diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,39132 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.999328768962277,
+  "eval_steps": 500,
+  "global_step": 27930,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.27031534910202026,
+      "learning_rate": 4.9999996046262024e-05,
+      "loss": 2.2625,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.3651030957698822,
+      "learning_rate": 4.999998418504935e-05,
+      "loss": 2.1664,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.4536094069480896,
+      "learning_rate": 4.999996441636572e-05,
+      "loss": 2.1371,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.3164426386356354,
+      "learning_rate": 4.999993674021739e-05,
+      "loss": 2.2912,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9441952109336853,
+      "learning_rate": 4.9999908905929944e-05,
+      "loss": 2.5717,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.45403921604156494,
+      "learning_rate": 4.99998669963689e-05,
+      "loss": 2.2767,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.38720938563346863,
+      "learning_rate": 4.9999817179373974e-05,
+      "loss": 2.2304,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2565215826034546,
+      "learning_rate": 4.9999759454960915e-05,
+      "loss": 1.9965,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.41144421696662903,
+      "learning_rate": 4.9999693823147996e-05,
+      "loss": 2.2625,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7766994833946228,
+      "learning_rate": 4.999962028395596e-05,
+      "loss": 2.253,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4188031852245331,
+      "learning_rate": 4.9999538837408077e-05,
+      "loss": 2.132,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6002323031425476,
+      "learning_rate": 4.999944948353011e-05,
+      "loss": 2.2579,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3897973597049713,
+      "learning_rate": 4.999935222235031e-05,
+      "loss": 2.0546,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2654597759246826,
+      "learning_rate": 4.9999247053899453e-05,
+      "loss": 2.1811,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.29960134625434875,
+      "learning_rate": 4.99991339782108e-05,
+      "loss": 1.9428,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.48218828439712524,
+      "learning_rate": 4.999901299532012e-05,
+      "loss": 2.2476,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6963359117507935,
+      "learning_rate": 4.9998884105265666e-05,
+      "loss": 1.9934,
+      "step": 85
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0964760780334473,
+      "learning_rate": 4.9998747308088226e-05,
+      "loss": 2.1915,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5373502969741821,
+      "learning_rate": 4.999860260383104e-05,
+      "loss": 1.988,
+      "step": 95
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4756152331829071,
+      "learning_rate": 4.999844999253991e-05,
+      "loss": 2.0278,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.35103705525398254,
+      "learning_rate": 4.9998289474263094e-05,
+      "loss": 2.1245,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6553156971931458,
+      "learning_rate": 4.999812104905136e-05,
+      "loss": 2.0964,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5029746890068054,
+      "learning_rate": 4.9997944716957985e-05,
+      "loss": 2.356,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3125253915786743,
+      "learning_rate": 4.999776047803873e-05,
+      "loss": 2.1678,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8542938232421875,
+      "learning_rate": 4.99975683323519e-05,
+      "loss": 2.1263,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3991641700267792,
+      "learning_rate": 4.9997368279958236e-05,
+      "loss": 2.0609,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4593062698841095,
+      "learning_rate": 4.999716032092103e-05,
+      "loss": 2.0401,
+      "step": 135
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3167307674884796,
+      "learning_rate": 4.999694445530607e-05,
+      "loss": 1.9401,
+      "step": 140
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7320476174354553,
+      "learning_rate": 4.9996720683181617e-05,
+      "loss": 2.0134,
+      "step": 145
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5831002593040466,
+      "learning_rate": 4.9996489004618455e-05,
+      "loss": 1.922,
+      "step": 150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0691373348236084,
+      "learning_rate": 4.999624941968986e-05,
+      "loss": 2.0451,
+      "step": 155
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5325329899787903,
+      "learning_rate": 4.999600192847162e-05,
+      "loss": 2.0023,
+      "step": 160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7082959413528442,
+      "learning_rate": 4.9995746531042006e-05,
+      "loss": 1.9057,
+      "step": 165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2313815355300903,
+      "learning_rate": 4.9995483227481824e-05,
+      "loss": 1.8793,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.24114160239696503,
+      "learning_rate": 4.9995212017874325e-05,
+      "loss": 2.1221,
+      "step": 175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4924792945384979,
+      "learning_rate": 4.9994932902305315e-05,
+      "loss": 2.0354,
+      "step": 180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.932770013809204,
+      "learning_rate": 4.9994645880863064e-05,
+      "loss": 2.0849,
+      "step": 185
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9561445713043213,
+      "learning_rate": 4.9994350953638366e-05,
+      "loss": 1.9205,
+      "step": 190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7701564431190491,
+      "learning_rate": 4.999404812072451e-05,
+      "loss": 2.2077,
+      "step": 195
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6854715943336487,
+      "learning_rate": 4.9993737382217265e-05,
+      "loss": 2.2886,
+      "step": 200
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7976120710372925,
+      "learning_rate": 4.999341873821493e-05,
+      "loss": 2.3515,
+      "step": 205
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1986960172653198,
+      "learning_rate": 4.999309218881829e-05,
+      "loss": 1.7781,
+      "step": 210
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5879755020141602,
+      "learning_rate": 4.999275773413063e-05,
+      "loss": 2.2107,
+      "step": 215
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0343559980392456,
+      "learning_rate": 4.9992415374257734e-05,
+      "loss": 1.8219,
+      "step": 220
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6155093312263489,
+      "learning_rate": 4.9992065109307904e-05,
+      "loss": 2.3041,
+      "step": 225
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1242767572402954,
+      "learning_rate": 4.9991706939391916e-05,
+      "loss": 2.2295,
+      "step": 230
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5417547225952148,
+      "learning_rate": 4.9991340864623057e-05,
+      "loss": 1.7807,
+      "step": 235
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.36535757780075073,
+      "learning_rate": 4.999096688511712e-05,
+      "loss": 1.8687,
+      "step": 240
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8204144835472107,
+      "learning_rate": 4.99905850009924e-05,
+      "loss": 2.1915,
+      "step": 245
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5473398566246033,
+      "learning_rate": 4.999019521236969e-05,
+      "loss": 1.9105,
+      "step": 250
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5916422605514526,
+      "learning_rate": 4.998979751937226e-05,
+      "loss": 1.8488,
+      "step": 255
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.40376555919647217,
+      "learning_rate": 4.998939192212591e-05,
+      "loss": 1.9348,
+      "step": 260
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4245920479297638,
+      "learning_rate": 4.998897842075894e-05,
+      "loss": 1.8736,
+      "step": 265
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.863102436065674,
+      "learning_rate": 4.998855701540213e-05,
+      "loss": 1.8208,
+      "step": 270
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9868331551551819,
+      "learning_rate": 4.998812770618877e-05,
+      "loss": 2.0812,
+      "step": 275
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3408450484275818,
+      "learning_rate": 4.998769049325465e-05,
+      "loss": 2.2077,
+      "step": 280
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3527214229106903,
+      "learning_rate": 4.998724537673807e-05,
+      "loss": 2.1004,
+      "step": 285
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1077073812484741,
+      "learning_rate": 4.9986792356779803e-05,
+      "loss": 2.3449,
+      "step": 290
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5574270486831665,
+      "learning_rate": 4.9986331433523156e-05,
+      "loss": 1.8885,
+      "step": 295
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4326357841491699,
+      "learning_rate": 4.99858626071139e-05,
+      "loss": 1.9964,
+      "step": 300
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9008758068084717,
+      "learning_rate": 4.998538587770034e-05,
+      "loss": 1.9349,
+      "step": 305
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8824496865272522,
+      "learning_rate": 4.998490124543327e-05,
+      "loss": 1.9985,
+      "step": 310
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6680270433425903,
+      "learning_rate": 4.998440871046596e-05,
+      "loss": 1.8583,
+      "step": 315
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.392929345369339,
+      "learning_rate": 4.99839082729542e-05,
+      "loss": 1.926,
+      "step": 320
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6838160753250122,
+      "learning_rate": 4.998339993305629e-05,
+      "loss": 2.0671,
+      "step": 325
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6205262541770935,
+      "learning_rate": 4.9982883690933014e-05,
+      "loss": 1.8439,
+      "step": 330
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5349203944206238,
+      "learning_rate": 4.998235954674766e-05,
+      "loss": 2.1677,
+      "step": 335
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5297353267669678,
+      "learning_rate": 4.9981827500666e-05,
+      "loss": 2.1915,
+      "step": 340
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5714792013168335,
+      "learning_rate": 4.9981287552856344e-05,
+      "loss": 1.8878,
+      "step": 345
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.546989917755127,
+      "learning_rate": 4.998073970348945e-05,
+      "loss": 2.1013,
+      "step": 350
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7602975964546204,
+      "learning_rate": 4.9980183952738634e-05,
+      "loss": 2.2048,
+      "step": 355
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0656648874282837,
+      "learning_rate": 4.997962030077965e-05,
+      "loss": 2.1493,
+      "step": 360
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6455523371696472,
+      "learning_rate": 4.997904874779079e-05,
+      "loss": 1.6774,
+      "step": 365
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.34377214312553406,
+      "learning_rate": 4.997846929395285e-05,
+      "loss": 1.8235,
+      "step": 370
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6402696967124939,
+      "learning_rate": 4.997788193944909e-05,
+      "loss": 1.9533,
+      "step": 375
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5523343086242676,
+      "learning_rate": 4.99772866844653e-05,
+      "loss": 2.0183,
+      "step": 380
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6676031947135925,
+      "learning_rate": 4.997668352918975e-05,
+      "loss": 1.8157,
+      "step": 385
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9743964672088623,
+      "learning_rate": 4.997607247381323e-05,
+      "loss": 2.1711,
+      "step": 390
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7307085990905762,
+      "learning_rate": 4.9975453518529014e-05,
+      "loss": 1.8769,
+      "step": 395
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.004335641860962,
+      "learning_rate": 4.997482666353287e-05,
+      "loss": 1.6377,
+      "step": 400
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7974880337715149,
+      "learning_rate": 4.997419190902307e-05,
+      "loss": 1.8962,
+      "step": 405
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4197213649749756,
+      "learning_rate": 4.99735492552004e-05,
+      "loss": 2.0266,
+      "step": 410
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5465844869613647,
+      "learning_rate": 4.997289870226812e-05,
+      "loss": 1.9882,
+      "step": 415
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7714999914169312,
+      "learning_rate": 4.9972240250432e-05,
+      "loss": 1.7723,
+      "step": 420
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4609231650829315,
+      "learning_rate": 4.9971573899900306e-05,
+      "loss": 2.1264,
+      "step": 425
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4901442527770996,
+      "learning_rate": 4.997089965088381e-05,
+      "loss": 1.8215,
+      "step": 430
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3669675588607788,
+      "learning_rate": 4.997021750359577e-05,
+      "loss": 1.9637,
+      "step": 435
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6169967651367188,
+      "learning_rate": 4.9969527458251964e-05,
+      "loss": 2.2501,
+      "step": 440
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.862839937210083,
+      "learning_rate": 4.9968829515070626e-05,
+      "loss": 2.0921,
+      "step": 445
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5367420315742493,
+      "learning_rate": 4.996812367427254e-05,
+      "loss": 1.9298,
+      "step": 450
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4946227967739105,
+      "learning_rate": 4.996740993608094e-05,
+      "loss": 2.0265,
+      "step": 455
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.68914395570755,
+      "learning_rate": 4.9966688300721603e-05,
+      "loss": 1.9182,
+      "step": 460
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.44019490480422974,
+      "learning_rate": 4.9965958768422775e-05,
+      "loss": 2.1147,
+      "step": 465
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7905745506286621,
+      "learning_rate": 4.9965221339415194e-05,
+      "loss": 1.909,
+      "step": 470
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5773944854736328,
+      "learning_rate": 4.996447601393211e-05,
+      "loss": 1.8616,
+      "step": 475
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7853548526763916,
+      "learning_rate": 4.9963722792209286e-05,
+      "loss": 2.2229,
+      "step": 480
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6674456596374512,
+      "learning_rate": 4.996296167448495e-05,
+      "loss": 1.9064,
+      "step": 485
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.360712766647339,
+      "learning_rate": 4.9962192660999866e-05,
+      "loss": 1.932,
+      "step": 490
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3946886360645294,
+      "learning_rate": 4.996141575199723e-05,
+      "loss": 2.112,
+      "step": 495
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.100876808166504,
+      "learning_rate": 4.996063094772281e-05,
