973244e-06,
+      "loss": 0.1605,
+      "step": 65525
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 2.076807380471068,
+      "learning_rate": 4.907235785953178e-06,
+      "loss": 0.1692,
+      "step": 65550
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.9389772507642347,
+      "learning_rate": 4.907193979933111e-06,
+      "loss": 0.1554,
+      "step": 65575
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 2.548984641939041,
+      "learning_rate": 4.907152173913044e-06,
+      "loss": 0.1627,
+      "step": 65600
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 2.6788514190264143,
+      "learning_rate": 4.907110367892977e-06,
+      "loss": 0.1436,
+      "step": 65625
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 3.1995893686446286,
+      "learning_rate": 4.90706856187291e-06,
+      "loss": 0.178,
+      "step": 65650
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.6593985754679415,
+      "learning_rate": 4.907026755852843e-06,
+      "loss": 0.1685,
+      "step": 65675
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 2.317242944867583,
+      "learning_rate": 4.906984949832777e-06,
+      "loss": 0.1656,
+      "step": 65700
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 2.8102006179505628,
+      "learning_rate": 4.906943143812709e-06,
+      "loss": 0.1587,
+      "step": 65725
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 2.201863264339552,
+      "learning_rate": 4.906901337792643e-06,
+      "loss": 0.1607,
+      "step": 65750
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.702427793526987,
+      "learning_rate": 4.906859531772576e-06,
+      "loss": 0.1755,
+      "step": 65775
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 2.6686775607203446,
+      "learning_rate": 4.906817725752509e-06,
+      "loss": 0.1396,
+      "step": 65800
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 2.6544320471662886,
+      "learning_rate": 4.906775919732442e-06,
+      "loss": 0.1568,
+      "step": 65825
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7099841241182607,
+      "learning_rate": 4.906735785953178e-06,
+      "loss": 0.1543,
+      "step": 65850
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.9112198873388229,
+      "learning_rate": 4.906693979933111e-06,
+      "loss": 0.1526,
+      "step": 65875
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 2.3657780976079748,
+      "learning_rate": 4.906652173913044e-06,
+      "loss": 0.1566,
+      "step": 65900
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.9480778857317624,
+      "learning_rate": 4.9066103678929775e-06,
+      "loss": 0.1585,
+      "step": 65925
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 2.942927544287174,
+      "learning_rate": 4.90656856187291e-06,
+      "loss": 0.1671,
+      "step": 65950
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.9627043489734612,
+      "learning_rate": 4.906526755852844e-06,
+      "loss": 0.1725,
+      "step": 65975
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.696701666683548,
+      "learning_rate": 4.906486622073579e-06,
+      "loss": 0.1576,
+      "step": 66000
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.4217429930751502,
+      "learning_rate": 4.906444816053512e-06,
+      "loss": 0.164,
+      "step": 66025
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 2.549382249171763,
+      "learning_rate": 4.906403010033445e-06,
+      "loss": 0.1618,
+      "step": 66050
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 2.2060418166923053,
+      "learning_rate": 4.9063612040133785e-06,
+      "loss": 0.1752,
+      "step": 66075
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 2.526241507559611,
+      "learning_rate": 4.906319397993311e-06,
+      "loss": 0.1661,
+      "step": 66100
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 2.024390593256092,
+      "learning_rate": 4.906277591973245e-06,
+      "loss": 0.1674,
+      "step": 66125
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 2.5548120313238436,
+      "learning_rate": 4.9062357859531774e-06,
+      "loss": 0.1555,
+      "step": 66150
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.7203300785919162,
+      "learning_rate": 4.906193979933111e-06,
+      "loss": 0.1654,
+      "step": 66175
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 2.1977595422357936,
+      "learning_rate": 4.9061521739130445e-06,
+      "loss": 0.1771,
+      "step": 66200
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.7535093547919756,
+      "learning_rate": 4.906110367892977e-06,
+      "loss": 0.1603,
+      "step": 66225
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 2.154775437570024,
+      "learning_rate": 4.906068561872911e-06,
+      "loss": 0.1671,
+      "step": 66250
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 2.957733586615425,
+      "learning_rate": 4.906026755852843e-06,
+      "loss": 0.1551,
+      "step": 66275
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 2.017968097634726,
+      "learning_rate": 4.905984949832776e-06,
+      "loss": 0.1718,
+      "step": 66300
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 2.1178933664306134,
+      "learning_rate": 4.905943143812709e-06,
+      "loss": 0.1418,
+      "step": 66325
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 2.2771629555356636,
+      "learning_rate": 4.905901337792642e-06,
+      "loss": 0.1591,
+      "step": 66350
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 2.114269650050633,
+      "learning_rate": 4.905859531772575e-06,
+      "loss": 0.157,
+      "step": 66375
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.8226274914884266,
+      "learning_rate": 4.905817725752509e-06,
+      "loss": 0.1586,
+      "step": 66400
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 2.2937073243363506,
+      "learning_rate": 4.905775919732441e-06,
+      "loss": 0.1605,
+      "step": 66425
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 1.521627303225354,
+      "learning_rate": 4.905734113712375e-06,
+      "loss": 0.1731,
+      "step": 66450
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.1538938575588253,
+      "learning_rate": 4.905692307692308e-06,
+      "loss": 0.1662,
+      "step": 66475
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.1191327411063714,
+      "learning_rate": 4.905650501672241e-06,
+      "loss": 0.1563,
+      "step": 66500
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.9074138182924498,
+      "learning_rate": 4.905608695652174e-06,
+      "loss": 0.155,
+      "step": 66525
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.914625498132579,
+      "learning_rate": 4.905566889632107e-06,
+      "loss": 0.1624,
+      "step": 66550
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.6539242263530385,
+      "learning_rate": 4.90552508361204e-06,
+      "loss": 0.1714,
+      "step": 66575
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.088674751899696,
+      "learning_rate": 4.905483277591974e-06,
+      "loss": 0.1714,
+      "step": 66600
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.2135373112639694,
+      "learning_rate": 4.905441471571907e-06,
+      "loss": 0.1584,
+      "step": 66625
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.4192960242700194,
+      "learning_rate": 4.90539966555184e-06,
+      "loss": 0.1591,
+      "step": 66650
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.7390386910080664,
+      "learning_rate": 4.905357859531773e-06,
+      "loss": 0.1769,
+      "step": 66675
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.232094694142498,
+      "learning_rate": 4.905316053511706e-06,
+      "loss": 0.1455,
+      "step": 66700
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.0138843173390297,
+      "learning_rate": 4.90527424749164e-06,
+      "loss": 0.1514,
+      "step": 66725
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.0070506284026393,
+      "learning_rate": 4.905232441471572e-06,
+      "loss": 0.1866,
+      "step": 66750
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.4365348345881692,
+      "learning_rate": 4.905190635451506e-06,
+      "loss": 0.1739,
+      "step": 66775
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.0958303312111766,
+      "learning_rate": 4.905148829431439e-06,
+      "loss": 0.1554,
+      "step": 66800
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.4725145319946504,
+      "learning_rate": 4.905107023411372e-06,
+      "loss": 0.1675,
+      "step": 66825
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.718854874162693,
+      "learning_rate": 4.905065217391305e-06,
+      "loss": 0.1548,
+      "step": 66850
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.0116226602644884,
+      "learning_rate": 4.905023411371238e-06,
+      "loss": 0.1478,
+      "step": 66875
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.8899799082771738,
+      "learning_rate": 4.904981605351171e-06,
+      "loss": 0.161,
+      "step": 66900
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.3658188452538966,
+      "learning_rate": 4.904939799331105e-06,
+      "loss": 0.1599,
+      "step": 66925
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 2.509051426616327,
+      "learning_rate": 4.904897993311037e-06,
+      "loss": 0.1614,
+      "step": 66950
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 3.211165300889533,
+      "learning_rate": 4.904856187290971e-06,
+      "loss": 0.1553,
+      "step": 66975
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.6733992628510892,
+      "learning_rate": 4.904814381270903e-06,
+      "loss": 0.1711,
+      "step": 67000
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.3085595836756876,
+      "learning_rate": 4.904772575250836e-06,
+      "loss": 0.167,
+      "step": 67025
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.9511587121489784,
+      "learning_rate": 4.90473076923077e-06,
+      "loss": 0.1802,
+      "step": 67050
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.7009959258288003,
+      "learning_rate": 4.9046889632107025e-06,
+      "loss": 0.1695,
+      "step": 67075
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.861255834759058,
+      "learning_rate": 4.904647157190636e-06,
+      "loss": 0.1708,
+      "step": 67100
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.332246734287938,
+      "learning_rate": 4.904605351170569e-06,
+      "loss": 0.1647,
+      "step": 67125
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 3.0582558350326523,
+      "learning_rate": 4.904563545150502e-06,
+      "loss": 0.1568,
+      "step": 67150
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.838261030151132,
+      "learning_rate": 4.904521739130435e-06,
+      "loss": 0.1527,
+      "step": 67175
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.9315235968632813,
+      "learning_rate": 4.9044799331103686e-06,
+      "loss": 0.1829,
+      "step": 67200
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 2.167459167955874,
+      "learning_rate": 4.904438127090301e-06,
+      "loss": 0.1524,
+      "step": 67225
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.8651054355321226,
+      "learning_rate": 4.904396321070235e-06,
+      "loss": 0.1765,
+      "step": 67250
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.974004440932649,
+      "learning_rate": 4.9043545150501675e-06,
+      "loss": 0.179,
+      "step": 67275
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.7036311477514448,
+      "learning_rate": 4.904312709030101e-06,
+      "loss": 0.1502,
+      "step": 67300
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 2.3390162169708817,
+      "learning_rate": 4.904270903010034e-06,
+      "loss": 0.1632,
+      "step": 67325
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 2.763794342227092,
+      "learning_rate": 4.904229096989967e-06,
+      "loss": 0.177,
+      "step": 67350
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 2.5359226898254685,
+      "learning_rate": 4.9041872909699e-06,
+      "loss": 0.1675,
+      "step": 67375
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 3.1775710087532154,
+      "learning_rate": 4.9041454849498335e-06,
+      "loss": 0.1738,
+      "step": 67400
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 2.345033829775269,
+      "learning_rate": 4.904103678929766e-06,
+      "loss": 0.1675,
+      "step": 67425
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 2.336885577490091,
+      "learning_rate": 4.9040618729097e-06,
+      "loss": 0.1517,
+      "step": 67450
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 2.4565329151898867,
+      "learning_rate": 4.9040200668896325e-06,
+      "loss": 0.1669,
+      "step": 67475
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 2.5585596954556937,
+      "learning_rate": 4.903978260869566e-06,
+      "loss": 0.1685,
+      "step": 67500
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.7937763982225357,
+      "learning_rate": 4.903936454849499e-06,
+      "loss": 0.1479,
+      "step": 67525
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.6019104450697434,
+      "learning_rate": 4.903894648829432e-06,
+      "loss": 0.1623,
+      "step": 67550
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 2.4719412168073753,
+      "learning_rate": 4.903852842809365e-06,
+      "loss": 0.1754,
+      "step": 67575
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 2.370947438562601,
+      "learning_rate": 4.9038110367892985e-06,
+      "loss": 0.1644,
+      "step": 67600
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 3.131913579404752,
+      "learning_rate": 4.903769230769231e-06,
+      "loss": 0.169,
+      "step": 67625
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.743261368460137,
+      "learning_rate": 4.903727424749165e-06,
+      "loss": 0.1668,
+      "step": 67650
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.3480808608244685,
+      "learning_rate": 4.9036856187290975e-06,
+      "loss": 0.164,
+      "step": 67675
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.8534317999026895,
+      "learning_rate": 4.90364381270903e-06,
+      "loss": 0.1497,
+      "step": 67700
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 2.960335057951582,
+      "learning_rate": 4.903602006688964e-06,
+      "loss": 0.1727,
+      "step": 67725
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.423990432992723,
+      "learning_rate": 4.903560200668896e-06,
+      "loss": 0.1529,
+      "step": 67750
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.0462023472811737,
+      "learning_rate": 4.90351839464883e-06,
+      "loss": 0.1727,
+      "step": 67775
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.6099807901854786,
+      "learning_rate": 4.903476588628763e-06,
+      "loss": 0.1578,
+      "step": 67800
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.170286265230943,
+      "learning_rate": 4.903434782608696e-06,
+      "loss": 0.1584,
+      "step": 67825
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.3937380773545813,
+      "learning_rate": 4.903392976588629e-06,
+      "loss": 0.1791,
+      "step": 67850
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.5374488563219373,
+      "learning_rate": 4.9033511705685624e-06,
+      "loss": 0.1427,
+      "step": 67875
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.2583070199750472,
+      "learning_rate": 4.903309364548495e-06,
+      "loss": 0.1636,
+      "step": 67900
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 1.8468048091641243,
+      "learning_rate": 4.903267558528429e-06,
+      "loss": 0.1519,
+      "step": 67925
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.6291588159527373,
+      "learning_rate": 4.903225752508361e-06,
+      "loss": 0.1758,
+      "step": 67950
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 3.7050360639335347,
+      "learning_rate": 4.903183946488295e-06,
+      "loss": 0.1749,
+      "step": 67975
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.7682431370811424,
+      "learning_rate": 4.90314381270903e-06,
+      "loss": 0.1573,
+      "step": 68000
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 2.3824688003115226,
+      "learning_rate": 4.9031020066889635e-06,
+      "loss": 0.1468,
+      "step": 68025
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.7485707314388916,
+      "learning_rate": 4.903060200668896e-06,
+      "loss": 0.1615,
+      "step": 68050
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.7128641702363026,
+      "learning_rate": 4.90301839464883e-06,
+      "loss": 0.1592,
+      "step": 68075
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 2.075482147796586,
+      "learning_rate": 4.902976588628762e-06,
+      "loss": 0.1352,
+      "step": 68100
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.6209920469074521,
+      "learning_rate": 4.902934782608696e-06,
+      "loss": 0.1712,
+      "step": 68125
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 2.2059154225617035,
+      "learning_rate": 4.902892976588629e-06,
+      "loss": 0.1582,
+      "step": 68150
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.5679563332864526,
+      "learning_rate": 4.902851170568562e-06,
+      "loss": 0.1644,
+      "step": 68175
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.3764731345936694,
+      "learning_rate": 4.902811036789298e-06,
+      "loss": 0.1656,
+      "step": 68200
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.9139328926308926,
+      "learning_rate": 4.902769230769231e-06,
+      "loss": 0.167,
+      "step": 68225
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 2.2718914414232003,
+      "learning_rate": 4.902727424749164e-06,
+      "loss": 0.1524,
+      "step": 68250
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 2.0111405659125294,
+      "learning_rate": 4.902685618729097e-06,
+      "loss": 0.1858,
+      "step": 68275
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 2.2865058081144105,
+      "learning_rate": 4.9026438127090305e-06,
+      "loss": 0.1537,
+      "step": 68300
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 2.1726475243792116,
+      "learning_rate": 4.902602006688963e-06,
+      "loss": 0.1633,
+      "step": 68325
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 2.8799278645516595,
+      "learning_rate": 4.902560200668897e-06,
+      "loss": 0.1432,
+      "step": 68350
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.1468444524866888,
+      "learning_rate": 4.9025183946488294e-06,
+      "loss": 0.1729,
+      "step": 68375
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.6869292960241524,
+      "learning_rate": 4.902476588628763e-06,
+      "loss": 0.1692,
+      "step": 68400
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 2.2943763214968387,
+      "learning_rate": 4.902434782608696e-06,
+      "loss": 0.1523,
+      "step": 68425
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.9078772412947974,
+      "learning_rate": 4.902392976588629e-06,
+      "loss": 0.1489,
+      "step": 68450
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 2.624638040883813,
+      "learning_rate": 4.902351170568562e-06,
+      "loss": 0.1663,
+      "step": 68475
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 2.0599344273565725,
+      "learning_rate": 4.9023093645484955e-06,
+      "loss": 0.1476,
+      "step": 68500
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 2.132918351620154,
+      "learning_rate": 4.902267558528428e-06,
+      "loss": 0.1706,
+      "step": 68525
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 2.2024926639954443,
+      "learning_rate": 4.902225752508362e-06,
+      "loss": 0.1638,
+      "step": 68550
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 2.00228229652154,
+      "learning_rate": 4.9021839464882944e-06,
+      "loss": 0.1745,
+      "step": 68575
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 2.4173513208178314,
+      "learning_rate": 4.902142140468228e-06,
+      "loss": 0.1667,
+      "step": 68600
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.4535956046540373,
+      "learning_rate": 4.902100334448161e-06,
+      "loss": 0.1872,
+      "step": 68625
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.5770697206236601,
+      "learning_rate": 4.902058528428094e-06,
+      "loss": 0.1584,
+      "step": 68650
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 2.1907711996947206,
+      "learning_rate": 4.902016722408027e-06,
+      "loss": 0.1614,
+      "step": 68675
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.5996563388129679,
+      "learning_rate": 4.9019749163879605e-06,
+      "loss": 0.1528,
+      "step": 68700
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.5217021335901246,
+      "learning_rate": 4.901933110367893e-06,
+      "loss": 0.1583,
+      "step": 68725
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 2.830636659279085,
+      "learning_rate": 4.901891304347827e-06,
+      "loss": 0.16,
+      "step": 68750
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 3.132431011223764,
+      "learning_rate": 4.901849498327759e-06,
+      "loss": 0.1755,
+      "step": 68775
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 2.954810412804257,
+      "learning_rate": 4.901807692307693e-06,
+      "loss": 0.1593,
+      "step": 68800
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.9873971899648006,
+      "learning_rate": 4.901765886287626e-06,
+      "loss": 0.1576,
+      "step": 68825
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.8546844581500215,
+      "learning_rate": 4.901724080267559e-06,
+      "loss": 0.1486,
+      "step": 68850
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 2.699256024899676,
+      "learning_rate": 4.901682274247492e-06,
+      "loss": 0.1497,
+      "step": 68875
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 2.1603180169557,
+      "learning_rate": 4.901640468227425e-06,
+      "loss": 0.159,
+      "step": 68900
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.9479327767344141,
+      "learning_rate": 4.901598662207358e-06,
+      "loss": 0.1679,
+      "step": 68925
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.3827271970959603,
+      "learning_rate": 4.901556856187291e-06,
+      "loss": 0.1696,
+      "step": 68950
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.806564358309027,
+      "learning_rate": 4.901515050167224e-06,
+      "loss": 0.1659,
+      "step": 68975
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 2.0332614028768843,
+      "learning_rate": 4.901473244147157e-06,
+      "loss": 0.174,
+      "step": 69000
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 2.079579653117786,
+      "learning_rate": 4.901431438127091e-06,
+      "loss": 0.1801,
+      "step": 69025
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 2.6221209766597116,
+      "learning_rate": 4.901389632107023e-06,
+      "loss": 0.1836,
+      "step": 69050
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 2.0314819729265583,
+      "learning_rate": 4.901347826086957e-06,
+      "loss": 0.1723,
+      "step": 69075
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 2.561945317804322,
+      "learning_rate": 4.9013060200668896e-06,
+      "loss": 0.1716,
+      "step": 69100
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 2.22404829656543,
+      "learning_rate": 4.901264214046823e-06,
+      "loss": 0.1616,
+      "step": 69125
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.0273108921916183,
+      "learning_rate": 4.901222408026756e-06,
+      "loss": 0.1553,
+      "step": 69150
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.9271862459705924,
+      "learning_rate": 4.901180602006689e-06,
+      "loss": 0.1535,
+      "step": 69175
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 1.8153291263804212,
+      "learning_rate": 4.901138795986622e-06,
+      "loss": 0.167,
+      "step": 69200
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 3.3250437575275225,
+      "learning_rate": 4.901096989966556e-06,
+      "loss": 0.1723,
+      "step": 69225
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 2.767320385273223,
+      "learning_rate": 4.901055183946488e-06,
+      "loss": 0.1731,
+      "step": 69250
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 2.0717377921678923,
+      "learning_rate": 4.901013377926422e-06,
+      "loss": 0.1831,
+      "step": 69275
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 2.044160851544343,
+      "learning_rate": 4.900971571906355e-06,
+      "loss": 0.1707,
+      "step": 69300
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 2.3873404711946207,
+      "learning_rate": 4.900929765886288e-06,
+      "loss": 0.1549,
+      "step": 69325
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 2.2695647905827445,
+      "learning_rate": 4.900887959866222e-06,
+      "loss": 0.1818,
+      "step": 69350
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.4378326071784568,
+      "learning_rate": 4.900846153846154e-06,
+      "loss": 0.1617,
+      "step": 69375
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.9957570406745557,
+      "learning_rate": 4.900804347826088e-06,
+      "loss": 0.1649,
+      "step": 69400
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 2.2610619693808034,
+      "learning_rate": 4.9007625418060206e-06,
+      "loss": 0.1614,
+      "step": 69425
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.9526855098138778,
+      "learning_rate": 4.900720735785954e-06,
+      "loss": 0.1709,
+      "step": 69450
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.8988605064657225,
+      "learning_rate": 4.900678929765887e-06,
+      "loss": 0.1671,
+      "step": 69475
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 2.2746147588855834,
+      "learning_rate": 4.90063712374582e-06,
+      "loss": 0.1577,
+      "step": 69500
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 2.3770617254178164,
+      "learning_rate": 4.900595317725753e-06,
+      "loss": 0.1719,
+      "step": 69525
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.6617922742946107,
+      "learning_rate": 4.900553511705687e-06,
+      "loss": 0.1661,
+      "step": 69550
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.783213648386689,
+      "learning_rate": 4.900511705685619e-06,
+      "loss": 0.1591,
+      "step": 69575
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 2.981878987936314,
+      "learning_rate": 4.900469899665552e-06,
+      "loss": 0.1658,
+      "step": 69600
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.6476488160209035,
+      "learning_rate": 4.900428093645485e-06,
+      "loss": 0.1536,
+      "step": 69625
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 2.325940941846499,
+      "learning_rate": 4.900386287625418e-06,
+      "loss": 0.169,
+      "step": 69650
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 2.4505268528811524,
+      "learning_rate": 4.900344481605351e-06,
+      "loss": 0.1702,
+      "step": 69675
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.9311266368341664,