+      "loss": 2.0544,
+      "step": 500
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3931337594985962,
+      "learning_rate": 4.9959838248424836e-05,
+      "loss": 1.7863,
+      "step": 505
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4513574540615082,
+      "learning_rate": 4.995903765435402e-05,
+      "loss": 2.1613,
+      "step": 510
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6865633130073547,
+      "learning_rate": 4.9958229165763614e-05,
+      "loss": 2.0138,
+      "step": 515
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6642078757286072,
+      "learning_rate": 4.9957412782909324e-05,
+      "loss": 1.8269,
+      "step": 520
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0124841928482056,
+      "learning_rate": 4.9956588506049374e-05,
+      "loss": 2.0759,
+      "step": 525
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8257017731666565,
+      "learning_rate": 4.9955756335444495e-05,
+      "loss": 1.7972,
+      "step": 530
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7027864456176758,
+      "learning_rate": 4.9954916271357876e-05,
+      "loss": 2.1897,
+      "step": 535
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6689014434814453,
+      "learning_rate": 4.9954068314055255e-05,
+      "loss": 2.0164,
+      "step": 540
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6984819769859314,
+      "learning_rate": 4.9953212463804824e-05,
+      "loss": 1.8525,
+      "step": 545
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8252110481262207,
+      "learning_rate": 4.9952348720877294e-05,
+      "loss": 1.906,
+      "step": 550
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7641300559043884,
+      "learning_rate": 4.995147708554587e-05,
+      "loss": 2.179,
+      "step": 555
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5277909636497498,
+      "learning_rate": 4.9950597558086246e-05,
+      "loss": 2.078,
+      "step": 560
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3785237073898315,
+      "learning_rate": 4.99497101387766e-05,
+      "loss": 2.1335,
+      "step": 565
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9279726147651672,
+      "learning_rate": 4.994881482789764e-05,
+      "loss": 1.823,
+      "step": 570
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6728343963623047,
+      "learning_rate": 4.994791162573256e-05,
+      "loss": 2.1131,
+      "step": 575
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7369157075881958,
+      "learning_rate": 4.9947000532567014e-05,
+      "loss": 1.7872,
+      "step": 580
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.544488549232483,
+      "learning_rate": 4.994608154868921e-05,
+      "loss": 2.2275,
+      "step": 585
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0794744491577148,
+      "learning_rate": 4.9945154674389796e-05,
+      "loss": 2.0575,
+      "step": 590
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5891600847244263,
+      "learning_rate": 4.994421990996195e-05,
+      "loss": 2.0061,
+      "step": 595
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5102590322494507,
+      "learning_rate": 4.994327725570135e-05,
+      "loss": 1.8905,
+      "step": 600
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9400714635848999,
+      "learning_rate": 4.994232671190614e-05,
+      "loss": 1.9824,
+      "step": 605
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7190917730331421,
+      "learning_rate": 4.994136827887699e-05,
+      "loss": 1.9437,
+      "step": 610
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5271459221839905,
+      "learning_rate": 4.994040195691703e-05,
+      "loss": 1.9421,
+      "step": 615
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.935598611831665,
+      "learning_rate": 4.9939427746331937e-05,
+      "loss": 2.1908,
+      "step": 620
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 14.462722778320312,
+      "learning_rate": 4.993844564742982e-05,
+      "loss": 2.0414,
+      "step": 625
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.596210479736328,
+      "learning_rate": 4.9937455660521345e-05,
+      "loss": 1.966,
+      "step": 630
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.547702431678772,
+      "learning_rate": 4.993645778591963e-05,
+      "loss": 1.8967,
+      "step": 635
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.006232976913452,
+      "learning_rate": 4.99354520239403e-05,
+      "loss": 1.8511,
+      "step": 640
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9103983640670776,
+      "learning_rate": 4.993443837490148e-05,
+      "loss": 1.7829,
+      "step": 645
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8740873336791992,
+      "learning_rate": 4.993341683912378e-05,
+      "loss": 2.0906,
+      "step": 650
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9075211882591248,
+      "learning_rate": 4.9932387416930326e-05,
+      "loss": 2.1868,
+      "step": 655
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5851132273674011,
+      "learning_rate": 4.993135010864671e-05,
+      "loss": 1.8202,
+      "step": 660
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7303404808044434,
+      "learning_rate": 4.993030491460104e-05,
+      "loss": 1.8032,
+      "step": 665
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5029420256614685,
+      "learning_rate": 4.9929251835123904e-05,
+      "loss": 1.8013,
+      "step": 670
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8666778802871704,
+      "learning_rate": 4.9928190870548384e-05,
+      "loss": 2.1201,
+      "step": 675
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3739973306655884,
+      "learning_rate": 4.9927122021210076e-05,
+      "loss": 1.8149,
+      "step": 680
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6640814542770386,
+      "learning_rate": 4.992604528744705e-05,
+      "loss": 1.9143,
+      "step": 685
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9875686764717102,
+      "learning_rate": 4.992496066959987e-05,
+      "loss": 1.7527,
+      "step": 690
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5057778358459473,
+      "learning_rate": 4.992386816801161e-05,
+      "loss": 1.9771,
+      "step": 695
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7259483337402344,
+      "learning_rate": 4.992276778302782e-05,
+      "loss": 1.891,
+      "step": 700
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7736199498176575,
+      "learning_rate": 4.9921659514996555e-05,
+      "loss": 1.8235,
+      "step": 705
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6016276478767395,
+      "learning_rate": 4.992054336426836e-05,
+      "loss": 1.868,
+      "step": 710
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6753050088882446,
+      "learning_rate": 4.991941933119626e-05,
+      "loss": 1.8746,
+      "step": 715
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6114783883094788,
+      "learning_rate": 4.99182874161358e-05,
+      "loss": 1.9586,
+      "step": 720
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7114779949188232,
+      "learning_rate": 4.9917147619445005e-05,
+      "loss": 2.3431,
+      "step": 725
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6027891635894775,
+      "learning_rate": 4.991599994148438e-05,
+      "loss": 2.1642,
+      "step": 730
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7011446952819824,
+      "learning_rate": 4.991484438261693e-05,
+      "loss": 1.9181,
+      "step": 735
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5473892092704773,
+      "learning_rate": 4.991368094320818e-05,
+      "loss": 2.1816,
+      "step": 740
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.4778834879398346,
+      "learning_rate": 4.991250962362611e-05,
+      "loss": 1.7557,
+      "step": 745
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6992874145507812,
+      "learning_rate": 4.99113304242412e-05,
+      "loss": 2.0037,
+      "step": 750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2139089107513428,
+      "learning_rate": 4.9910143345426446e-05,
+      "loss": 1.9781,
+      "step": 755
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7376035451889038,
+      "learning_rate": 4.9908948387557306e-05,
+      "loss": 1.9145,
+      "step": 760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9813640117645264,
+      "learning_rate": 4.9907745551011764e-05,
+      "loss": 1.9012,
+      "step": 765
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6064789295196533,
+      "learning_rate": 4.990653483617025e-05,
+      "loss": 2.0977,
+      "step": 770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8322803378105164,
+      "learning_rate": 4.990531624341573e-05,
+      "loss": 2.0924,
+      "step": 775
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39520594477653503,
+      "learning_rate": 4.990408977313363e-05,
+      "loss": 2.0109,
+      "step": 780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7123391628265381,
+      "learning_rate": 4.9902855425711905e-05,
+      "loss": 2.1073,
+      "step": 785
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.617445170879364,
+      "learning_rate": 4.9901613201540945e-05,
+      "loss": 2.081,
+      "step": 790
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2326832264661789,
+      "learning_rate": 4.990036310101369e-05,
+      "loss": 2.0825,
+      "step": 795
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.828574001789093,
+      "learning_rate": 4.989910512452554e-05,
+      "loss": 1.9948,
+      "step": 800
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5718833804130554,
+      "learning_rate": 4.989783927247439e-05,
+      "loss": 1.9836,
+      "step": 805
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5843701958656311,
+      "learning_rate": 4.989656554526062e-05,
+      "loss": 1.8482,
+      "step": 810
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.36369892954826355,
+      "learning_rate": 4.989528394328712e-05,
+      "loss": 1.9934,
+      "step": 815
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5329487323760986,
+      "learning_rate": 4.9893994466959246e-05,
+      "loss": 1.7943,
+      "step": 820
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9195080399513245,
+      "learning_rate": 4.989269711668487e-05,
+      "loss": 1.8812,
+      "step": 825
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.639167845249176,
+      "learning_rate": 4.9891391892874354e-05,
+      "loss": 2.0915,
+      "step": 830
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9392881393432617,
+      "learning_rate": 4.989007879594051e-05,
+      "loss": 1.9084,
+      "step": 835
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4047281742095947,
+      "learning_rate": 4.988875782629869e-05,
+      "loss": 2.1699,
+      "step": 840
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1853729486465454,
+      "learning_rate": 4.9887428984366705e-05,
+      "loss": 2.3892,
+      "step": 845
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.790334701538086,
+      "learning_rate": 4.9886092270564876e-05,
+      "loss": 1.9039,
+      "step": 850
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5937879681587219,
+      "learning_rate": 4.9884747685315996e-05,
+      "loss": 2.0544,
+      "step": 855
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8132315278053284,
+      "learning_rate": 4.988339522904536e-05,
+      "loss": 1.9541,
+      "step": 860
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7449702620506287,
+      "learning_rate": 4.988203490218075e-05,
+      "loss": 1.9042,
+      "step": 865
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5091820359230042,
+      "learning_rate": 4.9880666705152434e-05,
+      "loss": 1.9421,
+      "step": 870
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.45351913571357727,
+      "learning_rate": 4.987929063839317e-05,
+      "loss": 1.8925,
+      "step": 875
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5173969268798828,
+      "learning_rate": 4.987790670233821e-05,
+      "loss": 1.9209,
+      "step": 880
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5571516156196594,
+      "learning_rate": 4.987651489742528e-05,
+      "loss": 1.9871,
+      "step": 885
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9875330328941345,
+      "learning_rate": 4.987511522409463e-05,
+      "loss": 1.7102,
+      "step": 890
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.513702392578125,
+      "learning_rate": 4.987370768278895e-05,
+      "loss": 2.0374,
+      "step": 895
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7539028525352478,
+      "learning_rate": 4.987229227395346e-05,
+      "loss": 1.9778,
+      "step": 900
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5421642661094666,
+      "learning_rate": 4.9870868998035845e-05,
+      "loss": 1.7694,
+      "step": 905
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8099106550216675,
+      "learning_rate": 4.986943785548629e-05,
+      "loss": 1.7683,
+      "step": 910
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.797715425491333,
+      "learning_rate": 4.986799884675747e-05,
+      "loss": 1.8887,
+      "step": 915
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6067102551460266,
+      "learning_rate": 4.9866551972304526e-05,
+      "loss": 1.9329,
+      "step": 920
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.49789664149284363,
+      "learning_rate": 4.986509723258511e-05,
+      "loss": 2.0683,
+      "step": 925
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9681997299194336,
+      "learning_rate": 4.9863634628059366e-05,
+      "loss": 1.9902,
+      "step": 930
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5609699487686157,
+      "learning_rate": 4.98621641591899e-05,
+      "loss": 1.8425,
+      "step": 935
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5346946120262146,
+      "learning_rate": 4.986068582644182e-05,
+      "loss": 2.0863,
+      "step": 940
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.893808126449585,
+      "learning_rate": 4.985919963028274e-05,
+      "loss": 2.0174,
+      "step": 945
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4146747887134552,
+      "learning_rate": 4.985770557118272e-05,
+      "loss": 1.9704,
+      "step": 950
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.030465006828308,
+      "learning_rate": 4.985620364961433e-05,
+      "loss": 2.1478,
+      "step": 955
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6539037823677063,
+      "learning_rate": 4.985469386605265e-05,
+      "loss": 2.1638,
+      "step": 960
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.698945164680481,
+      "learning_rate": 4.9853176220975195e-05,
+      "loss": 1.93,
+      "step": 965
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6615301370620728,
+      "learning_rate": 4.9851650714862006e-05,
+      "loss": 2.0131,
+      "step": 970
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6794114112854004,
+      "learning_rate": 4.9850117348195606e-05,
+      "loss": 1.9938,
+      "step": 975
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4842730164527893,
+      "learning_rate": 4.984857612146099e-05,
+      "loss": 1.9502,
+      "step": 980
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0093274116516113,
+      "learning_rate": 4.984702703514565e-05,
+      "loss": 1.6196,
+      "step": 985
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6907001733779907,