+      "learning_rate": 4.9003026755852845e-06,
+      "loss": 0.1548,
+      "step": 69700
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.612976220389442,
+      "learning_rate": 4.900260869565218e-06,
+      "loss": 0.1612,
+      "step": 69725
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 2.8961876952169523,
+      "learning_rate": 4.900219063545151e-06,
+      "loss": 0.1839,
+      "step": 69750
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.7490468429363928,
+      "learning_rate": 4.900177257525084e-06,
+      "loss": 0.1601,
+      "step": 69775
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 2.701145780751548,
+      "learning_rate": 4.900135451505017e-06,
+      "loss": 0.1826,
+      "step": 69800
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.8448000328926708,
+      "learning_rate": 4.9000936454849505e-06,
+      "loss": 0.1698,
+      "step": 69825
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.6674747031267765,
+      "learning_rate": 4.900051839464883e-06,
+      "loss": 0.1673,
+      "step": 69850
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 2.9567258027871106,
+      "learning_rate": 4.900010033444817e-06,
+      "loss": 0.1816,
+      "step": 69875
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.9510130741773436,
+      "learning_rate": 4.8999682274247495e-06,
+      "loss": 0.1765,
+      "step": 69900
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 2.3497872557880353,
+      "learning_rate": 4.899926421404683e-06,
+      "loss": 0.1721,
+      "step": 69925
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 2.5324795074581137,
+      "learning_rate": 4.899884615384616e-06,
+      "loss": 0.1654,
+      "step": 69950
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.497124922083692,
+      "learning_rate": 4.899842809364549e-06,
+      "loss": 0.1709,
+      "step": 69975
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 2.0770713159583654,
+      "learning_rate": 4.899801003344482e-06,
+      "loss": 0.1692,
+      "step": 70000
+    },
+    {
+      "epoch": 2.75,
+      "eval_loss": 0.60400390625,
+      "eval_runtime": 11576.636,
+      "eval_samples_per_second": 0.818,
+      "eval_steps_per_second": 0.051,
+      "eval_wer": 0.11463410710551293,
+      "step": 70000
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 3.404522336853661,
+      "learning_rate": 4.8997591973244155e-06,
+      "loss": 0.1641,
+      "step": 70025
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 2.604856475398737,
+      "learning_rate": 4.899717391304348e-06,
+      "loss": 0.163,
+      "step": 70050
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.9673677744338196,
+      "learning_rate": 4.899675585284282e-06,
+      "loss": 0.1639,
+      "step": 70075
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 2.522347353607375,
+      "learning_rate": 4.8996337792642144e-06,
+      "loss": 0.1478,
+      "step": 70100
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 2.0517890231441367,
+      "learning_rate": 4.899591973244148e-06,
+      "loss": 0.1658,
+      "step": 70125
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 2.0923956047064167,
+      "learning_rate": 4.899550167224081e-06,
+      "loss": 0.1676,
+      "step": 70150
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 2.301297191893959,
+      "learning_rate": 4.899508361204014e-06,
+      "loss": 0.1651,
+      "step": 70175
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 2.583427012599905,
+      "learning_rate": 4.899468227424749e-06,
+      "loss": 0.1932,
+      "step": 70200
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 2.2125202227042093,
+      "learning_rate": 4.899426421404683e-06,
+      "loss": 0.1605,
+      "step": 70225
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 2.443241511177795,
+      "learning_rate": 4.8993846153846155e-06,
+      "loss": 0.1649,
+      "step": 70250
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 2.2941758774610634,
+      "learning_rate": 4.899342809364549e-06,
+      "loss": 0.1624,
+      "step": 70275
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 2.041645811171096,
+      "learning_rate": 4.899301003344482e-06,
+      "loss": 0.1718,
+      "step": 70300
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 3.4730930100201176,
+      "learning_rate": 4.899259197324415e-06,
+      "loss": 0.175,
+      "step": 70325
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 2.557520966431368,
+      "learning_rate": 4.899217391304348e-06,
+      "loss": 0.177,
+      "step": 70350
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 2.0185704278932226,
+      "learning_rate": 4.8991755852842815e-06,
+      "loss": 0.1474,
+      "step": 70375
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 3.1796598269669145,
+      "learning_rate": 4.899133779264215e-06,
+      "loss": 0.1533,
+      "step": 70400
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 2.000036451479218,
+      "learning_rate": 4.899091973244148e-06,
+      "loss": 0.1616,
+      "step": 70425
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 2.416211515477164,
+      "learning_rate": 4.899050167224081e-06,
+      "loss": 0.1909,
+      "step": 70450
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 2.3646053101731757,
+      "learning_rate": 4.899008361204014e-06,
+      "loss": 0.1586,
+      "step": 70475
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 2.2500982911348384,
+      "learning_rate": 4.8989665551839475e-06,
+      "loss": 0.169,
+      "step": 70500
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.9615764608426884,
+      "learning_rate": 4.898924749163879e-06,
+      "loss": 0.1608,
+      "step": 70525
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 2.359861228050328,
+      "learning_rate": 4.898882943143813e-06,
+      "loss": 0.1743,
+      "step": 70550
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.986395507242446,
+      "learning_rate": 4.898841137123746e-06,
+      "loss": 0.1612,
+      "step": 70575
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 2.650973993750791,
+      "learning_rate": 4.898799331103679e-06,
+      "loss": 0.1907,
+      "step": 70600
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.5054656396604489,
+      "learning_rate": 4.898757525083612e-06,
+      "loss": 0.1614,
+      "step": 70625
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 2.099185235357897,
+      "learning_rate": 4.898715719063545e-06,
+      "loss": 0.1599,
+      "step": 70650
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.752903980750435,
+      "learning_rate": 4.898673913043478e-06,
+      "loss": 0.1727,
+      "step": 70675
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 2.215584387164041,
+      "learning_rate": 4.898632107023412e-06,
+      "loss": 0.1555,
+      "step": 70700
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 2.7082133233014423,
+      "learning_rate": 4.898590301003344e-06,
+      "loss": 0.1586,
+      "step": 70725
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.8680573169122017,
+      "learning_rate": 4.898548494983278e-06,
+      "loss": 0.1922,
+      "step": 70750
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.5818802598899575,
+      "learning_rate": 4.898506688963211e-06,
+      "loss": 0.1632,
+      "step": 70775
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 2.6198897893488717,
+      "learning_rate": 4.898464882943144e-06,
+      "loss": 0.1765,
+      "step": 70800
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 2.112835040259786,
+      "learning_rate": 4.898423076923078e-06,
+      "loss": 0.174,
+      "step": 70825
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 2.5854543693695047,
+      "learning_rate": 4.89838127090301e-06,
+      "loss": 0.1454,
+      "step": 70850
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.339170672832636,
+      "learning_rate": 4.898339464882944e-06,
+      "loss": 0.1358,
+      "step": 70875
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.6926476355982136,
+      "learning_rate": 4.898297658862877e-06,
+      "loss": 0.1496,
+      "step": 70900
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.5075863109608603,
+      "learning_rate": 4.89825585284281e-06,
+      "loss": 0.1552,
+      "step": 70925
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 2.2298242138770483,
+      "learning_rate": 4.898214046822743e-06,
+      "loss": 0.1658,
+      "step": 70950
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 2.56403234832974,
+      "learning_rate": 4.8981722408026764e-06,
+      "loss": 0.1713,
+      "step": 70975
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 2.266105914417862,
+      "learning_rate": 4.898130434782609e-06,
+      "loss": 0.1721,
+      "step": 71000
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 2.625059941253109,
+      "learning_rate": 4.898088628762543e-06,
+      "loss": 0.1688,
+      "step": 71025
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 2.192109072550119,
+      "learning_rate": 4.898046822742475e-06,
+      "loss": 0.1651,
+      "step": 71050
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 4.035841787276603,
+      "learning_rate": 4.898005016722409e-06,
+      "loss": 0.164,
+      "step": 71075
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 3.4732146936336017,
+      "learning_rate": 4.897963210702342e-06,
+      "loss": 0.1783,
+      "step": 71100
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.9311210420583675,
+      "learning_rate": 4.897921404682275e-06,
+      "loss": 0.1717,
+      "step": 71125
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.8756949688926736,
+      "learning_rate": 4.897879598662208e-06,
+      "loss": 0.1659,
+      "step": 71150
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 2.422990879289581,
+      "learning_rate": 4.897837792642141e-06,
+      "loss": 0.1725,
+      "step": 71175
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 2.8900976364859154,
+      "learning_rate": 4.897797658862876e-06,
+      "loss": 0.1763,
+      "step": 71200
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 2.381504042282298,
+      "learning_rate": 4.89775585284281e-06,
+      "loss": 0.1592,
+      "step": 71225
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.8401869313416392,
+      "learning_rate": 4.897714046822743e-06,
+      "loss": 0.1719,
+      "step": 71250
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.2582691293928905,
+      "learning_rate": 4.897672240802676e-06,
+      "loss": 0.1661,
+      "step": 71275
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.3065157469687312,
+      "learning_rate": 4.897630434782609e-06,
+      "loss": 0.1574,
+      "step": 71300
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 2.8882287169254925,
+      "learning_rate": 4.8975886287625424e-06,
+      "loss": 0.1781,
+      "step": 71325
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.8066680445158254,
+      "learning_rate": 4.897546822742475e-06,
+      "loss": 0.1514,
+      "step": 71350
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 2.1135129603595155,
+      "learning_rate": 4.897505016722409e-06,
+      "loss": 0.168,
+      "step": 71375
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 2.4731969275899046,
+      "learning_rate": 4.897463210702341e-06,
+      "loss": 0.1679,
+      "step": 71400
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 2.1670796062973263,
+      "learning_rate": 4.897421404682275e-06,
+      "loss": 0.1721,
+      "step": 71425
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 2.1632396701972914,
+      "learning_rate": 4.897379598662208e-06,
+      "loss": 0.1728,
+      "step": 71450
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.8934608853543096,
+      "learning_rate": 4.89733779264214e-06,
+      "loss": 0.1585,
+      "step": 71475
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.8398068522285902,
+      "learning_rate": 4.897295986622074e-06,
+      "loss": 0.1697,
+      "step": 71500
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.8325654805684302,
+      "learning_rate": 4.8972541806020066e-06,
+      "loss": 0.1704,
+      "step": 71525
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 2.3199751539353652,
+      "learning_rate": 4.89721237458194e-06,
+      "loss": 0.161,
+      "step": 71550
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.147554620731119,
+      "learning_rate": 4.897170568561873e-06,
+      "loss": 0.1531,
+      "step": 71575
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 2.0353602477046118,
+      "learning_rate": 4.897128762541806e-06,
+      "loss": 0.163,
+      "step": 71600
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 2.371029050435062,
+      "learning_rate": 4.897086956521739e-06,
+      "loss": 0.1813,
+      "step": 71625
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 2.1762205039573836,
+      "learning_rate": 4.897045150501673e-06,
+      "loss": 0.1736,
+      "step": 71650
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.7956537476456738,
+      "learning_rate": 4.897003344481605e-06,
+      "loss": 0.1716,
+      "step": 71675
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 2.2524746958016393,
+      "learning_rate": 4.896961538461539e-06,
+      "loss": 0.1801,
+      "step": 71700
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 2.136139836516011,
+      "learning_rate": 4.8969197324414715e-06,
+      "loss": 0.1635,
+      "step": 71725
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 2.307095824383635,
+      "learning_rate": 4.896877926421405e-06,
+      "loss": 0.1487,
+      "step": 71750
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.9994033511834624,
+      "learning_rate": 4.896836120401338e-06,
+      "loss": 0.1698,
+      "step": 71775
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 2.0458057349311662,
+      "learning_rate": 4.896794314381271e-06,
+      "loss": 0.1664,
+      "step": 71800
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 2.7828406872393536,
+      "learning_rate": 4.896752508361204e-06,
+      "loss": 0.1835,
+      "step": 71825
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 2.3490321431100463,
+      "learning_rate": 4.8967107023411376e-06,
+      "loss": 0.1772,
+      "step": 71850
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 2.1239436779644456,
+      "learning_rate": 4.89666889632107e-06,
+      "loss": 0.1698,
+      "step": 71875
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 2.55078053521957,
+      "learning_rate": 4.896627090301004e-06,
+      "loss": 0.1865,
+      "step": 71900
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 2.0722007158937537,
+      "learning_rate": 4.896585284280937e-06,
+      "loss": 0.1534,
+      "step": 71925
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 3.2040867285924763,
+      "learning_rate": 4.89654347826087e-06,
+      "loss": 0.1615,
+      "step": 71950
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 2.2719001769767058,
+      "learning_rate": 4.896501672240804e-06,
+      "loss": 0.16,
+      "step": 71975
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 2.003118329948211,
+      "learning_rate": 4.896459866220736e-06,
+      "loss": 0.157,
+      "step": 72000
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 2.933295145712109,
+      "learning_rate": 4.89641806020067e-06,
+      "loss": 0.1549,
+      "step": 72025
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.89512122355957,
+      "learning_rate": 4.8963762541806025e-06,
+      "loss": 0.174,
+      "step": 72050
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.9653910668438528,
+      "learning_rate": 4.896334448160536e-06,
+      "loss": 0.1537,
+      "step": 72075
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.8822786208686735,
+      "learning_rate": 4.896292642140469e-06,
+      "loss": 0.1483,
+      "step": 72100
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.6974680148122512,
+      "learning_rate": 4.896250836120402e-06,
+      "loss": 0.1842,
+      "step": 72125
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 2.1232240441546324,
+      "learning_rate": 4.896209030100335e-06,
+      "loss": 0.1708,
+      "step": 72150
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 2.5147761693595014,
+      "learning_rate": 4.896167224080269e-06,
+      "loss": 0.1734,
+      "step": 72175
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.5814979109687117,
+      "learning_rate": 4.8961270903010036e-06,
+      "loss": 0.1753,
+      "step": 72200
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.9271487770120417,
+      "learning_rate": 4.896085284280937e-06,
+      "loss": 0.16,
+      "step": 72225
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.6807905811978427,
+      "learning_rate": 4.89604347826087e-06,
+      "loss": 0.1638,
+      "step": 72250
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 3.0101811846023243,
+      "learning_rate": 4.896001672240803e-06,
+      "loss": 0.1482,
+      "step": 72275
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 2.1706275732886335,
+      "learning_rate": 4.895959866220736e-06,
+      "loss": 0.1694,
+      "step": 72300
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.8417541316276584,
+      "learning_rate": 4.89591806020067e-06,
+      "loss": 0.1714,
+      "step": 72325
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.6630468895130261,
+      "learning_rate": 4.895876254180602e-06,
+      "loss": 0.1551,
+      "step": 72350
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 3.309966479941043,
+      "learning_rate": 4.895834448160536e-06,
+      "loss": 0.1719,
+      "step": 72375
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.7480764555332018,
+      "learning_rate": 4.8957926421404685e-06,
+      "loss": 0.1615,
+      "step": 72400
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.8651253589940062,
+      "learning_rate": 4.895750836120401e-06,
+      "loss": 0.1673,
+      "step": 72425
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.367768522432886,
+      "learning_rate": 4.895709030100335e-06,
+      "loss": 0.1614,
+      "step": 72450
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 2.47582620859887,
+      "learning_rate": 4.8956672240802675e-06,
+      "loss": 0.1607,
+      "step": 72475
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 3.023070237184101,
+      "learning_rate": 4.895625418060201e-06,
+      "loss": 0.1719,
+      "step": 72500
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 2.177541987436765,
+      "learning_rate": 4.895583612040134e-06,
+      "loss": 0.1612,
+      "step": 72525
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 2.257043109303166,
+      "learning_rate": 4.895541806020067e-06,
+      "loss": 0.1707,
+      "step": 72550
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 2.0137969172613395,
+      "learning_rate": 4.8955e-06,
+      "loss": 0.1643,
+      "step": 72575
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.9763477092821808,
+      "learning_rate": 4.8954581939799335e-06,
+      "loss": 0.1709,
+      "step": 72600
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.6234222767388702,
+      "learning_rate": 4.895416387959866e-06,
+      "loss": 0.1556,
+      "step": 72625
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.8976361029757107,
+      "learning_rate": 4.8953745819398e-06,
+      "loss": 0.1585,
+      "step": 72650
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.739742768865686,
+      "learning_rate": 4.8953327759197325e-06,
+      "loss": 0.1721,
+      "step": 72675
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 2.1472365854087303,
+      "learning_rate": 4.895290969899666e-06,
+      "loss": 0.1681,
+      "step": 72700
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 2.4942342308769527,
+      "learning_rate": 4.895249163879599e-06,
+      "loss": 0.1653,
+      "step": 72725
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 2.213462965668188,
+      "learning_rate": 4.895207357859532e-06,
+      "loss": 0.1533,
+      "step": 72750
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.757886571286951,
+      "learning_rate": 4.895165551839465e-06,
+      "loss": 0.1628,
+      "step": 72775
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 3.161382454759714,
+      "learning_rate": 4.8951237458193985e-06,
+      "loss": 0.1746,
+      "step": 72800
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 2.898760234748562,
+      "learning_rate": 4.895081939799331e-06,
+      "loss": 0.1644,
+      "step": 72825
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.9181331685697647,
+      "learning_rate": 4.895040133779265e-06,
+      "loss": 0.1744,
+      "step": 72850
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.2622034792223826,
+      "learning_rate": 4.8949983277591974e-06,
+      "loss": 0.1627,
+      "step": 72875
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 2.0311044399909335,
+      "learning_rate": 4.894956521739131e-06,
+      "loss": 0.1644,
+      "step": 72900
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.7264522011793653,
+      "learning_rate": 4.894914715719064e-06,
+      "loss": 0.162,
+      "step": 72925
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 2.5284533556414153,
+      "learning_rate": 4.894872909698997e-06,
+      "loss": 0.1708,
+      "step": 72950
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 2.3584564833935526,
+      "learning_rate": 4.89483110367893e-06,
+      "loss": 0.1553,
+      "step": 72975
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 2.6811106192462244,
+      "learning_rate": 4.8947892976588635e-06,
+      "loss": 0.1768,
+      "step": 73000
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 2.880359819740993,
+      "learning_rate": 4.894747491638796e-06,
+      "loss": 0.1638,
+      "step": 73025
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.8098189143936174,
+      "learning_rate": 4.89470568561873e-06,
+      "loss": 0.1694,
+      "step": 73050
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 2.1960220587817347,
+      "learning_rate": 4.894663879598663e-06,
+      "loss": 0.1529,
+      "step": 73075
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.879513441542289,
+      "learning_rate": 4.894622073578596e-06,
+      "loss": 0.1602,
+      "step": 73100
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 2.354208570461897,
+      "learning_rate": 4.894580267558529e-06,
+      "loss": 0.154,
+      "step": 73125
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.2171654991734855,
+      "learning_rate": 4.894538461538461e-06,
+      "loss": 0.1699,
+      "step": 73150
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 2.2940717027810593,
+      "learning_rate": 4.894496655518395e-06,
+      "loss": 0.1675,
+      "step": 73175
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.7813725219024068,
+      "learning_rate": 4.894456521739131e-06,
+      "loss": 0.1726,
+      "step": 73200
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.5471775285839873,
+      "learning_rate": 4.894414715719064e-06,
+      "loss": 0.1506,
+      "step": 73225
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 2.3179111030180164,
+      "learning_rate": 4.894372909698997e-06,
+      "loss": 0.1549,
+      "step": 73250
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 2.381683834465129,
+      "learning_rate": 4.8943311036789305e-06,
+      "loss": 0.1578,
+      "step": 73275
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 2.8932283117028095,
+      "learning_rate": 4.894289297658863e-06,
+      "loss": 0.1596,
+      "step": 73300
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 3.0652628362276526,
+      "learning_rate": 4.894247491638797e-06,
+      "loss": 0.1757,
+      "step": 73325
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 2.2515578022167846,
+      "learning_rate": 4.8942056856187295e-06,
+      "loss": 0.1531,
+      "step": 73350
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.9302780505682737,
+      "learning_rate": 4.894163879598662e-06,
+      "loss": 0.1751,
+      "step": 73375
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 2.3296661369560576,
+      "learning_rate": 4.894122073578596e-06,
+      "loss": 0.1579,
+      "step": 73400
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 2.317270323434963,
+      "learning_rate": 4.894080267558528e-06,
+      "loss": 0.1521,
+      "step": 73425
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.8791542432395285,
+      "learning_rate": 4.894038461538462e-06,
+      "loss": 0.1666,
+      "step": 73450
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 3.1141465597924354,
+      "learning_rate": 4.893996655518395e-06,
+      "loss": 0.1779,
+      "step": 73475
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.83550212334721,
+      "learning_rate": 4.893954849498328e-06,
+      "loss": 0.1573,
+      "step": 73500
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.869911326816695,
+      "learning_rate": 4.893913043478261e-06,
+      "loss": 0.1505,
+      "step": 73525
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.875677676542583,
+      "learning_rate": 4.8938712374581944e-06,
+      "loss": 0.1749,
+      "step": 73550
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 2.877401745073335,
+      "learning_rate": 4.893829431438127e-06,
+      "loss": 0.1756,
+      "step": 73575
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 2.8537228949707183,
+      "learning_rate": 4.893787625418061e-06,
+      "loss": 0.1553,
+      "step": 73600
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 3.132534812602467,
+      "learning_rate": 4.893745819397993e-06,
+      "loss": 0.1593,
+      "step": 73625
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.5367775675098951,
+      "learning_rate": 4.893704013377927e-06,
+      "loss": 0.1658,
+      "step": 73650
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 2.51031029076394,
+      "learning_rate": 4.89366220735786e-06,
+      "loss": 0.1599,
+      "step": 73675
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.9305209052598873,
+      "learning_rate": 4.893620401337793e-06,
+      "loss": 0.1642,
+      "step": 73700
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.8180335977528552,
+      "learning_rate": 4.893578595317726e-06,
+      "loss": 0.1697,
+      "step": 73725
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.7564659182371012,
+      "learning_rate": 4.893536789297659e-06,
+      "loss": 0.1408,
+      "step": 73750
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 2.707073171305364,
+      "learning_rate": 4.893494983277592e-06,
+      "loss": 0.1761,
+      "step": 73775
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 2.0197959277765496,
+      "learning_rate": 4.893453177257526e-06,
+      "loss": 0.1544,
+      "step": 73800
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.592459155442505,