+      "learning_rate": 4.984547008973956e-05,
+      "loss": 1.951,
+      "step": 990
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6231794357299805,
+      "learning_rate": 4.984390528573517e-05,
+      "loss": 2.0397,
+      "step": 995
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7842473387718201,
+      "learning_rate": 4.984233262362745e-05,
+      "loss": 2.0751,
+      "step": 1000
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2024447917938232,
+      "learning_rate": 4.98407521039138e-05,
+      "loss": 1.5978,
+      "step": 1005
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8643293976783752,
+      "learning_rate": 4.983916372709416e-05,
+      "loss": 2.2393,
+      "step": 1010
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6391337513923645,
+      "learning_rate": 4.9837567493670936e-05,
+      "loss": 1.9106,
+      "step": 1015
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7475038766860962,
+      "learning_rate": 4.983596340414899e-05,
+      "loss": 2.0462,
+      "step": 1020
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4926930665969849,
+      "learning_rate": 4.9834351459035714e-05,
+      "loss": 1.7617,
+      "step": 1025
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6800340414047241,
+      "learning_rate": 4.9832731658840956e-05,
+      "loss": 1.7933,
+      "step": 1030
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5612903237342834,
+      "learning_rate": 4.983110400407707e-05,
+      "loss": 1.9801,
+      "step": 1035
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5487276315689087,
+      "learning_rate": 4.982946849525886e-05,
+      "loss": 1.8137,
+      "step": 1040
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.1689997911453247,
+      "learning_rate": 4.982782513290365e-05,
+      "loss": 2.0348,
+      "step": 1045
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5050978660583496,
+      "learning_rate": 4.9826173917531235e-05,
+      "loss": 2.1974,
+      "step": 1050
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6541104912757874,
+      "learning_rate": 4.982451484966389e-05,
+      "loss": 1.824,
+      "step": 1055
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4202798306941986,
+      "learning_rate": 4.9822847929826374e-05,
+      "loss": 1.9036,
+      "step": 1060
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6690078377723694,
+      "learning_rate": 4.9821173158545936e-05,
+      "loss": 1.8924,
+      "step": 1065
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.41221898794174194,
+      "learning_rate": 4.98194905363523e-05,
+      "loss": 2.2559,
+      "step": 1070
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.146337628364563,
+      "learning_rate": 4.981780006377769e-05,
+      "loss": 1.7768,
+      "step": 1075
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.376248836517334,
+      "learning_rate": 4.981610174135678e-05,
+      "loss": 1.8807,
+      "step": 1080
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8797820806503296,
+      "learning_rate": 4.981439556962676e-05,
+      "loss": 1.5606,
+      "step": 1085
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9308553338050842,
+      "learning_rate": 4.981268154912729e-05,
+      "loss": 1.7338,
+      "step": 1090
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8552083969116211,
+      "learning_rate": 4.981095968040053e-05,
+      "loss": 1.7534,
+      "step": 1095
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9000282287597656,
+      "learning_rate": 4.9809229963991064e-05,
+      "loss": 1.7684,
+      "step": 1100
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6335359811782837,
+      "learning_rate": 4.980749240044604e-05,
+      "loss": 1.9627,
+      "step": 1105
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8010488152503967,
+      "learning_rate": 4.9805746990315024e-05,
+      "loss": 2.0931,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5501634478569031,
+      "learning_rate": 4.980399373415009e-05,
+      "loss": 2.1531,
+      "step": 1115
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.876747190952301,
+      "learning_rate": 4.980223263250581e-05,
+      "loss": 1.755,
+      "step": 1120
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8235836625099182,
+      "learning_rate": 4.98004636859392e-05,
+      "loss": 2.2103,
+      "step": 1125
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3459777534008026,
+      "learning_rate": 4.979868689500978e-05,
+      "loss": 1.8597,
+      "step": 1130
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6202408075332642,
+      "learning_rate": 4.9796902260279546e-05,
+      "loss": 1.8657,
+      "step": 1135
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.912277340888977,
+      "learning_rate": 4.9795109782312974e-05,
+      "loss": 1.7297,
+      "step": 1140
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5778829455375671,
+      "learning_rate": 4.979330946167704e-05,
+      "loss": 1.939,
+      "step": 1145
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8134537935256958,
+      "learning_rate": 4.9791501298941165e-05,
+      "loss": 1.9645,
+      "step": 1150
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8636032938957214,
+      "learning_rate": 4.978968529467728e-05,
+      "loss": 1.6756,
+      "step": 1155
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.87062668800354,
+      "learning_rate": 4.978786144945977e-05,
+      "loss": 1.7352,
+      "step": 1160
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9337445497512817,
+      "learning_rate": 4.978602976386554e-05,
+      "loss": 1.6599,
+      "step": 1165
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8104198575019836,
+      "learning_rate": 4.978419023847393e-05,
+      "loss": 1.7557,
+      "step": 1170
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3826814591884613,
+      "learning_rate": 4.978234287386678e-05,
+      "loss": 2.0881,
+      "step": 1175
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.47126176953315735,
+      "learning_rate": 4.978048767062843e-05,
+      "loss": 2.2003,
+      "step": 1180
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4971998333930969,
+      "learning_rate": 4.977862462934566e-05,
+      "loss": 1.8297,
+      "step": 1185
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5127992033958435,
+      "learning_rate": 4.977675375060775e-05,
+      "loss": 1.8892,
+      "step": 1190
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7723473310470581,
+      "learning_rate": 4.9774875035006464e-05,
+      "loss": 1.9711,
+      "step": 1195
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5480608344078064,
+      "learning_rate": 4.977298848313604e-05,
+      "loss": 1.7975,
+      "step": 1200
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4244174659252167,
+      "learning_rate": 4.9771094095593176e-05,
+      "loss": 2.1068,
+      "step": 1205
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5436844825744629,
+      "learning_rate": 4.9769191872977085e-05,
+      "loss": 2.09,
+      "step": 1210
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4076632261276245,
+      "learning_rate": 4.9767281815889425e-05,
+      "loss": 2.038,
+      "step": 1215
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5819802284240723,
+      "learning_rate": 4.976536392493435e-05,
+      "loss": 1.8444,
+      "step": 1220
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7633818984031677,
+      "learning_rate": 4.976343820071849e-05,
+      "loss": 2.0922,
+      "step": 1225
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6349738836288452,
+      "learning_rate": 4.9761504643850945e-05,
+      "loss": 1.9147,
+      "step": 1230
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5177556276321411,
+      "learning_rate": 4.9759563254943306e-05,
+      "loss": 1.7667,
+      "step": 1235
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4259908199310303,
+      "learning_rate": 4.9757614034609615e-05,
+      "loss": 1.9067,
+      "step": 1240
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8151448369026184,
+      "learning_rate": 4.975565698346642e-05,
+      "loss": 2.0555,
+      "step": 1245
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.9753692102132735e-05,
+      "loss": 1.9399,
+      "step": 1250
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.86171954870224,
+      "learning_rate": 4.975171939123005e-05,
+      "loss": 2.203,
+      "step": 1255
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.751837432384491,
+      "learning_rate": 4.9749738851382326e-05,
+      "loss": 1.7247,
+      "step": 1260
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.416966438293457,
+      "learning_rate": 4.9747750483216015e-05,
+      "loss": 2.1371,
+      "step": 1265
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8290925621986389,
+      "learning_rate": 4.974575428736002e-05,
+      "loss": 1.9527,
+      "step": 1270
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5154220461845398,
+      "learning_rate": 4.974375026444575e-05,
+      "loss": 2.2402,
+      "step": 1275
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5878034830093384,
+      "learning_rate": 4.974173841510708e-05,
+      "loss": 2.008,
+      "step": 1280
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9623072743415833,
+      "learning_rate": 4.9739718739980346e-05,
+      "loss": 1.734,
+      "step": 1285
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6699478030204773,
+      "learning_rate": 4.9737691239704366e-05,
+      "loss": 1.9048,
+      "step": 1290
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6063868403434753,
+      "learning_rate": 4.9735655914920445e-05,
+      "loss": 2.3764,
+      "step": 1295
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.598061740398407,
+      "learning_rate": 4.973361276627235e-05,
+      "loss": 1.901,
+      "step": 1300
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7591137886047363,
+      "learning_rate": 4.9731561794406326e-05,
+      "loss": 1.8706,
+      "step": 1305
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5705850124359131,
+      "learning_rate": 4.9729502999971086e-05,
+      "loss": 2.0533,
+      "step": 1310
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9160748720169067,
+      "learning_rate": 4.9727436383617853e-05,
+      "loss": 1.8877,
+      "step": 1315
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8184042572975159,
+      "learning_rate": 4.9725361946000267e-05,
+      "loss": 1.8283,
+      "step": 1320
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9235120415687561,
+      "learning_rate": 4.9723279687774485e-05,
+      "loss": 2.1376,
+      "step": 1325
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4321657121181488,
+      "learning_rate": 4.9721189609599106e-05,
+      "loss": 1.9237,
+      "step": 1330
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.47391876578330994,
+      "learning_rate": 4.9719091712135244e-05,
+      "loss": 1.9186,
+      "step": 1335
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.293057441711426,
+      "learning_rate": 4.9716985996046443e-05,
+      "loss": 1.8054,
+      "step": 1340
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6522160768508911,
+      "learning_rate": 4.971487246199875e-05,
+      "loss": 1.9026,
+      "step": 1345
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6272959113121033,
+      "learning_rate": 4.971275111066067e-05,
+      "loss": 2.1646,
+      "step": 1350
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.32946717739105225,
+      "learning_rate": 4.9710621942703186e-05,
+      "loss": 1.8629,
+      "step": 1355
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5091928839683533,
+      "learning_rate": 4.970848495879975e-05,
+      "loss": 2.0348,
+      "step": 1360
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.6853437423706055,
+      "learning_rate": 4.9706340159626284e-05,
+      "loss": 1.6108,
+      "step": 1365
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7039760947227478,
+      "learning_rate": 4.970418754586119e-05,
+      "loss": 2.1027,
+      "step": 1370
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4680666923522949,
+      "learning_rate": 4.970202711818535e-05,
+      "loss": 2.0027,
+      "step": 1375
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.49407467246055603,
+      "learning_rate": 4.969985887728208e-05,
+      "loss": 2.144,
+      "step": 1380
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6088310480117798,
+      "learning_rate": 4.96976828238372e-05,
+      "loss": 2.2039,
+      "step": 1385
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.39288827776908875,
+      "learning_rate": 4.9695498958539014e-05,
+      "loss": 1.9803,
+      "step": 1390
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5070251226425171,
+      "learning_rate": 4.969330728207825e-05,
+      "loss": 2.078,
+      "step": 1395
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4578852653503418,
+      "learning_rate": 4.969110779514815e-05,
+      "loss": 1.7425,
+      "step": 1400
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8231094479560852,
+      "learning_rate": 4.96889004984444e-05,
+      "loss": 2.1139,
+      "step": 1405
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5976428985595703,
+      "learning_rate": 4.9686685392665174e-05,
+      "loss": 2.0963,
+      "step": 1410
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8335079550743103,
+      "learning_rate": 4.968446247851111e-05,
+      "loss": 1.9815,
+      "step": 1415
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8027464747428894,
+      "learning_rate": 4.96822317566853e-05,
+      "loss": 1.878,
+      "step": 1420
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3042505979537964,
+      "learning_rate": 4.9679993227893326e-05,
+      "loss": 2.0152,
+      "step": 1425
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6054438352584839,
+      "learning_rate": 4.967774689284323e-05,
+      "loss": 2.0464,
+      "step": 1430
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.845056414604187,
+      "learning_rate": 4.9675492752245536e-05,
+      "loss": 2.0495,
+      "step": 1435
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.930960476398468,
+      "learning_rate": 4.967323080681322e-05,
+      "loss": 2.0878,
+      "step": 1440
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.2143394947052,
+      "learning_rate": 4.967096105726173e-05,
+      "loss": 2.1571,
+      "step": 1445
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4864540696144104,
+      "learning_rate": 4.966868350430899e-05,
+      "loss": 1.9011,
+      "step": 1450
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6188135147094727,
+      "learning_rate": 4.9666398148675374e-05,
+      "loss": 2.0124,
+      "step": 1455
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.174637794494629,
+      "learning_rate": 4.9664104991083757e-05,
+      "loss": 1.8437,
+      "step": 1460
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6095223426818848,
+      "learning_rate": 4.9661804032259464e-05,
+      "loss": 1.8675,
+      "step": 1465
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4776439964771271,
+      "learning_rate": 4.9659495272930266e-05,
+      "loss": 2.1693,
+      "step": 1470
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6089404225349426,
+      "learning_rate": 4.9657178713826434e-05,