+      "learning_rate": 4.893411371237458e-06,
+      "loss": 0.169,
+      "step": 73825
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 2.1128575826279286,
+      "learning_rate": 4.893369565217392e-06,
+      "loss": 0.1821,
+      "step": 73850
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 2.330641493994605,
+      "learning_rate": 4.893327759197325e-06,
+      "loss": 0.1599,
+      "step": 73875
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 2.289273202147065,
+      "learning_rate": 4.893285953177258e-06,
+      "loss": 0.1665,
+      "step": 73900
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 2.3967166690376724,
+      "learning_rate": 4.893244147157191e-06,
+      "loss": 0.1537,
+      "step": 73925
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.7436469370309389,
+      "learning_rate": 4.893202341137124e-06,
+      "loss": 0.1681,
+      "step": 73950
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 2.593160901601273,
+      "learning_rate": 4.893160535117057e-06,
+      "loss": 0.1549,
+      "step": 73975
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.7292694124000343,
+      "learning_rate": 4.893118729096991e-06,
+      "loss": 0.1704,
+      "step": 74000
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 2.7513280274403336,
+      "learning_rate": 4.893076923076923e-06,
+      "loss": 0.1762,
+      "step": 74025
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 2.245015646917485,
+      "learning_rate": 4.893035117056857e-06,
+      "loss": 0.1545,
+      "step": 74050
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 2.0277778975395058,
+      "learning_rate": 4.89299331103679e-06,
+      "loss": 0.1585,
+      "step": 74075
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 2.6304291462822813,
+      "learning_rate": 4.892951505016722e-06,
+      "loss": 0.1635,
+      "step": 74100
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 2.1099036734597085,
+      "learning_rate": 4.892909698996656e-06,
+      "loss": 0.165,
+      "step": 74125
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 2.3841230259715886,
+      "learning_rate": 4.8928678929765885e-06,
+      "loss": 0.1541,
+      "step": 74150
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.4945382544604393,
+      "learning_rate": 4.892826086956522e-06,
+      "loss": 0.159,
+      "step": 74175
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 3.160768317692789,
+      "learning_rate": 4.892785953177258e-06,
+      "loss": 0.1554,
+      "step": 74200
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 2.307190734455112,
+      "learning_rate": 4.8927441471571914e-06,
+      "loss": 0.1572,
+      "step": 74225
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 2.5864605527863525,
+      "learning_rate": 4.892702341137124e-06,
+      "loss": 0.1478,
+      "step": 74250
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 2.3159481412550114,
+      "learning_rate": 4.892660535117058e-06,
+      "loss": 0.1712,
+      "step": 74275
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.7017971769327145,
+      "learning_rate": 4.89261872909699e-06,
+      "loss": 0.1576,
+      "step": 74300
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 2.916033378211683,
+      "learning_rate": 4.892576923076923e-06,
+      "loss": 0.1415,
+      "step": 74325
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 2.589428720427529,
+      "learning_rate": 4.892535117056857e-06,
+      "loss": 0.1514,
+      "step": 74350
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 2.4781537139443732,
+      "learning_rate": 4.892493311036789e-06,
+      "loss": 0.1753,
+      "step": 74375
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 2.087506703608687,
+      "learning_rate": 4.892451505016723e-06,
+      "loss": 0.1665,
+      "step": 74400
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.4517621145289306,
+      "learning_rate": 4.8924096989966556e-06,
+      "loss": 0.1673,
+      "step": 74425
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.549222778728874,
+      "learning_rate": 4.892367892976589e-06,
+      "loss": 0.1623,
+      "step": 74450
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 2.6905760289098284,
+      "learning_rate": 4.892326086956522e-06,
+      "loss": 0.1688,
+      "step": 74475
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.7294195336741673,
+      "learning_rate": 4.892284280936455e-06,
+      "loss": 0.1792,
+      "step": 74500
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 2.2205866806833905,
+      "learning_rate": 4.892242474916388e-06,
+      "loss": 0.1703,
+      "step": 74525
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.512048417276666,
+      "learning_rate": 4.892200668896322e-06,
+      "loss": 0.1727,
+      "step": 74550
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.6165734990093756,
+      "learning_rate": 4.892158862876254e-06,
+      "loss": 0.1689,
+      "step": 74575
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.7105503061939633,
+      "learning_rate": 4.892117056856188e-06,
+      "loss": 0.1645,
+      "step": 74600
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 2.241186103825974,
+      "learning_rate": 4.8920752508361206e-06,
+      "loss": 0.174,
+      "step": 74625
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 3.0478956139162046,
+      "learning_rate": 4.892033444816054e-06,
+      "loss": 0.1732,
+      "step": 74650
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 2.2023056589981613,
+      "learning_rate": 4.891991638795987e-06,
+      "loss": 0.1704,
+      "step": 74675
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 2.1750560553924156,
+      "learning_rate": 4.89194983277592e-06,
+      "loss": 0.1757,
+      "step": 74700
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.7338152881128504,
+      "learning_rate": 4.891908026755853e-06,
+      "loss": 0.1625,
+      "step": 74725
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.5947374991763381,
+      "learning_rate": 4.891866220735787e-06,
+      "loss": 0.1742,
+      "step": 74750
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.9887138661903652,
+      "learning_rate": 4.891824414715719e-06,
+      "loss": 0.1811,
+      "step": 74775
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 2.436010532096877,
+      "learning_rate": 4.891782608695653e-06,
+      "loss": 0.1537,
+      "step": 74800
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 2.926286576820829,
+      "learning_rate": 4.8917408026755855e-06,
+      "loss": 0.16,
+      "step": 74825
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.824564621763424,
+      "learning_rate": 4.891698996655519e-06,
+      "loss": 0.1629,
+      "step": 74850
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 2.683636286974673,
+      "learning_rate": 4.891657190635452e-06,
+      "loss": 0.1557,
+      "step": 74875
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 2.1385204177371224,
+      "learning_rate": 4.891615384615385e-06,
+      "loss": 0.1675,
+      "step": 74900
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.5571459930052227,
+      "learning_rate": 4.891573578595318e-06,
+      "loss": 0.1468,
+      "step": 74925
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 2.19393355253032,
+      "learning_rate": 4.8915317725752516e-06,
+      "loss": 0.1668,
+      "step": 74950
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 2.1955950411829646,
+      "learning_rate": 4.891489966555184e-06,
+      "loss": 0.1705,
+      "step": 74975
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.8842544208995302,
+      "learning_rate": 4.891448160535118e-06,
+      "loss": 0.1853,
+      "step": 75000
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.9456707154959463,
+      "learning_rate": 4.8914063545150505e-06,
+      "loss": 0.1676,
+      "step": 75025
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.720712384593594,
+      "learning_rate": 4.891364548494983e-06,
+      "loss": 0.1691,
+      "step": 75050
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 2.732269103665272,
+      "learning_rate": 4.891322742474917e-06,
+      "loss": 0.157,
+      "step": 75075
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 2.093993948698275,
+      "learning_rate": 4.8912809364548494e-06,
+      "loss": 0.1579,
+      "step": 75100
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 2.2596660812012637,
+      "learning_rate": 4.891239130434783e-06,
+      "loss": 0.1612,
+      "step": 75125
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.607362977009435,
+      "learning_rate": 4.891197324414716e-06,
+      "loss": 0.1539,
+      "step": 75150
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.750020200184701,
+      "learning_rate": 4.891155518394649e-06,
+      "loss": 0.1653,
+      "step": 75175
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 2.65398912558923,
+      "learning_rate": 4.891115384615385e-06,
+      "loss": 0.1536,
+      "step": 75200
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 2.8119697714667136,
+      "learning_rate": 4.891073578595319e-06,
+      "loss": 0.1533,
+      "step": 75225
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 2.249098596865318,
+      "learning_rate": 4.891031772575251e-06,
+      "loss": 0.1636,
+      "step": 75250
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 2.380834954836527,
+      "learning_rate": 4.890989966555184e-06,
+      "loss": 0.1592,
+      "step": 75275
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 2.074689075288214,
+      "learning_rate": 4.8909481605351176e-06,
+      "loss": 0.1662,
+      "step": 75300
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 2.2019387377694404,
+      "learning_rate": 4.89090635451505e-06,
+      "loss": 0.1697,
+      "step": 75325
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 2.4940970023751756,
+      "learning_rate": 4.890864548494984e-06,
+      "loss": 0.1524,
+      "step": 75350
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 3.0020918616337213,
+      "learning_rate": 4.8908227424749165e-06,
+      "loss": 0.167,
+      "step": 75375
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 2.04633813344427,
+      "learning_rate": 4.89078093645485e-06,
+      "loss": 0.1857,
+      "step": 75400
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.8578961402822785,
+      "learning_rate": 4.890739130434783e-06,
+      "loss": 0.1712,
+      "step": 75425
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 2.568127490750955,
+      "learning_rate": 4.890697324414716e-06,
+      "loss": 0.1824,
+      "step": 75450
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.7273995103916189,
+      "learning_rate": 4.890655518394649e-06,
+      "loss": 0.1633,
+      "step": 75475
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.8705828417286539,
+      "learning_rate": 4.8906137123745825e-06,
+      "loss": 0.1555,
+      "step": 75500
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.2009745533949896,
+      "learning_rate": 4.890571906354515e-06,
+      "loss": 0.1711,
+      "step": 75525
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 2.4521956369026072,
+      "learning_rate": 4.890530100334449e-06,
+      "loss": 0.1702,
+      "step": 75550
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.6330318191278905,
+      "learning_rate": 4.8904882943143815e-06,
+      "loss": 0.1622,
+      "step": 75575
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.8895480021883273,
+      "learning_rate": 4.890446488294315e-06,
+      "loss": 0.1484,
+      "step": 75600
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 2.990507862504963,
+      "learning_rate": 4.890404682274248e-06,
+      "loss": 0.1704,
+      "step": 75625
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 2.254372856378548,
+      "learning_rate": 4.890362876254181e-06,
+      "loss": 0.1781,
+      "step": 75650
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 2.038112544222414,
+      "learning_rate": 4.890321070234114e-06,
+      "loss": 0.1593,
+      "step": 75675
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 3.040909300567587,
+      "learning_rate": 4.8902792642140475e-06,
+      "loss": 0.1729,
+      "step": 75700
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 2.7786710261443406,
+      "learning_rate": 4.89023745819398e-06,
+      "loss": 0.1461,
+      "step": 75725
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 2.3393555386740212,
+      "learning_rate": 4.890195652173914e-06,
+      "loss": 0.1595,
+      "step": 75750
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 5.199257311504345,
+      "learning_rate": 4.8901538461538465e-06,
+      "loss": 0.1753,
+      "step": 75775
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.703194333068814,
+      "learning_rate": 4.89011204013378e-06,
+      "loss": 0.1589,
+      "step": 75800
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 2.780370147699592,
+      "learning_rate": 4.890070234113713e-06,
+      "loss": 0.1715,
+      "step": 75825
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 2.3580394840206327,
+      "learning_rate": 4.890028428093646e-06,
+      "loss": 0.1559,
+      "step": 75850
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 2.1728983607540133,
+      "learning_rate": 4.889986622073579e-06,
+      "loss": 0.1591,
+      "step": 75875
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 2.277580980241663,