+      "loss": 2.2899,
+      "step": 1475
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6431376934051514,
+      "learning_rate": 4.96548543556807e-05,
+      "loss": 1.9234,
+      "step": 1480
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5474340319633484,
+      "learning_rate": 4.965252219922825e-05,
+      "loss": 2.0113,
+      "step": 1485
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.44193366169929504,
+      "learning_rate": 4.9650182245206726e-05,
+      "loss": 1.8976,
+      "step": 1490
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8061282634735107,
+      "learning_rate": 4.9647834494356274e-05,
+      "loss": 2.0584,
+      "step": 1495
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5918030738830566,
+      "learning_rate": 4.964547894741949e-05,
+      "loss": 2.2051,
+      "step": 1500
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7036905288696289,
+      "learning_rate": 4.964311560514141e-05,
+      "loss": 2.1186,
+      "step": 1505
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5520045757293701,
+      "learning_rate": 4.964074446826957e-05,
+      "loss": 2.0825,
+      "step": 1510
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.34511449933052063,
+      "learning_rate": 4.9638365537553956e-05,
+      "loss": 1.9604,
+      "step": 1515
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0205286741256714,
+      "learning_rate": 4.963597881374702e-05,
+      "loss": 1.5741,
+      "step": 1520
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0340920686721802,
+      "learning_rate": 4.9633584297603686e-05,
+      "loss": 2.076,
+      "step": 1525
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.757640540599823,
+      "learning_rate": 4.963118198988133e-05,
+      "loss": 2.2935,
+      "step": 1530
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.603891134262085,
+      "learning_rate": 4.9628771891339806e-05,
+      "loss": 2.0039,
+      "step": 1535
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5811793804168701,
+      "learning_rate": 4.962635400274142e-05,
+      "loss": 1.8798,
+      "step": 1540
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6387256383895874,
+      "learning_rate": 4.962392832485095e-05,
+      "loss": 1.814,
+      "step": 1545
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7429436445236206,
+      "learning_rate": 4.962149485843564e-05,
+      "loss": 1.8334,
+      "step": 1550
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6524866819381714,
+      "learning_rate": 4.9619053604265185e-05,
+      "loss": 1.9859,
+      "step": 1555
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7228694558143616,
+      "learning_rate": 4.961660456311176e-05,
+      "loss": 1.8701,
+      "step": 1560
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.49972599744796753,
+      "learning_rate": 4.961414773574998e-05,
+      "loss": 2.0535,
+      "step": 1565
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3244457542896271,
+      "learning_rate": 4.961168312295696e-05,
+      "loss": 1.7632,
+      "step": 1570
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6015215516090393,
+      "learning_rate": 4.9609210725512234e-05,
+      "loss": 1.8848,
+      "step": 1575
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5646616220474243,
+      "learning_rate": 4.9606730544197834e-05,
+      "loss": 1.8899,
+      "step": 1580
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.66986483335495,
+      "learning_rate": 4.960424257979822e-05,
+      "loss": 2.0326,
+      "step": 1585
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.0103802680969238,
+      "learning_rate": 4.960174683310035e-05,
+      "loss": 1.9453,
+      "step": 1590
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6305530667304993,
+      "learning_rate": 4.9599243304893625e-05,
+      "loss": 1.6763,
+      "step": 1595
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6252381801605225,
+      "learning_rate": 4.95967319959699e-05,
+      "loss": 1.7975,
+      "step": 1600
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2989176511764526,
+      "learning_rate": 4.95942129071235e-05,
+      "loss": 2.024,
+      "step": 1605
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5174963474273682,
+      "learning_rate": 4.959168603915122e-05,
+      "loss": 1.7532,
+      "step": 1610
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.730141282081604,
+      "learning_rate": 4.95891513928523e-05,
+      "loss": 1.7375,
+      "step": 1615
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.690668523311615,
+      "learning_rate": 4.958660896902844e-05,
+      "loss": 2.0428,
+      "step": 1620
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5279943346977234,
+      "learning_rate": 4.958405876848382e-05,
+      "loss": 1.8309,
+      "step": 1625
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4510974586009979,
+      "learning_rate": 4.958150079202505e-05,
+      "loss": 1.7792,
+      "step": 1630
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.628743052482605,
+      "learning_rate": 4.957893504046123e-05,
+      "loss": 2.0351,
+      "step": 1635
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7251464128494263,
+      "learning_rate": 4.95763615146039e-05,
+      "loss": 1.9267,
+      "step": 1640
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6540181040763855,
+      "learning_rate": 4.957378021526705e-05,
+      "loss": 1.8565,
+      "step": 1645
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.511382520198822,
+      "learning_rate": 4.957119114326717e-05,
+      "loss": 2.0029,
+      "step": 1650
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8582594990730286,
+      "learning_rate": 4.9568594299423154e-05,
+      "loss": 1.9714,
+      "step": 1655
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.29603156447410583,
+      "learning_rate": 4.9565989684556405e-05,
+      "loss": 2.0099,
+      "step": 1660
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5270999073982239,
+      "learning_rate": 4.956337729949074e-05,
+      "loss": 1.9188,
+      "step": 1665
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8146567940711975,
+      "learning_rate": 4.9560757145052465e-05,
+      "loss": 2.0585,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.967383861541748,
+      "learning_rate": 4.955812922207033e-05,
+      "loss": 1.86,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5086208581924438,
+      "learning_rate": 4.955549353137554e-05,
+      "loss": 1.8033,
+      "step": 1680
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7942809462547302,
+      "learning_rate": 4.955285007380177e-05,
+      "loss": 1.9515,
+      "step": 1685
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9034421443939209,
+      "learning_rate": 4.9550198850185136e-05,
+      "loss": 1.747,
+      "step": 1690
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7966691851615906,
+      "learning_rate": 4.954753986136422e-05,
+      "loss": 2.2449,
+      "step": 1695
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7726231813430786,
+      "learning_rate": 4.954487310818006e-05,
+      "loss": 2.1002,
+      "step": 1700
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5616759061813354,
+      "learning_rate": 4.9542198591476144e-05,
+      "loss": 2.0595,
+      "step": 1705
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9209871888160706,
+      "learning_rate": 4.953951631209842e-05,
+      "loss": 1.7772,
+      "step": 1710
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5064948201179504,
+      "learning_rate": 4.953682627089529e-05,
+      "loss": 1.9936,
+      "step": 1715
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5640893578529358,
+      "learning_rate": 4.953412846871761e-05,
+      "loss": 2.0519,
+      "step": 1720
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8482123017311096,
+      "learning_rate": 4.95314229064187e-05,
+      "loss": 1.573,
+      "step": 1725
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2031537294387817,
+      "learning_rate": 4.952870958485432e-05,
+      "loss": 1.8774,
+      "step": 1730
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6398987174034119,
+      "learning_rate": 4.952598850488269e-05,
+      "loss": 2.3118,
+      "step": 1735
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2966880798339844,
+      "learning_rate": 4.952325966736449e-05,
+      "loss": 1.9382,
+      "step": 1740
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0261720418930054,
+      "learning_rate": 4.952052307316284e-05,
+      "loss": 1.8073,
+      "step": 1745
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.920386791229248,
+      "learning_rate": 4.951777872314333e-05,
+      "loss": 1.752,
+      "step": 1750
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.49061957001686096,
+      "learning_rate": 4.951502661817399e-05,
+      "loss": 2.0995,
+      "step": 1755
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8655161261558533,
+      "learning_rate": 4.951226675912532e-05,
+      "loss": 1.8597,
+      "step": 1760
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9834558367729187,
+      "learning_rate": 4.9509499146870236e-05,
+      "loss": 1.9147,
+      "step": 1765
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.557944893836975,
+      "learning_rate": 4.950672378228416e-05,
+      "loss": 1.8975,
+      "step": 1770
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3454476296901703,
+      "learning_rate": 4.950394066624492e-05,
+      "loss": 1.9411,
+      "step": 1775
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5218179225921631,
+      "learning_rate": 4.950114979963282e-05,
+      "loss": 1.8548,
+      "step": 1780
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9810051321983337,
+      "learning_rate": 4.94983511833306e-05,
+      "loss": 2.064,
+      "step": 1785
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9594024419784546,
+      "learning_rate": 4.949554481822347e-05,
+      "loss": 1.996,
+      "step": 1790
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.664357602596283,
+      "learning_rate": 4.949273070519907e-05,
+      "loss": 1.8579,
+      "step": 1795
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5438924431800842,
+      "learning_rate": 4.948990884514752e-05,
+      "loss": 1.829,
+      "step": 1800
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6013981699943542,
+      "learning_rate": 4.948707923896134e-05,
+      "loss": 1.6491,
+      "step": 1805
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6347360014915466,
+      "learning_rate": 4.948424188753556e-05,
+      "loss": 2.038,
+      "step": 1810
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27879294753074646,
+      "learning_rate": 4.948139679176762e-05,
+      "loss": 1.9321,
+      "step": 1815
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.48399272561073303,
+      "learning_rate": 4.9478543952557425e-05,
+      "loss": 1.7748,
+      "step": 1820
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5349772572517395,
+      "learning_rate": 4.9475683370807326e-05,
+      "loss": 2.0395,
+      "step": 1825
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5943443775177002,
+      "learning_rate": 4.9472815047422115e-05,
+      "loss": 1.759,
+      "step": 1830
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6108463406562805,
+      "learning_rate": 4.9469938983309045e-05,
+      "loss": 2.1658,
+      "step": 1835
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8649053573608398,
+      "learning_rate": 4.946705517937782e-05,
+      "loss": 1.9245,
+      "step": 1840
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3017230033874512,
+      "learning_rate": 4.946416363654056e-05,
+      "loss": 1.7443,
+      "step": 1845
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.48545828461647034,
+      "learning_rate": 4.946126435571188e-05,
+      "loss": 1.9818,
+      "step": 1850
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7874136567115784,
+      "learning_rate": 4.945835733780881e-05,
+      "loss": 1.8089,
+      "step": 1855
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6153427958488464,
+      "learning_rate": 4.945544258375086e-05,
+      "loss": 1.896,
+      "step": 1860
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7674331665039062,
+      "learning_rate": 4.945252009445992e-05,
+      "loss": 2.0585,
+      "step": 1865
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3186115026474,
+      "learning_rate": 4.94495898708604e-05,
+      "loss": 2.1453,
+      "step": 1870
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5109472274780273,
+      "learning_rate": 4.944665191387913e-05,
+      "loss": 1.7282,
+      "step": 1875
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5344210267066956,
+      "learning_rate": 4.9443706224445366e-05,
+      "loss": 1.8201,
+      "step": 1880
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.217705726623535,
+      "learning_rate": 4.9440752803490844e-05,
+      "loss": 2.2501,
+      "step": 1885
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9253365993499756,
+      "learning_rate": 4.9437791651949704e-05,
+      "loss": 1.9486,
+      "step": 1890
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6712260842323303,
+      "learning_rate": 4.943482277075858e-05,
+      "loss": 1.8439,
+      "step": 1895
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6539827585220337,
+      "learning_rate": 4.943184616085652e-05,
+      "loss": 1.9844,
+      "step": 1900
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4726671278476715,
+      "learning_rate": 4.9428861823185016e-05,
+      "loss": 1.7578,
+      "step": 1905
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7064640522003174,
+      "learning_rate": 4.942586975868801e-05,
+      "loss": 2.0314,
+      "step": 1910
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1073232889175415,
+      "learning_rate": 4.94228699683119e-05,
+      "loss": 1.9822,
+      "step": 1915
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4439001679420471,
+      "learning_rate": 4.941986245300552e-05,
+      "loss": 1.9749,
+      "step": 1920
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8952843546867371,
+      "learning_rate": 4.941684721372012e-05,
+      "loss": 1.9725,
+      "step": 1925
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6618747115135193,
+      "learning_rate": 4.941382425140944e-05,
+      "loss": 1.7512,
+      "step": 1930
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7568296790122986,
+      "learning_rate": 4.941079356702963e-05,
+      "loss": 1.9417,
+      "step": 1935
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7739149332046509,
+      "learning_rate": 4.9407755161539295e-05,
+      "loss": 2.0415,
+      "step": 1940
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6339311003684998,
+      "learning_rate": 4.9404709035899475e-05,
+      "loss": 2.0725,
+      "step": 1945
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8696390390396118,
+      "learning_rate": 4.940165519107367e-05,
+      "loss": 1.9008,
+      "step": 1950
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9669596552848816,
+      "learning_rate": 4.939859362802779e-05,
+      "loss": 2.1317,
+      "step": 1955
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.293050765991211,
+      "learning_rate": 4.9395524347730224e-05,
+      "loss": 2.0905,
+      "step": 1960