+      "learning_rate": 4.8899448160535125e-06,
+      "loss": 0.1659,
+      "step": 75900
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 2.3986761264261127,
+      "learning_rate": 4.889903010033445e-06,
+      "loss": 0.1672,
+      "step": 75925
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.931152925984652,
+      "learning_rate": 4.889861204013378e-06,
+      "loss": 0.1657,
+      "step": 75950
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.68257502997896,
+      "learning_rate": 4.8898193979933114e-06,
+      "loss": 0.1614,
+      "step": 75975
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.470089170915271,
+      "learning_rate": 4.889777591973244e-06,
+      "loss": 0.1782,
+      "step": 76000
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 3.2551584773611544,
+      "learning_rate": 4.889735785953178e-06,
+      "loss": 0.1503,
+      "step": 76025
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 2.1093960304591755,
+      "learning_rate": 4.88969397993311e-06,
+      "loss": 0.1657,
+      "step": 76050
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 2.672877955706056,
+      "learning_rate": 4.889652173913044e-06,
+      "loss": 0.1649,
+      "step": 76075
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 3.1470669315639115,
+      "learning_rate": 4.889610367892977e-06,
+      "loss": 0.1746,
+      "step": 76100
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 3.0579894482513397,
+      "learning_rate": 4.88956856187291e-06,
+      "loss": 0.1657,
+      "step": 76125
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 2.947301172535142,
+      "learning_rate": 4.889526755852843e-06,
+      "loss": 0.1516,
+      "step": 76150
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 3.3863874587051725,
+      "learning_rate": 4.889484949832776e-06,
+      "loss": 0.1697,
+      "step": 76175
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.9291008913623715,
+      "learning_rate": 4.889444816053512e-06,
+      "loss": 0.1776,
+      "step": 76200
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 2.319491892854814,
+      "learning_rate": 4.889403010033445e-06,
+      "loss": 0.1781,
+      "step": 76225
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 3.488180039037966,
+      "learning_rate": 4.889362876254181e-06,
+      "loss": 0.1857,
+      "step": 76250
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 2.412768379326752,
+      "learning_rate": 4.889321070234114e-06,
+      "loss": 0.1438,
+      "step": 76275
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 3.8203948129859513,
+      "learning_rate": 4.889279264214047e-06,
+      "loss": 0.1328,
+      "step": 76300
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 2.8638845252884635,
+      "learning_rate": 4.8892374581939806e-06,
+      "loss": 0.1195,
+      "step": 76325
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 3.0186144490512676,
+      "learning_rate": 4.889195652173913e-06,
+      "loss": 0.1231,
+      "step": 76350
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 2.40564050431268,
+      "learning_rate": 4.889153846153847e-06,
+      "loss": 0.1288,
+      "step": 76375
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 3.223902399031433,
+      "learning_rate": 4.88911204013378e-06,
+      "loss": 0.1253,
+      "step": 76400
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 2.524120816578796,
+      "learning_rate": 4.889070234113712e-06,
+      "loss": 0.1186,
+      "step": 76425
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 4.33608755350878,
+      "learning_rate": 4.889028428093646e-06,
+      "loss": 0.1302,
+      "step": 76450
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 3.0812374326190763,
+      "learning_rate": 4.8889866220735784e-06,
+      "loss": 0.1154,
+      "step": 76475
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 3.8611643250629357,
+      "learning_rate": 4.888944816053512e-06,
+      "loss": 0.1182,
+      "step": 76500
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 4.562201565976749,
+      "learning_rate": 4.888903010033445e-06,
+      "loss": 0.1163,
+      "step": 76525
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 5.0482039643712335,
+      "learning_rate": 4.888861204013378e-06,
+      "loss": 0.1294,
+      "step": 76550
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 2.7435608656097723,
+      "learning_rate": 4.888819397993311e-06,
+      "loss": 0.1384,
+      "step": 76575
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 2.2602926708641857,
+      "learning_rate": 4.8887775919732445e-06,
+      "loss": 0.1116,
+      "step": 76600
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 5.6874792345843135,
+      "learning_rate": 4.888735785953177e-06,
+      "loss": 0.1224,
+      "step": 76625
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 2.720275109336373,
+      "learning_rate": 4.888693979933111e-06,
+      "loss": 0.123,
+      "step": 76650
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 4.395574166086428,
+      "learning_rate": 4.888652173913043e-06,
+      "loss": 0.1168,
+      "step": 76675
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 2.703916615126388,
+      "learning_rate": 4.888610367892977e-06,
+      "loss": 0.1214,
+      "step": 76700
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 3.181411929507769,
+      "learning_rate": 4.88856856187291e-06,
+      "loss": 0.1138,
+      "step": 76725
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 3.509556295339099,
+      "learning_rate": 4.888526755852843e-06,
+      "loss": 0.1232,
+      "step": 76750
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 3.8902899973523626,
+      "learning_rate": 4.888484949832776e-06,
+      "loss": 0.1212,
+      "step": 76775
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 2.0686794778491895,
+      "learning_rate": 4.8884431438127095e-06,
+      "loss": 0.123,
+      "step": 76800
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 3.16510341619209,
+      "learning_rate": 4.888401337792643e-06,
+      "loss": 0.1183,
+      "step": 76825
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 4.458808729537471,
+      "learning_rate": 4.888359531772576e-06,
+      "loss": 0.1277,
+      "step": 76850
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 2.6194750814850947,
+      "learning_rate": 4.888317725752509e-06,
+      "loss": 0.1167,
+      "step": 76875
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 2.4407102227373794,
+      "learning_rate": 4.888275919732442e-06,
+      "loss": 0.1298,
+      "step": 76900
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 2.6333427547264407,
+      "learning_rate": 4.8882341137123755e-06,
+      "loss": 0.1372,
+      "step": 76925
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 2.4980445842112227,
+      "learning_rate": 4.888192307692308e-06,
+      "loss": 0.1213,
+      "step": 76950
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 3.410736806496833,
+      "learning_rate": 4.888150501672242e-06,
+      "loss": 0.1324,
+      "step": 76975
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 1.9758670528816349,
+      "learning_rate": 4.8881086956521744e-06,
+      "loss": 0.1302,
+      "step": 77000
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 3.4608232211672334,
+      "learning_rate": 4.888066889632108e-06,
+      "loss": 0.1157,
+      "step": 77025
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 3.9971640082189137,
+      "learning_rate": 4.888025083612041e-06,
+      "loss": 0.1116,
+      "step": 77050
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 3.7059743493192046,
+      "learning_rate": 4.887983277591974e-06,
+      "loss": 0.1168,
+      "step": 77075
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 3.525869534264302,
+      "learning_rate": 4.887941471571907e-06,
+      "loss": 0.1272,
+      "step": 77100
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 3.3799153930239405,
+      "learning_rate": 4.88789966555184e-06,
+      "loss": 0.1373,
+      "step": 77125
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 2.979968630998421,
+      "learning_rate": 4.887857859531772e-06,
+      "loss": 0.1173,
+      "step": 77150
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 3.116452957430253,
+      "learning_rate": 4.887816053511706e-06,
+      "loss": 0.1313,
+      "step": 77175
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 4.502809671470845,
+      "learning_rate": 4.8877742474916386e-06,
+      "loss": 0.1211,
+      "step": 77200
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 3.3430787138363804,
+      "learning_rate": 4.887732441471572e-06,
+      "loss": 0.1137,
+      "step": 77225
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 3.4815832417449935,
+      "learning_rate": 4.887690635451506e-06,
+      "loss": 0.124,
+      "step": 77250
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 2.8682967863246294,
+      "learning_rate": 4.887648829431438e-06,
+      "loss": 0.1292,
+      "step": 77275
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 3.5200313476756526,
+      "learning_rate": 4.887607023411372e-06,
+      "loss": 0.1175,
+      "step": 77300
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 3.7267567794932743,
+      "learning_rate": 4.887565217391305e-06,
+      "loss": 0.1252,
+      "step": 77325
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 3.9971687272656258,
+      "learning_rate": 4.887523411371238e-06,
+      "loss": 0.1274,
+      "step": 77350
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 4.970670446361493,
+      "learning_rate": 4.887481605351171e-06,
+      "loss": 0.1329,
+      "step": 77375
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 2.4841506585451643,
+      "learning_rate": 4.887439799331104e-06,
+      "loss": 0.1214,
+      "step": 77400
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 4.295568567553117,
+      "learning_rate": 4.887397993311037e-06,
+      "loss": 0.1232,
+      "step": 77425
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 4.063655366491959,
+      "learning_rate": 4.887356187290971e-06,
+      "loss": 0.1326,
+      "step": 77450
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 2.8435059530091262,
+      "learning_rate": 4.887314381270903e-06,
+      "loss": 0.1242,
+      "step": 77475
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 2.1310397377270602,
+      "learning_rate": 4.887272575250837e-06,
+      "loss": 0.1181,
+      "step": 77500
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 3.3838400346860897,
+      "learning_rate": 4.8872307692307696e-06,
+      "loss": 0.1243,
+      "step": 77525
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 3.642961468985254,
+      "learning_rate": 4.887188963210703e-06,
+      "loss": 0.1402,
+      "step": 77550
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 2.593211405878639,
+      "learning_rate": 4.887147157190636e-06,
+      "loss": 0.1219,
+      "step": 77575
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 4.739286033182659,
+      "learning_rate": 4.887105351170569e-06,
+      "loss": 0.1313,
+      "step": 77600
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 3.05581489965152,
+      "learning_rate": 4.887063545150502e-06,
+      "loss": 0.1201,
+      "step": 77625
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 3.0899269653481745,
+      "learning_rate": 4.887021739130436e-06,
+      "loss": 0.1247,
+      "step": 77650
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 3.758598812637565,
+      "learning_rate": 4.886979933110368e-06,
+      "loss": 0.1213,
+      "step": 77675
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 3.6037831919337875,
+      "learning_rate": 4.886938127090302e-06,
+      "loss": 0.128,
+      "step": 77700
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 4.451175536710578,
+      "learning_rate": 4.8868963210702345e-06,
+      "loss": 0.1311,
+      "step": 77725
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 3.09548461829267,
+      "learning_rate": 4.886854515050168e-06,
+      "loss": 0.1207,
+      "step": 77750
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 3.689827710615845,
+      "learning_rate": 4.886812709030101e-06,
+      "loss": 0.1152,
+      "step": 77775
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 3.7771578128301226,
+      "learning_rate": 4.886770903010034e-06,
+      "loss": 0.1242,
+      "step": 77800
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 4.031573600779883,
+      "learning_rate": 4.886729096989967e-06,
+      "loss": 0.1324,
+      "step": 77825
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 3.426503558716008,
+      "learning_rate": 4.8866872909699e-06,
+      "loss": 0.1173,
+      "step": 77850
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 2.4533360047703456,
+      "learning_rate": 4.886645484949833e-06,
+      "loss": 0.1177,
+      "step": 77875
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 3.6081638322635046,
+      "learning_rate": 4.886603678929766e-06,
+      "loss": 0.1318,
+      "step": 77900
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 3.651931508167466,
+      "learning_rate": 4.8865618729096995e-06,
+      "loss": 0.1278,
+      "step": 77925
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 3.133415821529739,
+      "learning_rate": 4.886520066889632e-06,
+      "loss": 0.1227,
+      "step": 77950
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 2.6432438963375477,