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.0222151279449463,
+      "learning_rate": 4.9392447351151766e-05,
+      "loss": 2.1604,
+      "step": 1965
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7745553255081177,
+      "learning_rate": 4.9389362639265673e-05,
+      "loss": 1.7385,
+      "step": 1970
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5706347227096558,
+      "learning_rate": 4.938627021304764e-05,
+      "loss": 1.6271,
+      "step": 1975
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8289479613304138,
+      "learning_rate": 4.9383170073475795e-05,
+      "loss": 2.0758,
+      "step": 1980
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.584398090839386,
+      "learning_rate": 4.938006222153071e-05,
+      "loss": 2.0172,
+      "step": 1985
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0959614515304565,
+      "learning_rate": 4.9376946658195385e-05,
+      "loss": 1.963,
+      "step": 1990
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6466994881629944,
+      "learning_rate": 4.93738233844553e-05,
+      "loss": 2.0027,
+      "step": 1995
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.194799780845642,
+      "learning_rate": 4.93706924012983e-05,
+      "loss": 2.0338,
+      "step": 2000
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.551344871520996,
+      "learning_rate": 4.9367553709714754e-05,
+      "loss": 2.0308,
+      "step": 2005
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7082229852676392,
+      "learning_rate": 4.9364407310697394e-05,
+      "loss": 1.6489,
+      "step": 2010
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5904958248138428,
+      "learning_rate": 4.9361253205241446e-05,
+      "loss": 1.8373,
+      "step": 2015
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.1805599927902222,
+      "learning_rate": 4.9358091394344543e-05,
+      "loss": 1.8094,
+      "step": 2020
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.45778775215148926,
+      "learning_rate": 4.9354921879006755e-05,
+      "loss": 1.8818,
+      "step": 2025
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8848242163658142,
+      "learning_rate": 4.93517446602306e-05,
+      "loss": 1.923,
+      "step": 2030
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4579658508300781,
+      "learning_rate": 4.934855973902105e-05,
+      "loss": 2.0447,
+      "step": 2035
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5983306169509888,
+      "learning_rate": 4.934536711638546e-05,
+      "loss": 2.0134,
+      "step": 2040
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4230965375900269,
+      "learning_rate": 4.934216679333367e-05,
+      "loss": 2.1751,
+      "step": 2045
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8285879492759705,
+      "learning_rate": 4.933895877087794e-05,
+      "loss": 1.8737,
+      "step": 2050
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.829188346862793,
+      "learning_rate": 4.933574305003296e-05,
+      "loss": 1.9408,
+      "step": 2055
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6285657286643982,
+      "learning_rate": 4.933251963181586e-05,
+      "loss": 1.8272,
+      "step": 2060
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5883930921554565,
+      "learning_rate": 4.932928851724621e-05,
+      "loss": 1.8055,
+      "step": 2065
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6375488042831421,
+      "learning_rate": 4.9326049707346e-05,
+      "loss": 1.8569,
+      "step": 2070
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6939455270767212,
+      "learning_rate": 4.9322803203139666e-05,
+      "loss": 1.5819,
+      "step": 2075
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8837314248085022,
+      "learning_rate": 4.931954900565408e-05,
+      "loss": 1.8026,
+      "step": 2080
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8050704598426819,
+      "learning_rate": 4.931628711591854e-05,
+      "loss": 1.9522,
+      "step": 2085
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8270391821861267,
+      "learning_rate": 4.931301753496476e-05,
+      "loss": 1.9587,
+      "step": 2090
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.1010689735412598,
+      "learning_rate": 4.930974026382693e-05,
+      "loss": 2.0557,
+      "step": 2095
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.389054298400879,
+      "learning_rate": 4.930645530354163e-05,
+      "loss": 1.7961,
+      "step": 2100
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7086299061775208,
+      "learning_rate": 4.93031626551479e-05,
+      "loss": 1.8631,
+      "step": 2105
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6452075242996216,
+      "learning_rate": 4.9299862319687204e-05,
+      "loss": 2.1335,
+      "step": 2110
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5691851377487183,
+      "learning_rate": 4.9296554298203423e-05,
+      "loss": 2.1724,
+      "step": 2115
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.40636715292930603,
+      "learning_rate": 4.929323859174289e-05,
+      "loss": 1.908,
+      "step": 2120
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7919933795928955,
+      "learning_rate": 4.928991520135436e-05,
+      "loss": 1.8257,
+      "step": 2125
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.567937970161438,
+      "learning_rate": 4.928658412808901e-05,
+      "loss": 1.8642,
+      "step": 2130
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.49311545491218567,
+      "learning_rate": 4.9283245373000465e-05,
+      "loss": 1.8398,
+      "step": 2135
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7059115767478943,
+      "learning_rate": 4.927989893714477e-05,
+      "loss": 1.879,
+      "step": 2140
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5949462652206421,
+      "learning_rate": 4.927654482158039e-05,
+      "loss": 1.9516,
+      "step": 2145
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8240839838981628,
+      "learning_rate": 4.927318302736824e-05,
+      "loss": 2.192,
+      "step": 2150
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7169739603996277,
+      "learning_rate": 4.926981355557164e-05,
+      "loss": 1.7331,
+      "step": 2155
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.85520076751709,
+      "learning_rate": 4.926643640725635e-05,
+      "loss": 1.9134,
+      "step": 2160
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6009440422058105,
+      "learning_rate": 4.9263051583490584e-05,
+      "loss": 1.8528,
+      "step": 2165
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7495572566986084,
+      "learning_rate": 4.9259659085344925e-05,
+      "loss": 1.7285,
+      "step": 2170
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4452281892299652,
+      "learning_rate": 4.925625891389244e-05,
+      "loss": 1.9832,
+      "step": 2175
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6736578345298767,
+      "learning_rate": 4.925285107020859e-05,
+      "loss": 2.4924,
+      "step": 2180
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6530075073242188,
+      "learning_rate": 4.924943555537128e-05,
+      "loss": 2.2028,
+      "step": 2185
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.629514753818512,
+      "learning_rate": 4.924601237046082e-05,
+      "loss": 1.9472,
+      "step": 2190
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8466399908065796,
+      "learning_rate": 4.924258151655997e-05,
+      "loss": 1.8294,
+      "step": 2195
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8130293488502502,
+      "learning_rate": 4.92391429947539e-05,
+      "loss": 2.1553,
+      "step": 2200
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7702954411506653,
+      "learning_rate": 4.923569680613023e-05,
+      "loss": 2.0827,
+      "step": 2205
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4999507963657379,
+      "learning_rate": 4.923224295177897e-05,
+      "loss": 1.736,
+      "step": 2210
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1567188501358032,
+      "learning_rate": 4.922878143279256e-05,
+      "loss": 1.7352,
+      "step": 2215
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.736792802810669,
+      "learning_rate": 4.92253122502659e-05,
+      "loss": 1.606,
+      "step": 2220
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5883451700210571,
+      "learning_rate": 4.922183540529627e-05,
+      "loss": 1.8151,
+      "step": 2225
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6268717646598816,
+      "learning_rate": 4.921835089898341e-05,
+      "loss": 2.3119,
+      "step": 2230
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0409209728240967,
+      "learning_rate": 4.9214858732429444e-05,
+      "loss": 1.9868,
+      "step": 2235
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6836051344871521,
+      "learning_rate": 4.921135890673896e-05,
+      "loss": 1.8223,
+      "step": 2240
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6672108173370361,
+      "learning_rate": 4.920785142301893e-05,
+      "loss": 2.0905,
+      "step": 2245
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7968783378601074,
+      "learning_rate": 4.920433628237879e-05,
+      "loss": 1.8637,
+      "step": 2250
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7804060578346252,
+      "learning_rate": 4.9200813485930375e-05,
+      "loss": 1.749,
+      "step": 2255
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4104400873184204,
+      "learning_rate": 4.9197283034787925e-05,
+      "loss": 1.8482,
+      "step": 2260
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.480487585067749,
+      "learning_rate": 4.919374493006812e-05,
+      "loss": 1.966,
+      "step": 2265
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7626844644546509,
+      "learning_rate": 4.919019917289006e-05,
+      "loss": 1.796,
+      "step": 2270
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4077138304710388,
+      "learning_rate": 4.918664576437528e-05,
+      "loss": 2.1519,
+      "step": 2275
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5030271410942078,
+      "learning_rate": 4.918308470564771e-05,
+      "loss": 1.8834,
+      "step": 2280
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5887728929519653,
+      "learning_rate": 4.9179515997833706e-05,
+      "loss": 2.0374,
+      "step": 2285
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.766281247138977,
+      "learning_rate": 4.917593964206205e-05,
+      "loss": 1.8359,
+      "step": 2290
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6358659267425537,
+      "learning_rate": 4.917235563946394e-05,
+      "loss": 1.7785,
+      "step": 2295
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4064447283744812,
+      "learning_rate": 4.916876399117299e-05,
+      "loss": 2.0286,
+      "step": 2300
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7041420936584473,
+      "learning_rate": 4.916516469832524e-05,
+      "loss": 1.7674,
+      "step": 2305
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7361471056938171,
+      "learning_rate": 4.916155776205913e-05,
+      "loss": 1.7543,
+      "step": 2310
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4390116930007935,
+      "learning_rate": 4.915794318351555e-05,
+      "loss": 1.7741,
+      "step": 2315
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5468139052391052,
+      "learning_rate": 4.915432096383779e-05,
+      "loss": 1.8634,
+      "step": 2320
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.36798736453056335,
+      "learning_rate": 4.915069110417152e-05,
+      "loss": 1.9809,
+      "step": 2325
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6476161479949951,
+      "learning_rate": 4.914705360566489e-05,
+      "loss": 1.8003,
+      "step": 2330
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4177202880382538,
+      "learning_rate": 4.914340846946844e-05,
+      "loss": 1.8988,
+      "step": 2335
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6585730314254761,
+      "learning_rate": 4.91397556967351e-05,
+      "loss": 2.0532,
+      "step": 2340
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7709575295448303,
+      "learning_rate": 4.913609528862027e-05,
+      "loss": 1.8792,
+      "step": 2345
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.782526969909668,
+      "learning_rate": 4.9132427246281705e-05,
+      "loss": 1.7209,
+      "step": 2350
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9264500737190247,
+      "learning_rate": 4.9128751570879616e-05,
+      "loss": 1.9869,
+      "step": 2355
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5954318642616272,
+      "learning_rate": 4.912506826357661e-05,
+      "loss": 1.9642,
+      "step": 2360
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.49041104316711426,
+      "learning_rate": 4.9121377325537724e-05,
+      "loss": 1.6972,
+      "step": 2365
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8424311280250549,
+      "learning_rate": 4.911767875793039e-05,
+      "loss": 1.7111,
+      "step": 2370
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7687463760375977,
+      "learning_rate": 4.911397256192446e-05,
+      "loss": 1.8755,
+      "step": 2375
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6133802533149719,
+      "learning_rate": 4.9110258738692204e-05,
+      "loss": 2.5453,
+      "step": 2380
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6727443337440491,
+      "learning_rate": 4.9106537289408305e-05,
+      "loss": 1.6229,
+      "step": 2385
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0047177076339722,
+      "learning_rate": 4.910280821524985e-05,
+      "loss": 2.0174,
+      "step": 2390
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6900597810745239,
+      "learning_rate": 4.909907151739633e-05,
+      "loss": 1.8986,
+      "step": 2395
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.730687141418457,
+      "learning_rate": 4.909532719702968e-05,
+      "loss": 1.8659,
+      "step": 2400
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8315011858940125,
+      "learning_rate": 4.909157525533421e-05,
+      "loss": 1.6193,
+      "step": 2405
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3859171867370605,
+      "learning_rate": 4.908781569349666e-05,
+      "loss": 2.1098,
+      "step": 2410
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7538830637931824,
+      "learning_rate": 4.9084048512706174e-05,
+      "loss": 1.7888,
+      "step": 2415
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.840951681137085,
+      "learning_rate": 4.908027371415431e-05,
+      "loss": 1.6484,
+      "step": 2420
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.1713069677352905,
+      "learning_rate": 4.907649129903504e-05,
+      "loss": 1.9293,
+      "step": 2425
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6097638010978699,
+      "learning_rate": 4.9072701268544726e-05,
+      "loss": 1.9388,
+      "step": 2430
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6106088161468506,
+      "learning_rate": 4.906890362388215e-05,
+      "loss": 1.6406,
+      "step": 2435
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9904814958572388,
+      "learning_rate": 4.906509836624852e-05,
+      "loss": 1.6748,
+      "step": 2440
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0274256467819214,
+      "learning_rate": 4.906128549684741e-05,
+      "loss": 1.9952,
+      "step": 2445
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5849592685699463,