+      "learning_rate": 4.886478260869566e-06,
+      "loss": 0.1226,
+      "step": 77975
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 4.072468610495336,
+      "learning_rate": 4.8864364548494985e-06,
+      "loss": 0.1336,
+      "step": 78000
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 3.607977716821972,
+      "learning_rate": 4.886394648829432e-06,
+      "loss": 0.1259,
+      "step": 78025
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 3.192830764597763,
+      "learning_rate": 4.886352842809365e-06,
+      "loss": 0.1216,
+      "step": 78050
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 3.583472664299709,
+      "learning_rate": 4.886311036789298e-06,
+      "loss": 0.1296,
+      "step": 78075
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 3.1812193947409777,
+      "learning_rate": 4.886269230769231e-06,
+      "loss": 0.1129,
+      "step": 78100
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 3.5681957955883603,
+      "learning_rate": 4.8862274247491645e-06,
+      "loss": 0.1257,
+      "step": 78125
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 4.3655757139060585,
+      "learning_rate": 4.886185618729097e-06,
+      "loss": 0.1176,
+      "step": 78150
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 3.627286736907382,
+      "learning_rate": 4.886143812709031e-06,
+      "loss": 0.128,
+      "step": 78175
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 4.110026082315242,
+      "learning_rate": 4.8861020066889634e-06,
+      "loss": 0.1314,
+      "step": 78200
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 3.540369512342808,
+      "learning_rate": 4.886060200668897e-06,
+      "loss": 0.1242,
+      "step": 78225
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 3.0234131517149936,
+      "learning_rate": 4.886020066889632e-06,
+      "loss": 0.1132,
+      "step": 78250
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 3.4765331389271035,
+      "learning_rate": 4.8859782608695655e-06,
+      "loss": 0.1241,
+      "step": 78275
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 3.7796806155271914,
+      "learning_rate": 4.885936454849498e-06,
+      "loss": 0.1158,
+      "step": 78300
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 2.8841477452018838,
+      "learning_rate": 4.885894648829432e-06,
+      "loss": 0.1493,
+      "step": 78325
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 4.4433914203454625,
+      "learning_rate": 4.885852842809365e-06,
+      "loss": 0.1196,
+      "step": 78350
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 2.75884085554516,
+      "learning_rate": 4.885811036789298e-06,
+      "loss": 0.1188,
+      "step": 78375
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 4.273808052650839,
+      "learning_rate": 4.8857692307692316e-06,
+      "loss": 0.1217,
+      "step": 78400
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 3.897334135564084,
+      "learning_rate": 4.885727424749164e-06,
+      "loss": 0.1286,
+      "step": 78425
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 4.615939251349839,
+      "learning_rate": 4.885685618729098e-06,
+      "loss": 0.1362,
+      "step": 78450
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 3.730304960317534,
+      "learning_rate": 4.8856438127090305e-06,
+      "loss": 0.1344,
+      "step": 78475
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 3.5702453678071944,
+      "learning_rate": 4.885602006688964e-06,
+      "loss": 0.1254,
+      "step": 78500
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 2.9459198258987245,
+      "learning_rate": 4.885560200668897e-06,
+      "loss": 0.1155,
+      "step": 78525
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 2.9422269730368202,
+      "learning_rate": 4.88551839464883e-06,
+      "loss": 0.1341,
+      "step": 78550
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 4.345971398090705,
+      "learning_rate": 4.885476588628763e-06,
+      "loss": 0.125,
+      "step": 78575
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 3.4203734766216543,
+      "learning_rate": 4.8854347826086965e-06,
+      "loss": 0.1219,
+      "step": 78600
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 2.7450396553454763,
+      "learning_rate": 4.885392976588629e-06,
+      "loss": 0.1273,
+      "step": 78625
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 2.6718974934824007,
+      "learning_rate": 4.885351170568563e-06,
+      "loss": 0.1431,
+      "step": 78650
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 2.853838303655648,
+      "learning_rate": 4.8853093645484955e-06,
+      "loss": 0.125,
+      "step": 78675
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 2.9709753361358064,
+      "learning_rate": 4.885267558528429e-06,
+      "loss": 0.1244,
+      "step": 78700
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 4.023857278065838,
+      "learning_rate": 4.885225752508362e-06,
+      "loss": 0.1293,
+      "step": 78725
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 4.373864745184071,
+      "learning_rate": 4.885183946488295e-06,
+      "loss": 0.1319,
+      "step": 78750
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 2.508205959991612,
+      "learning_rate": 4.885142140468228e-06,
+      "loss": 0.1273,
+      "step": 78775
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 2.6411187927711772,
+      "learning_rate": 4.885100334448161e-06,
+      "loss": 0.1311,
+      "step": 78800
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 3.3179749804960834,
+      "learning_rate": 4.885058528428094e-06,
+      "loss": 0.1206,
+      "step": 78825
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 3.7366552197499647,
+      "learning_rate": 4.885016722408027e-06,
+      "loss": 0.1246,
+      "step": 78850
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 3.118992258293641,
+      "learning_rate": 4.8849749163879605e-06,
+      "loss": 0.1126,
+      "step": 78875
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 2.056430415904338,
+      "learning_rate": 4.884933110367893e-06,
+      "loss": 0.1069,
+      "step": 78900
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 3.785563939242501,
+      "learning_rate": 4.884891304347827e-06,
+      "loss": 0.1279,
+      "step": 78925
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 3.522011901121123,
+      "learning_rate": 4.884849498327759e-06,
+      "loss": 0.1221,
+      "step": 78950
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 2.815780165397175,
+      "learning_rate": 4.884807692307693e-06,
+      "loss": 0.1252,
+      "step": 78975
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 3.6460270553802965,
+      "learning_rate": 4.884765886287626e-06,
+      "loss": 0.1291,
+      "step": 79000
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 3.267409610114439,
+      "learning_rate": 4.884724080267559e-06,
+      "loss": 0.131,
+      "step": 79025
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 3.512502522504111,
+      "learning_rate": 4.884682274247492e-06,
+      "loss": 0.1193,
+      "step": 79050
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 4.876716295837919,
+      "learning_rate": 4.8846404682274254e-06,
+      "loss": 0.1279,
+      "step": 79075
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 4.068141572812643,
+      "learning_rate": 4.884598662207358e-06,
+      "loss": 0.1191,
+      "step": 79100
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 2.57695746518627,
+      "learning_rate": 4.884556856187292e-06,
+      "loss": 0.1218,
+      "step": 79125
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 2.8362615333679826,
+      "learning_rate": 4.884515050167224e-06,
+      "loss": 0.1169,
+      "step": 79150
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 3.7181091078142487,
+      "learning_rate": 4.884473244147158e-06,
+      "loss": 0.1271,
+      "step": 79175
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 2.562858519582811,
+      "learning_rate": 4.884431438127091e-06,
+      "loss": 0.132,
+      "step": 79200
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 3.226476454470988,
+      "learning_rate": 4.884389632107024e-06,
+      "loss": 0.1202,
+      "step": 79225
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 3.251373898366226,
+      "learning_rate": 4.884349498327759e-06,
+      "loss": 0.1468,
+      "step": 79250
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 3.198441179981739,
+      "learning_rate": 4.884307692307693e-06,
+      "loss": 0.1273,
+      "step": 79275
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 3.0629650236020756,
+      "learning_rate": 4.884265886287625e-06,
+      "loss": 0.1231,
+      "step": 79300
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 3.2827139225909976,
+      "learning_rate": 4.884224080267559e-06,
+      "loss": 0.1404,
+      "step": 79325
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 3.144707188115353,
+      "learning_rate": 4.884182274247492e-06,
+      "loss": 0.1298,
+      "step": 79350
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 2.9421101700328185,
+      "learning_rate": 4.884140468227425e-06,
+      "loss": 0.1405,
+      "step": 79375
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 4.59682086895325,
+      "learning_rate": 4.884098662207358e-06,
+      "loss": 0.1441,
+      "step": 79400
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 3.391160492130251,
+      "learning_rate": 4.884056856187291e-06,
+      "loss": 0.1375,
+      "step": 79425
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 3.180088795295559,
+      "learning_rate": 4.884015050167225e-06,
+      "loss": 0.1287,
+      "step": 79450
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 4.570107888816917,
+      "learning_rate": 4.883973244147158e-06,
+      "loss": 0.1255,
+      "step": 79475
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 3.0789019235619275,
+      "learning_rate": 4.883931438127091e-06,
+      "loss": 0.1254,
+      "step": 79500
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 3.5793084776409003,
+      "learning_rate": 4.883889632107024e-06,
+      "loss": 0.1181,
+      "step": 79525
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 3.797918053156867,
+      "learning_rate": 4.8838478260869575e-06,
+      "loss": 0.1248,
+      "step": 79550
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 3.027925691307738,
+      "learning_rate": 4.88380602006689e-06,
+      "loss": 0.1218,
+      "step": 79575
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 5.0368441312464,
+      "learning_rate": 4.883764214046824e-06,
+      "loss": 0.1157,
+      "step": 79600
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 3.8288632838026757,
+      "learning_rate": 4.883722408026756e-06,
+      "loss": 0.1156,
+      "step": 79625
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 2.597067146027184,
+      "learning_rate": 4.88368060200669e-06,
+      "loss": 0.1297,
+      "step": 79650
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 3.500369841183393,
+      "learning_rate": 4.883638795986623e-06,
+      "loss": 0.1239,
+      "step": 79675
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 2.9150129908770857,
+      "learning_rate": 4.883596989966556e-06,
+      "loss": 0.1264,
+      "step": 79700
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 3.720662911601689,
+      "learning_rate": 4.883555183946488e-06,
+      "loss": 0.1289,
+      "step": 79725
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 3.340766578368334,
+      "learning_rate": 4.883513377926422e-06,
+      "loss": 0.1416,
+      "step": 79750
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 3.3314964950611614,
+      "learning_rate": 4.883471571906354e-06,
+      "loss": 0.1245,
+      "step": 79775
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 4.481987339852263,
+      "learning_rate": 4.883429765886288e-06,
+      "loss": 0.1423,
+      "step": 79800
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 2.6168443489008637,
+      "learning_rate": 4.8833879598662205e-06,
+      "loss": 0.1343,
+      "step": 79825
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 3.9455102466510907,
+      "learning_rate": 4.883346153846154e-06,
+      "loss": 0.1227,
+      "step": 79850
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 2.6209413472131025,
+      "learning_rate": 4.883304347826088e-06,
+      "loss": 0.1405,
+      "step": 79875
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 4.187224029783121,
+      "learning_rate": 4.88326254180602e-06,
+      "loss": 0.1346,
+      "step": 79900
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 3.9706536290794507,
+      "learning_rate": 4.883220735785954e-06,
+      "loss": 0.1334,
+      "step": 79925
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 3.1222225652663607,
+      "learning_rate": 4.8831789297658866e-06,
+      "loss": 0.1165,
+      "step": 79950
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 3.1495103383983034,
+      "learning_rate": 4.88313712374582e-06,
+      "loss": 0.119,
+      "step": 79975
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 3.5192842049873176,
+      "learning_rate": 4.883095317725753e-06,
+      "loss": 0.121,
+      "step": 80000
+    },
+    {
+      "epoch": 3.15,
+      "eval_loss": 0.5361328125,
+      "eval_runtime": 11580.0151,
+      "eval_samples_per_second": 0.818,
+      "eval_steps_per_second": 0.051,
+      "eval_wer": 0.11300934639576597,