+      "learning_rate": 4.905746501688485e-05,
+      "loss": 1.7149,
+      "step": 2450
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4269081354141235,
+      "learning_rate": 4.9053636927569246e-05,
+      "loss": 2.0261,
+      "step": 2455
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9268386363983154,
+      "learning_rate": 4.904980123011142e-05,
+      "loss": 1.6845,
+      "step": 2460
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6730044484138489,
+      "learning_rate": 4.904595792572459e-05,
+      "loss": 1.9066,
+      "step": 2465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5785349011421204,
+      "learning_rate": 4.9042107015624405e-05,
+      "loss": 1.9116,
+      "step": 2470
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.519561767578125,
+      "learning_rate": 4.903824850102889e-05,
+      "loss": 1.9645,
+      "step": 2475
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5183913111686707,
+      "learning_rate": 4.903438238315851e-05,
+      "loss": 2.0975,
+      "step": 2480
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6047394871711731,
+      "learning_rate": 4.903050866323609e-05,
+      "loss": 1.9343,
+      "step": 2485
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.560421109199524,
+      "learning_rate": 4.9026627342486886e-05,
+      "loss": 1.9887,
+      "step": 2490
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5357934236526489,
+      "learning_rate": 4.902273842213857e-05,
+      "loss": 2.06,
+      "step": 2495
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7645955085754395,
+      "learning_rate": 4.901884190342121e-05,
+      "loss": 1.7665,
+      "step": 2500
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7442848086357117,
+      "learning_rate": 4.901493778756725e-05,
+      "loss": 1.7705,
+      "step": 2505
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5181974768638611,
+      "learning_rate": 4.901102607581156e-05,
+      "loss": 2.1552,
+      "step": 2510
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5318804979324341,
+      "learning_rate": 4.900710676939143e-05,
+      "loss": 1.925,
+      "step": 2515
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.42624104022979736,
+      "learning_rate": 4.900317986954651e-05,
+      "loss": 1.8521,
+      "step": 2520
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8136389255523682,
+      "learning_rate": 4.89992453775189e-05,
+      "loss": 1.8125,
+      "step": 2525
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8504977226257324,
+      "learning_rate": 4.8995303294553054e-05,
+      "loss": 1.704,
+      "step": 2530
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5383654236793518,
+      "learning_rate": 4.899135362189585e-05,
+      "loss": 1.8018,
+      "step": 2535
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.590053915977478,
+      "learning_rate": 4.898739636079658e-05,
+      "loss": 1.7633,
+      "step": 2540
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6471716165542603,
+      "learning_rate": 4.898343151250691e-05,
+      "loss": 1.8805,
+      "step": 2545
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8511287569999695,
+      "learning_rate": 4.8979459078280924e-05,
+      "loss": 1.5846,
+      "step": 2550
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3268200755119324,
+      "learning_rate": 4.8975479059375093e-05,
+      "loss": 1.6611,
+      "step": 2555
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7143041491508484,
+      "learning_rate": 4.8971491457048305e-05,
+      "loss": 1.9187,
+      "step": 2560
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5510767698287964,
+      "learning_rate": 4.896749627256182e-05,
+      "loss": 2.0012,
+      "step": 2565
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.34194180369377136,
+      "learning_rate": 4.8963493507179314e-05,
+      "loss": 1.8626,
+      "step": 2570
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2060723304748535,
+      "learning_rate": 4.895948316216687e-05,
+      "loss": 1.7144,
+      "step": 2575
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.33225440979003906,
+      "learning_rate": 4.895546523879294e-05,
+      "loss": 1.709,
+      "step": 2580
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.596155047416687,
+      "learning_rate": 4.895143973832841e-05,
+      "loss": 2.3011,
+      "step": 2585
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8214730620384216,
+      "learning_rate": 4.8947406662046516e-05,
+      "loss": 1.7885,
+      "step": 2590
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9236437082290649,
+      "learning_rate": 4.894336601122293e-05,
+      "loss": 2.1127,
+      "step": 2595
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6400870680809021,
+      "learning_rate": 4.893931778713572e-05,
+      "loss": 2.0813,
+      "step": 2600
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.41277340054512024,
+      "learning_rate": 4.893526199106531e-05,
+      "loss": 2.1293,
+      "step": 2605
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.0775052309036255,
+      "learning_rate": 4.893119862429455e-05,
+      "loss": 1.453,
+      "step": 2610
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6560648083686829,
+      "learning_rate": 4.892712768810869e-05,
+      "loss": 1.6368,
+      "step": 2615
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7044461369514465,
+      "learning_rate": 4.8923049183795355e-05,
+      "loss": 1.9884,
+      "step": 2620
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6141453385353088,
+      "learning_rate": 4.8918963112644576e-05,
+      "loss": 1.7104,
+      "step": 2625
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2917144298553467,
+      "learning_rate": 4.891486947594878e-05,
+      "loss": 1.9332,
+      "step": 2630
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.97393798828125,
+      "learning_rate": 4.8910768275002763e-05,
+      "loss": 1.9389,
+      "step": 2635
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5923995971679688,
+      "learning_rate": 4.890665951110374e-05,
+      "loss": 2.0709,
+      "step": 2640
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.865432620048523,
+      "learning_rate": 4.890254318555131e-05,
+      "loss": 2.1132,
+      "step": 2645
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7784671187400818,
+      "learning_rate": 4.889841929964746e-05,
+      "loss": 1.7297,
+      "step": 2650
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5125192999839783,
+      "learning_rate": 4.889428785469657e-05,
+      "loss": 1.9602,
+      "step": 2655
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7674616575241089,
+      "learning_rate": 4.8890148852005416e-05,
+      "loss": 1.9787,
+      "step": 2660
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3875999450683594,
+      "learning_rate": 4.888600229288317e-05,
+      "loss": 1.7716,
+      "step": 2665
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8313968181610107,
+      "learning_rate": 4.8881848178641364e-05,
+      "loss": 1.4916,
+      "step": 2670
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8791585564613342,
+      "learning_rate": 4.887768651059395e-05,
+      "loss": 1.9918,
+      "step": 2675
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9071698188781738,
+      "learning_rate": 4.887351729005726e-05,
+      "loss": 2.0969,
+      "step": 2680
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9274541139602661,
+      "learning_rate": 4.8869340518350024e-05,
+      "loss": 2.0151,
+      "step": 2685
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5907272100448608,
+      "learning_rate": 4.8865156196793334e-05,
+      "loss": 2.1632,
+      "step": 2690
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.831947922706604,
+      "learning_rate": 4.88609643267107e-05,
+      "loss": 2.2977,
+      "step": 2695
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7203566431999207,
+      "learning_rate": 4.885676490942799e-05,
+      "loss": 2.071,
+      "step": 2700
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.22580087184906,
+      "learning_rate": 4.885255794627349e-05,
+      "loss": 1.9375,
+      "step": 2705
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7402857542037964,
+      "learning_rate": 4.884834343857786e-05,
+      "loss": 2.1846,
+      "step": 2710
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.610673189163208,
+      "learning_rate": 4.884412138767414e-05,
+      "loss": 1.9636,
+      "step": 2715
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7960859537124634,
+      "learning_rate": 4.883989179489775e-05,
+      "loss": 2.0377,
+      "step": 2720
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5835393071174622,
+      "learning_rate": 4.883565466158653e-05,
+      "loss": 2.1319,
+      "step": 2725
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7329149842262268,
+      "learning_rate": 4.8831409989080656e-05,
+      "loss": 1.7587,
+      "step": 2730
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5767766833305359,
+      "learning_rate": 4.882715777872273e-05,
+      "loss": 1.9912,
+      "step": 2735
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.826280951499939,
+      "learning_rate": 4.882289803185772e-05,
+      "loss": 1.6964,
+      "step": 2740
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6336571574211121,
+      "learning_rate": 4.881863074983298e-05,
+      "loss": 2.1097,
+      "step": 2745
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4215265214443207,
+      "learning_rate": 4.8814355933998235e-05,
+      "loss": 1.8692,
+      "step": 2750
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4438379108905792,
+      "learning_rate": 4.881007358570562e-05,
+      "loss": 1.797,
+      "step": 2755
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6177186369895935,
+      "learning_rate": 4.880578370630963e-05,
+      "loss": 1.7992,
+      "step": 2760
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6064658761024475,
+      "learning_rate": 4.8801486297167156e-05,
+      "loss": 1.9718,
+      "step": 2765
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5778130888938904,
+      "learning_rate": 4.879718135963746e-05,
+      "loss": 2.0216,
+      "step": 2770
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0478150844573975,
+      "learning_rate": 4.879286889508219e-05,
+      "loss": 1.6562,
+      "step": 2775
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9771472811698914,
+      "learning_rate": 4.8788548904865375e-05,
+      "loss": 1.8386,
+      "step": 2780
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5626159310340881,
+      "learning_rate": 4.878422139035341e-05,
+      "loss": 2.2397,
+      "step": 2785
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8984270691871643,
+      "learning_rate": 4.877988635291511e-05,
+      "loss": 1.8268,
+      "step": 2790
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7879625558853149,
+      "learning_rate": 4.877554379392163e-05,
+      "loss": 1.9463,
+      "step": 2795
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.40900498628616333,
+      "learning_rate": 4.877119371474651e-05,
+      "loss": 1.8477,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.685408353805542,
+      "learning_rate": 4.876683611676569e-05,
+      "loss": 1.9898,
+      "step": 2805
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5986139178276062,
+      "learning_rate": 4.876247100135746e-05,
+      "loss": 2.0433,
+      "step": 2810
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7480238676071167,
+      "learning_rate": 4.875809836990252e-05,
+      "loss": 1.9656,
+      "step": 2815
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.595890522003174,
+      "learning_rate": 4.8753718223783905e-05,
+      "loss": 1.8644,
+      "step": 2820
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7732789516448975,
+      "learning_rate": 4.874933056438707e-05,
+      "loss": 1.8836,
+      "step": 2825
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.520476222038269,
+      "learning_rate": 4.874493539309982e-05,
+      "loss": 2.0753,
+      "step": 2830
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2731196880340576,
+      "learning_rate": 4.874053271131235e-05,
+      "loss": 2.1824,
+      "step": 2835
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.522787094116211,
+      "learning_rate": 4.8736122520417215e-05,
+      "loss": 1.776,
+      "step": 2840
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6246161460876465,
+      "learning_rate": 4.8731704821809354e-05,
+      "loss": 1.8492,
+      "step": 2845
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5303625464439392,
+      "learning_rate": 4.872727961688609e-05,
+      "loss": 1.7064,
+      "step": 2850
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.939147412776947,
+      "learning_rate": 4.87228469070471e-05,
+      "loss": 1.8514,
+      "step": 2855
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5877857208251953,
+      "learning_rate": 4.8718406693694465e-05,
+      "loss": 1.7117,
+      "step": 2860
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4529130458831787,
+      "learning_rate": 4.87139589782326e-05,
+      "loss": 1.8436,
+      "step": 2865
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8624293208122253,
+      "learning_rate": 4.870950376206831e-05,
+      "loss": 1.8674,
+      "step": 2870
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6913080811500549,
+      "learning_rate": 4.87050410466108e-05,
+      "loss": 2.4175,
+      "step": 2875
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0130589008331299,
+      "learning_rate": 4.87005708332716e-05,
+      "loss": 1.8629,
+      "step": 2880
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.40573757886886597,
+      "learning_rate": 4.869609312346465e-05,
+      "loss": 2.1532,
+      "step": 2885
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0451394319534302,
+      "learning_rate": 4.869160791860623e-05,
+      "loss": 1.9334,
+      "step": 2890
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0864038467407227,
+      "learning_rate": 4.868711522011502e-05,
+      "loss": 1.7183,
+      "step": 2895
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2843054234981537,
+      "learning_rate": 4.868261502941204e-05,
+      "loss": 2.0684,
+      "step": 2900
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6705717444419861,
+      "learning_rate": 4.867810734792072e-05,
+      "loss": 1.8852,
+      "step": 2905
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8164358735084534,
+      "learning_rate": 4.867359217706681e-05,
+      "loss": 2.0016,
+      "step": 2910
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6517994403839111,
+      "learning_rate": 4.866906951827847e-05,
+      "loss": 1.936,
+      "step": 2915
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3796793818473816,
+      "learning_rate": 4.866453937298621e-05,
+      "loss": 1.9839,
+      "step": 2920
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5019554495811462,
+      "learning_rate": 4.866000174262291e-05,
+      "loss": 1.6889,
+      "step": 2925
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4469722509384155,
+      "learning_rate": 4.865545662862381e-05,
+      "loss": 1.9799,
+      "step": 2930
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0865347385406494,
+      "learning_rate": 4.865090403242654e-05,
+      "loss": 1.7499,
+      "step": 2935
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9820341467857361,
+      "learning_rate": 4.864634395547106e-05,
+      "loss": 1.9783,
+      "step": 2940
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.41876184940338135,
+      "learning_rate": 4.8641776399199743e-05,
+      "loss": 1.991,
+      "step": 2945
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.171912431716919,
+      "learning_rate": 4.863720136505729e-05,
+      "loss": 1.7386,
+      "step": 2950
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8185043931007385,
+      "learning_rate": 4.863261885449078e-05,
+      "loss": 1.852,
+      "step": 2955
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5740297436714172,
+      "learning_rate": 4.862802886894966e-05,
+      "loss": 1.4815,
+      "step": 2960
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5629348754882812,
+      "learning_rate": 4.862343140988573e-05,
+      "loss": 1.9167,
+      "step": 2965
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5648974180221558,
+      "learning_rate": 4.861882647875318e-05,
+      "loss": 2.1185,
+      "step": 2970
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.46517616510391235,
+      "learning_rate": 4.861421407700853e-05,
+      "loss": 1.8994,
+      "step": 2975
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4326724410057068,
+      "learning_rate": 4.860959420611067e-05,
+      "loss": 1.9796,
+      "step": 2980
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0473462343215942,
+      "learning_rate": 4.860496686752088e-05,
+      "loss": 1.5259,
+      "step": 2985
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6728710532188416,
+      "learning_rate": 4.860033206270277e-05,
+      "loss": 2.0169,
+      "step": 2990
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2514013051986694,
+      "learning_rate": 4.859568979312233e-05,
+      "loss": 1.8368,
+      "step": 2995
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6908515095710754,
+      "learning_rate": 4.8591040060247914e-05,
+      "loss": 2.2935,
+      "step": 3000
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.934448778629303,
+      "learning_rate": 4.8586382865550216e-05,
+      "loss": 1.9923,
+      "step": 3005
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7105411887168884,
+      "learning_rate": 4.85817182105023e-05,
+      "loss": 1.8267,
+      "step": 3010
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6507242321968079,
+      "learning_rate": 4.8577046096579596e-05,
+      "loss": 1.7661,
+      "step": 3015
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.576221227645874,
+      "learning_rate": 4.857236652525989e-05,
+      "loss": 1.7314,
+      "step": 3020
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.374879002571106,
+      "learning_rate": 4.856767949802333e-05,
+      "loss": 2.1544,
+      "step": 3025
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6787958145141602,
+      "learning_rate": 4.856298501635241e-05,
+      "loss": 1.5669,
+      "step": 3030
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8718627095222473,
+      "learning_rate": 4.8558283081732004e-05,
+      "loss": 1.8672,
+      "step": 3035
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5774152278900146,
+      "learning_rate": 4.855357369564931e-05,
+      "loss": 1.7822,
+      "step": 3040
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5398827791213989,
+      "learning_rate": 4.854885685959391e-05,
+      "loss": 1.974,
+      "step": 3045
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1588878631591797,
+      "learning_rate": 4.8544132575057745e-05,
+      "loss": 1.5734,
+      "step": 3050
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9843418598175049,
+      "learning_rate": 4.8539400843535086e-05,
+      "loss": 2.267,
+      "step": 3055
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5535194277763367,
+      "learning_rate": 4.8534661666522584e-05,
+      "loss": 1.5894,
+      "step": 3060
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1449483633041382,
+      "learning_rate": 4.852991504551924e-05,
+      "loss": 1.7232,
+      "step": 3065
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4027251601219177,
+      "learning_rate": 4.852516098202639e-05,
+      "loss": 1.934,
+      "step": 3070
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9327101707458496,
+      "learning_rate": 4.852039947754775e-05,
+      "loss": 1.9557,
+      "step": 3075
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7538480162620544,
+      "learning_rate": 4.851563053358938e-05,
+      "loss": 1.759,
+      "step": 3080
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.788642942905426,
+      "learning_rate": 4.85108541516597e-05,
+      "loss": 1.9517,
+      "step": 3085
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7686721086502075,
+      "learning_rate": 4.850607033326946e-05,
+      "loss": 1.676,
+      "step": 3090
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1780756711959839,
+      "learning_rate": 4.850127907993178e-05,
+      "loss": 2.0255,
+      "step": 3095
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5100767612457275,
+      "learning_rate": 4.8496480393162144e-05,
+      "loss": 1.7199,
+      "step": 3100
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0231378078460693,
+      "learning_rate": 4.849167427447835e-05,
+      "loss": 2.0379,
+      "step": 3105
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6608045697212219,
+      "learning_rate": 4.8486860725400584e-05,
+      "loss": 1.9066,
+      "step": 3110
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8580523133277893,
+      "learning_rate": 4.848203974745136e-05,
+      "loss": 1.842,
+      "step": 3115
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7334781885147095,
+      "learning_rate": 4.847721134215555e-05,
+      "loss": 2.1667,
+      "step": 3120
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.904914379119873,
+      "learning_rate": 4.847237551104037e-05,
+      "loss": 2.0194,
+      "step": 3125
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9545987248420715,
+      "learning_rate": 4.846753225563541e-05,
+      "loss": 1.9362,
+      "step": 3130
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.530463457107544,
+      "learning_rate": 4.846268157747256e-05,
+      "loss": 2.1253,
+      "step": 3135
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.36319640278816223,
+      "learning_rate": 4.845782347808609e-05,
+      "loss": 1.9925,
+      "step": 3140
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2937252521514893,
+      "learning_rate": 4.845295795901262e-05,
+      "loss": 2.2135,
+      "step": 3145
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6018854379653931,
+      "learning_rate": 4.8448085021791115e-05,
+      "loss": 1.7548,
+      "step": 3150
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.49405333399772644,
+      "learning_rate": 4.844320466796286e-05,
+      "loss": 1.9319,
+      "step": 3155
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0258450508117676,
+      "learning_rate": 4.8438316899071525e-05,
+      "loss": 1.969,
+      "step": 3160
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.1361362934112549,
+      "learning_rate": 4.84334217166631e-05,
+      "loss": 1.9683,
+      "step": 3165
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8203340768814087,
+      "learning_rate": 4.842851912228592e-05,
+      "loss": 2.0704,
+      "step": 3170
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7416129112243652,
+      "learning_rate": 4.8423609117490686e-05,
+      "loss": 1.8224,
+      "step": 3175
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2014813423156738,
+      "learning_rate": 4.841869170383041e-05,
+      "loss": 1.8246,
+      "step": 3180
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5674131512641907,
+      "learning_rate": 4.841376688286048e-05,
+      "loss": 2.114,
+      "step": 3185
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6258872151374817,
+      "learning_rate": 4.8408834656138605e-05,
+      "loss": 1.9057,
+      "step": 3190
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6485921144485474,
+      "learning_rate": 4.840389502522484e-05,
+      "loss": 1.9694,
+      "step": 3195
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.786283016204834,
+      "learning_rate": 4.83989479916816e-05,
+      "loss": 1.9653,
+      "step": 3200
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5875077843666077,
+      "learning_rate": 4.8393993557073605e-05,
+      "loss": 1.8774,
+      "step": 3205
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5160438418388367,
+      "learning_rate": 4.838903172296796e-05,
+      "loss": 1.9262,
+      "step": 3210
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4956705868244171,
+      "learning_rate": 4.838406249093408e-05,
+      "loss": 2.0985,
+      "step": 3215
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6459410190582275,
+      "learning_rate": 4.837908586254372e-05,
+      "loss": 1.9046,
+      "step": 3220
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8042646050453186,
+      "learning_rate": 4.837410183937099e-05,
+      "loss": 2.1288,
+      "step": 3225
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6269195079803467,
+      "learning_rate": 4.836911042299233e-05,
+      "loss": 2.0259,
+      "step": 3230
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8424185514450073,
+      "learning_rate": 4.8364111614986527e-05,
+      "loss": 1.6815,
+      "step": 3235
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9412960410118103,
+      "learning_rate": 4.835910541693469e-05,
+      "loss": 1.8376,
+      "step": 3240
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6458824276924133,
+      "learning_rate": 4.835409183042028e-05,
+      "loss": 1.803,
+      "step": 3245
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5122315287590027,
+      "learning_rate": 4.834907085702908e-05,
+      "loss": 2.0321,
+      "step": 3250
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6900401711463928,
+      "learning_rate": 4.834404249834924e-05,
+      "loss": 2.0708,
+      "step": 3255
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.9199984669685364,
+      "learning_rate": 4.83390067559712e-05,
+      "loss": 1.7106,
+      "step": 3260
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7172281742095947,
+      "learning_rate": 4.8333963631487776e-05,
+      "loss": 2.0546,
+      "step": 3265
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.753948986530304,
+      "learning_rate": 4.8328913126494104e-05,
+      "loss": 1.8142,
+      "step": 3270
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8077734708786011,
+      "learning_rate": 4.8323855242587644e-05,
+      "loss": 1.908,
+      "step": 3275
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6292251944541931,
+      "learning_rate": 4.831878998136821e-05,
+      "loss": 2.0195,
+      "step": 3280
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6925063133239746,
+      "learning_rate": 4.8313717344437936e-05,
+      "loss": 1.9374,
+      "step": 3285
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5431442856788635,
+      "learning_rate": 4.8308637333401284e-05,
+      "loss": 1.9119,
+      "step": 3290
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8268663883209229,
+      "learning_rate": 4.830354994986507e-05,
+      "loss": 1.8918,
+      "step": 3295
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5958759784698486,
+      "learning_rate": 4.829845519543842e-05,
+      "loss": 1.6231,
+      "step": 3300
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7670718431472778,
+      "learning_rate": 4.82933530717328e-05,
+      "loss": 1.8069,
+      "step": 3305
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5632594227790833,
+      "learning_rate": 4.828824358036201e-05,
+      "loss": 1.8654,
+      "step": 3310
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0903352499008179,
+      "learning_rate": 4.8283126722942176e-05,
+      "loss": 2.0945,
+      "step": 3315
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4936974346637726,
+      "learning_rate": 4.8278002501091755e-05,
+      "loss": 2.1163,
+      "step": 3320
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2784171402454376,
+      "learning_rate": 4.827287091643154e-05,
+      "loss": 1.9931,
+      "step": 3325
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8564805388450623,
+      "learning_rate": 4.8267731970584626e-05,
+      "loss": 2.1167,
+      "step": 3330
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2154970169067383,
+      "learning_rate": 4.8262585665176485e-05,
+      "loss": 1.7391,
+      "step": 3335
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9126550555229187,
+      "learning_rate": 4.825743200183486e-05,
+      "loss": 2.0901,
+      "step": 3340
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4065916836261749,
+      "learning_rate": 4.825227098218987e-05,
+      "loss": 1.991,
+      "step": 3345
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.732071578502655,
+      "learning_rate": 4.824710260787394e-05,
+      "loss": 2.0214,
+      "step": 3350
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5489442348480225,
+      "learning_rate": 4.824192688052181e-05,
+      "loss": 1.8027,
+      "step": 3355
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5446940660476685,
+      "learning_rate": 4.8236743801770565e-05,
+      "loss": 1.7853,
+      "step": 3360
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5271531343460083,
+      "learning_rate": 4.823155337325961e-05,
+      "loss": 1.8151,
+      "step": 3365
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5761736631393433,
+      "learning_rate": 4.8226355596630655e-05,
+      "loss": 1.6954,
+      "step": 3370
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9606974124908447,
+      "learning_rate": 4.8221150473527774e-05,
+      "loss": 1.7548,
+      "step": 3375
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4853847026824951,
+      "learning_rate": 4.821593800559734e-05,
+      "loss": 1.9427,
+      "step": 3380
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9224936366081238,
+      "learning_rate": 4.8210718194488046e-05,
+      "loss": 1.707,
+      "step": 3385
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6053968071937561,
+      "learning_rate": 4.82054910418509e-05,
+      "loss": 2.1525,
+      "step": 3390
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5924646258354187,
+      "learning_rate": 4.8200256549339274e-05,
+      "loss": 1.6875,
+      "step": 3395
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.748364448547363,
+      "learning_rate": 4.8195014718608814e-05,
+      "loss": 1.9114,
+      "step": 3400
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8054745197296143,
+      "learning_rate": 4.81897655513175e-05,
+      "loss": 1.935,
+      "step": 3405
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8999778032302856,
+      "learning_rate": 4.8184509049125645e-05,
+      "loss": 1.9618,
+      "step": 3410
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8038251996040344,
+      "learning_rate": 4.8179245213695887e-05,
+      "loss": 1.5852,
+      "step": 3415
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.1532466411590576,
+      "learning_rate": 4.817397404669317e-05,