+      "step": 80000
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 3.382723320522212,
+      "learning_rate": 1.4347222222222223e-07,
+      "loss": 0.1257,
+      "step": 80025
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 3.7499859028087914,
+      "learning_rate": 1.417361111111111e-07,
+      "loss": 0.1128,
+      "step": 80050
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 2.9592785016890195,
+      "learning_rate": 1.4e-07,
+      "loss": 0.1195,
+      "step": 80075
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 3.4893216360464137,
+      "learning_rate": 1.3826388888888889e-07,
+      "loss": 0.1313,
+      "step": 80100
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 3.719554257846966,
+      "learning_rate": 1.365277777777778e-07,
+      "loss": 0.1113,
+      "step": 80125
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 3.3150805950550217,
+      "learning_rate": 1.347916666666667e-07,
+      "loss": 0.124,
+      "step": 80150
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 3.810734588250568,
+      "learning_rate": 1.3305555555555557e-07,
+      "loss": 0.1217,
+      "step": 80175
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 2.616845456697401,
+      "learning_rate": 1.3131944444444444e-07,
+      "loss": 0.1277,
+      "step": 80200
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 3.2366291515709578,
+      "learning_rate": 1.2958333333333335e-07,
+      "loss": 0.1152,
+      "step": 80225
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 3.7270587123667935,
+      "learning_rate": 1.2791666666666668e-07,
+      "loss": 0.1214,
+      "step": 80250
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 3.79323438662421,
+      "learning_rate": 1.2618055555555558e-07,
+      "loss": 0.1164,
+      "step": 80275
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 3.388340458302277,
+      "learning_rate": 1.2444444444444446e-07,
+      "loss": 0.1053,
+      "step": 80300
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 3.0068550159409613,
+      "learning_rate": 1.2270833333333333e-07,
+      "loss": 0.1363,
+      "step": 80325
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 2.6705820472591024,
+      "learning_rate": 1.2097222222222223e-07,
+      "loss": 0.1184,
+      "step": 80350
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 3.6711884295544013,
+      "learning_rate": 1.192361111111111e-07,
+      "loss": 0.1183,
+      "step": 80375
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 4.303935705314459,
+      "learning_rate": 1.1750000000000001e-07,
+      "loss": 0.1312,
+      "step": 80400
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 3.0047973076097407,
+      "learning_rate": 1.157638888888889e-07,
+      "loss": 0.1308,
+      "step": 80425
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 3.809379593211185,
+      "learning_rate": 1.1402777777777778e-07,
+      "loss": 0.1167,
+      "step": 80450
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 2.9141385074256534,
+      "learning_rate": 1.1229166666666668e-07,
+      "loss": 0.1215,
+      "step": 80475
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 3.6712555964990172,
+      "learning_rate": 1.1055555555555557e-07,
+      "loss": 0.1307,
+      "step": 80500
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 2.821579337006869,
+      "learning_rate": 1.0881944444444445e-07,
+      "loss": 0.1114,
+      "step": 80525
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 3.7286248730869658,
+      "learning_rate": 1.0708333333333335e-07,
+      "loss": 0.1346,
+      "step": 80550
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 3.0656348499738315,
+      "learning_rate": 1.0534722222222223e-07,
+      "loss": 0.1149,
+      "step": 80575
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 4.331509026966981,
+      "learning_rate": 1.0361111111111111e-07,
+      "loss": 0.1107,
+      "step": 80600
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 1.8774186348904527,
+      "learning_rate": 1.0187500000000002e-07,
+      "loss": 0.1117,
+      "step": 80625
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 2.706357801131824,
+      "learning_rate": 1.0013888888888889e-07,
+      "loss": 0.1129,
+      "step": 80650
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 3.5291983010831305,
+      "learning_rate": 9.840277777777778e-08,
+      "loss": 0.1276,
+      "step": 80675
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 2.5884193020093105,
+      "learning_rate": 9.666666666666669e-08,
+      "loss": 0.1228,
+      "step": 80700
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 3.065588647509654,
+      "learning_rate": 9.493055555555556e-08,
+      "loss": 0.1282,
+      "step": 80725
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 3.4441302429520957,
+      "learning_rate": 9.319444444444445e-08,
+      "loss": 0.1109,
+      "step": 80750
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 3.568478894528398,
+      "learning_rate": 9.145833333333335e-08,
+      "loss": 0.115,
+      "step": 80775
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 3.2298861894081052,
+      "learning_rate": 8.972222222222223e-08,
+      "loss": 0.1079,
+      "step": 80800
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 3.0351733682866784,
+      "learning_rate": 8.798611111111112e-08,
+      "loss": 0.1155,
+      "step": 80825
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 2.8423943298154777,
+      "learning_rate": 8.625000000000001e-08,
+      "loss": 0.1058,
+      "step": 80850
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 4.239375206161924,
+      "learning_rate": 8.45138888888889e-08,
+      "loss": 0.1202,
+      "step": 80875
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 3.9414210967413044,
+      "learning_rate": 8.277777777777779e-08,
+      "loss": 0.1226,
+      "step": 80900
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 2.6741087247987356,
+      "learning_rate": 8.104166666666668e-08,
+      "loss": 0.1131,
+      "step": 80925
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 2.470938835323759,
+      "learning_rate": 7.930555555555557e-08,
+      "loss": 0.1006,
+      "step": 80950
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 3.1023252839310533,
+      "learning_rate": 7.756944444444445e-08,
+      "loss": 0.1192,
+      "step": 80975
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 3.1292889463150133,
+      "learning_rate": 7.583333333333334e-08,
+      "loss": 0.1111,
+      "step": 81000
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 2.8191984119816373,
+      "learning_rate": 7.409722222222222e-08,
+      "loss": 0.1214,
+      "step": 81025
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 2.699005654746743,
+      "learning_rate": 7.236111111111111e-08,
+      "loss": 0.1083,
+      "step": 81050
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 2.5511639322171162,
+      "learning_rate": 7.062500000000001e-08,
+      "loss": 0.1028,
+      "step": 81075
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 4.017135101608909,
+      "learning_rate": 6.888888888888889e-08,
+      "loss": 0.108,
+      "step": 81100
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 3.354633095244795,
+      "learning_rate": 6.715277777777778e-08,
+      "loss": 0.1187,
+      "step": 81125
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 2.968280201052923,
+      "learning_rate": 6.541666666666668e-08,
+      "loss": 0.1117,
+      "step": 81150
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 3.441489150840829,
+      "learning_rate": 6.368055555555556e-08,
+      "loss": 0.1168,
+      "step": 81175
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 3.397176047085528,
+      "learning_rate": 6.194444444444445e-08,
+      "loss": 0.1167,
+      "step": 81200
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 3.7413484965620807,
+      "learning_rate": 6.020833333333335e-08,
+      "loss": 0.1245,
+      "step": 81225
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 3.368881934301906,
+      "learning_rate": 5.8541666666666666e-08,
+      "loss": 0.1164,
+      "step": 81250
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 3.5376665390407798,
+      "learning_rate": 5.680555555555556e-08,
+      "loss": 0.1125,
+      "step": 81275
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 3.071272749793487,
+      "learning_rate": 5.506944444444445e-08,
+      "loss": 0.1043,
+      "step": 81300
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 2.8957813289832415,
+      "learning_rate": 5.3333333333333334e-08,
+      "loss": 0.1227,
+      "step": 81325
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 3.177849245107898,
+      "learning_rate": 5.1597222222222224e-08,
+      "loss": 0.1151,
+      "step": 81350
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 3.036992433502032,
+      "learning_rate": 4.986111111111112e-08,
+      "loss": 0.1067,
+      "step": 81375
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 3.2947567806935134,
+      "learning_rate": 4.8125e-08,
+      "loss": 0.1169,
+      "step": 81400
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 2.6033716960489306,
+      "learning_rate": 4.638888888888889e-08,
+      "loss": 0.1308,
+      "step": 81425
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 3.271050187274925,
+      "learning_rate": 4.465277777777779e-08,
+      "loss": 0.1216,
+      "step": 81450
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 3.228010138283361,
+      "learning_rate": 4.291666666666667e-08,
+      "loss": 0.1178,
+      "step": 81475
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 1.4784343520529,
+      "learning_rate": 4.118055555555556e-08,
+      "loss": 0.1104,
+      "step": 81500
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 3.007137030167928,
+      "learning_rate": 3.944444444444445e-08,
+      "loss": 0.1063,
+      "step": 81525
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 3.391635916189553,
+      "learning_rate": 3.770833333333334e-08,
+      "loss": 0.1057,
+      "step": 81550
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 2.9441226132043647,
+      "learning_rate": 3.597222222222222e-08,
+      "loss": 0.1265,
+      "step": 81575
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 2.4559004956821036,
+      "learning_rate": 3.423611111111112e-08,
+      "loss": 0.107,
+      "step": 81600
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 3.1518532804348625,
+      "learning_rate": 3.25e-08,
+      "loss": 0.1375,
+      "step": 81625
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 2.8691924917711353,
+      "learning_rate": 3.076388888888889e-08,
+      "loss": 0.1073,
+      "step": 81650
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 2.866450101574571,
+      "learning_rate": 2.902777777777778e-08,
+      "loss": 0.1081,
+      "step": 81675
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 4.055904824711259,
+      "learning_rate": 2.7291666666666668e-08,
+      "loss": 0.1108,
+      "step": 81700
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 3.35548877818344,
+      "learning_rate": 2.555555555555556e-08,
+      "loss": 0.1175,
+      "step": 81725
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 3.0025022533511128,
+      "learning_rate": 2.3819444444444447e-08,
+      "loss": 0.1196,
+      "step": 81750
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 3.9601682959534625,
+      "learning_rate": 2.2083333333333336e-08,
+      "loss": 0.121,
+      "step": 81775
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 3.5576863922110955,
+      "learning_rate": 2.0347222222222222e-08,
+      "loss": 0.1156,
+      "step": 81800
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 3.0524980328514615,
+      "learning_rate": 1.861111111111111e-08,
+      "loss": 0.107,
+      "step": 81825
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 2.5465332535964564,
+      "learning_rate": 1.6875e-08,
+      "loss": 0.1104,
+      "step": 81850
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 3.9758255525233475,
+      "learning_rate": 1.513888888888889e-08,
+      "loss": 0.1192,
+      "step": 81875
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 3.156736393762649,
+      "learning_rate": 1.3402777777777778e-08,
+      "loss": 0.1246,
+      "step": 81900
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 2.193575947789099,
+      "learning_rate": 1.1666666666666669e-08,
+      "loss": 0.1183,
+      "step": 81925
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 4.202801227471746,
+      "learning_rate": 9.930555555555556e-09,
+      "loss": 0.1035,
+      "step": 81950
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 4.617004459055916,
+      "learning_rate": 8.194444444444446e-09,
+      "loss": 0.1142,
+      "step": 81975
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 3.2809986802623974,
+      "learning_rate": 6.458333333333334e-09,
+      "loss": 0.1231,
+      "step": 82000
+    },
+    {
+      "epoch": 3.22,
+      "step": 82000,
+      "total_flos": 7.219553262695875e+18,
+      "train_loss": 0.0028538422119326707,
+      "train_runtime": 23952.0519,
+      "train_samples_per_second": 109.552,
+      "train_steps_per_second": 3.424
+    }
+  ],
+  "logging_steps": 25,
+  "max_steps": 82000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 5000,
+  "total_flos": 7.219553262695875e+18,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}