+      "loss": 1.7721,
+      "step": 3420
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.0493075847625732,
+      "learning_rate": 4.816869554978475e-05,
+      "loss": 1.8461,
+      "step": 3425
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7406529784202576,
+      "learning_rate": 4.81634097246402e-05,
+      "loss": 1.9881,
+      "step": 3430
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6479585766792297,
+      "learning_rate": 4.8158116572931445e-05,
+      "loss": 1.7853,
+      "step": 3435
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9244452118873596,
+      "learning_rate": 4.815281609633269e-05,
+      "loss": 1.7463,
+      "step": 3440
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5907859802246094,
+      "learning_rate": 4.814750829652048e-05,
+      "loss": 1.9086,
+      "step": 3445
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5566309094429016,
+      "learning_rate": 4.814219317517365e-05,
+      "loss": 2.0991,
+      "step": 3450
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7224152088165283,
+      "learning_rate": 4.813687073397339e-05,
+      "loss": 1.9487,
+      "step": 3455
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6035553216934204,
+      "learning_rate": 4.8131540974603164e-05,
+      "loss": 2.0636,
+      "step": 3460
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7073687314987183,
+      "learning_rate": 4.8126203898748776e-05,
+      "loss": 1.9388,
+      "step": 3465
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.752922534942627,
+      "learning_rate": 4.812085950809835e-05,
+      "loss": 1.5911,
+      "step": 3470
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6370113492012024,
+      "learning_rate": 4.811550780434229e-05,
+      "loss": 1.7502,
+      "step": 3475
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7335951924324036,
+      "learning_rate": 4.811014878917335e-05,
+      "loss": 1.8573,
+      "step": 3480
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6506394743919373,
+      "learning_rate": 4.810478246428658e-05,
+      "loss": 1.7608,
+      "step": 3485
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5935482382774353,
+      "learning_rate": 4.8099408831379335e-05,
+      "loss": 1.9053,
+      "step": 3490
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0178114175796509,
+      "learning_rate": 4.80940278921513e-05,
+      "loss": 1.9641,
+      "step": 3495
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5880244970321655,
+      "learning_rate": 4.8088639648304455e-05,
+      "loss": 1.6671,
+      "step": 3500
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6311240792274475,
+      "learning_rate": 4.80832441015431e-05,
+      "loss": 1.6607,
+      "step": 3505
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4380526542663574,
+      "learning_rate": 4.807784125357383e-05,
+      "loss": 2.2915,
+      "step": 3510
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7547699213027954,
+      "learning_rate": 4.8072431106105577e-05,
+      "loss": 1.7945,
+      "step": 3515
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.0797021389007568,
+      "learning_rate": 4.806701366084956e-05,
+      "loss": 1.9746,
+      "step": 3520
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3622939586639404,
+      "learning_rate": 4.8061588919519304e-05,
+      "loss": 1.7607,
+      "step": 3525
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4650992751121521,
+      "learning_rate": 4.805615688383066e-05,
+      "loss": 1.9469,
+      "step": 3530
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.421215683221817,
+      "learning_rate": 4.805071755550177e-05,
+      "loss": 1.7305,
+      "step": 3535
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6225435733795166,
+      "learning_rate": 4.8045270936253086e-05,
+      "loss": 1.6456,
+      "step": 3540
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6064786314964294,
+      "learning_rate": 4.8039817027807374e-05,
+      "loss": 1.8573,
+      "step": 3545
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.45190367102622986,
+      "learning_rate": 4.803435583188969e-05,
+      "loss": 1.9488,
+      "step": 3550
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.44525858759880066,
+      "learning_rate": 4.8028887350227424e-05,
+      "loss": 2.1564,
+      "step": 3555
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6965821385383606,
+      "learning_rate": 4.802341158455024e-05,
+      "loss": 1.7877,
+      "step": 3560
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4658873975276947,
+      "learning_rate": 4.80179285365901e-05,
+      "loss": 1.9438,
+      "step": 3565
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8804203867912292,
+      "learning_rate": 4.801243820808132e-05,
+      "loss": 1.8994,
+      "step": 3570
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5602685809135437,
+      "learning_rate": 4.8006940600760475e-05,
+      "loss": 1.8585,
+      "step": 3575
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7118222713470459,
+      "learning_rate": 4.800143571636644e-05,
+      "loss": 1.9065,
+      "step": 3580
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5043877959251404,
+      "learning_rate": 4.799592355664041e-05,
+      "loss": 1.8838,
+      "step": 3585
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5556797981262207,
+      "learning_rate": 4.799040412332588e-05,
+      "loss": 1.8743,
+      "step": 3590
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5163377523422241,
+      "learning_rate": 4.798487741816864e-05,
+      "loss": 1.888,
+      "step": 3595
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7609609961509705,
+      "learning_rate": 4.797934344291678e-05,
+      "loss": 2.0205,
+      "step": 3600
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6983132362365723,
+      "learning_rate": 4.797380219932069e-05,
+      "loss": 1.9689,
+      "step": 3605
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.0645403861999512,
+      "learning_rate": 4.7968253689133074e-05,
+      "loss": 1.8991,
+      "step": 3610
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5911131501197815,
+      "learning_rate": 4.79626979141089e-05,
+      "loss": 2.0719,
+      "step": 3615
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.25667804479599,
+      "learning_rate": 4.795713487600547e-05,
+      "loss": 2.1776,
+      "step": 3620
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4448702037334442,
+      "learning_rate": 4.795156457658235e-05,
+      "loss": 2.1112,
+      "step": 3625
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5844559669494629,
+      "learning_rate": 4.7945987017601435e-05,
+      "loss": 2.0799,
+      "step": 3630
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5194575786590576,
+      "learning_rate": 4.79404022008269e-05,
+      "loss": 1.9454,
+      "step": 3635
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4168377220630646,
+      "learning_rate": 4.7934810128025216e-05,
+      "loss": 1.5136,
+      "step": 3640
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6346775889396667,
+      "learning_rate": 4.792921080096514e-05,
+      "loss": 2.225,
+      "step": 3645
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2728739976882935,
+      "learning_rate": 4.7923604221417754e-05,
+      "loss": 1.9655,
+      "step": 3650
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6890284419059753,
+      "learning_rate": 4.79179903911564e-05,
+      "loss": 1.9297,
+      "step": 3655
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.341615676879883,
+      "learning_rate": 4.7912369311956726e-05,
+      "loss": 2.0387,
+      "step": 3660
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8217077851295471,
+      "learning_rate": 4.790674098559668e-05,
+      "loss": 1.8114,
+      "step": 3665
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.9442257881164551,
+      "learning_rate": 4.790110541385649e-05,
+      "loss": 2.011,
+      "step": 3670
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0342832803726196,
+      "learning_rate": 4.789546259851869e-05,
+      "loss": 1.9695,
+      "step": 3675
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5166803598403931,
+      "learning_rate": 4.788981254136809e-05,
+      "loss": 1.6962,
+      "step": 3680
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.49156954884529114,
+      "learning_rate": 4.78841552441918e-05,
+      "loss": 1.9125,
+      "step": 3685
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.0156267881393433,
+      "learning_rate": 4.7878490708779225e-05,
+      "loss": 2.0294,
+      "step": 3690
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5214945673942566,
+      "learning_rate": 4.787281893692204e-05,
+      "loss": 2.1405,
+      "step": 3695
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5293641090393066,
+      "learning_rate": 4.7867139930414214e-05,
+      "loss": 1.8728,
+      "step": 3700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5196399688720703,
+      "learning_rate": 4.786145369105204e-05,
+      "loss": 1.8972,
+      "step": 3705
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6026878952980042,
+      "learning_rate": 4.785576022063405e-05,
+      "loss": 1.9923,
+      "step": 3710
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5810438394546509,
+      "learning_rate": 4.785005952096109e-05,
+      "loss": 1.9926,
+      "step": 3715
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6136882305145264,
+      "learning_rate": 4.784435159383627e-05,
+      "loss": 1.9561,
+      "step": 3720
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7523034811019897,
+      "learning_rate": 4.783863644106502e-05,
+      "loss": 1.9247,
+      "step": 3725
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.477094441652298,
+      "learning_rate": 4.783291406445504e-05,
+      "loss": 2.2932,
+      "step": 3730
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7570971250534058,
+      "learning_rate": 4.78271844658163e-05,
+      "loss": 2.1548,
+      "step": 3735
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6923122406005859,
+      "learning_rate": 4.7821447646961066e-05,
+      "loss": 1.7722,
+      "step": 3740
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4171823561191559,
+      "learning_rate": 4.7815703609703896e-05,
+      "loss": 2.1501,
+      "step": 3745
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8493704199790955,
+      "learning_rate": 4.7809952355861614e-05,
+      "loss": 2.1935,
+      "step": 3750
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7510491609573364,
+      "learning_rate": 4.7804193887253344e-05,
+      "loss": 1.7398,
+      "step": 3755
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6981149315834045,
+      "learning_rate": 4.779842820570049e-05,
+      "loss": 1.857,
+      "step": 3760
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6277537941932678,
+      "learning_rate": 4.7792655313026715e-05,
+      "loss": 1.9971,
+      "step": 3765
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.963358461856842,
+      "learning_rate": 4.7786875211058e-05,
+      "loss": 1.7726,
+      "step": 3770
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5438962578773499,
+      "learning_rate": 4.778108790162256e-05,
+      "loss": 1.6827,
+      "step": 3775
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7283356189727783,
+      "learning_rate": 4.777529338655094e-05,
+      "loss": 1.915,
+      "step": 3780
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.1045600175857544,
+      "learning_rate": 4.776949166767592e-05,
+      "loss": 2.3648,
+      "step": 3785
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6527469754219055,
+      "learning_rate": 4.77636827468326e-05,
+      "loss": 1.8445,
+      "step": 3790
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3629080057144165,
+      "learning_rate": 4.775786662585831e-05,
+      "loss": 1.949,
+      "step": 3795
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.49108001589775085,
+      "learning_rate": 4.77520433065927e-05,
+      "loss": 1.8602,
+      "step": 3800
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3771425783634186,
+      "learning_rate": 4.774621279087768e-05,
+      "loss": 1.8913,
+      "step": 3805
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.9145960807800293,
+      "learning_rate": 4.7740375080557434e-05,
+      "loss": 1.9314,
+      "step": 3810
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6728583574295044,
+      "learning_rate": 4.773453017747842e-05,
+      "loss": 1.6274,
+      "step": 3815
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5668622255325317,
+      "learning_rate": 4.772867808348938e-05,
+      "loss": 2.0638,
+      "step": 3820
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6772801876068115,
+      "learning_rate": 4.7722818800441316e-05,
+      "loss": 1.9404,
+      "step": 3825
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.99435043334961,
+      "learning_rate": 4.771695233018754e-05,
+      "loss": 1.8604,
+      "step": 3830
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8190693259239197,
+      "learning_rate": 4.7711078674583576e-05,
+      "loss": 1.8342,
+      "step": 3835
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4974558353424072,
+      "learning_rate": 4.770519783548728e-05,
+      "loss": 2.1466,
+      "step": 3840
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8280129432678223,
+      "learning_rate": 4.7699309814758746e-05,
+      "loss": 1.845,
+      "step": 3845
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6876835823059082,
+      "learning_rate": 4.7693414614260356e-05,
+      "loss": 2.0733,
+      "step": 3850
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2137017250061035,
+      "learning_rate": 4.7687512235856744e-05,
+      "loss": 2.1651,
+      "step": 3855
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7566338181495667,
+      "learning_rate": 4.768160268141484e-05,
+      "loss": 1.515,
+      "step": 3860
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.49978703260421753,
+      "learning_rate": 4.767568595280383e-05,
+      "loss": 1.8929,
+      "step": 3865
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3817705512046814,
+      "learning_rate": 4.766976205189516e-05,
+      "loss": 1.8465,
+      "step": 3870
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4288356900215149,
+      "learning_rate": 4.766383098056255e-05,
+      "loss": 1.8791,
+      "step": 3875
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5867508053779602,
+      "learning_rate": 4.7657892740682006e-05,
+      "loss": 1.9564,
+      "step": 3880
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2050591707229614,
+      "learning_rate": 4.765194733413178e-05,
+      "loss": 1.9315,
+      "step": 3885
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.9066573977470398,
+      "learning_rate": 4.764599476279241e-05,
+      "loss": 1.9297,
+      "step": 3890
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.0444132089614868,
+      "learning_rate": 4.764003502854668e-05,
+      "loss": 1.7961,
+      "step": 3895
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.708714485168457,
+      "learning_rate": 4.763406813327964e-05,
+      "loss": 1.9499,
+      "step": 3900
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7246438264846802,
+      "learning_rate": 4.762809407887862e-05,
+      "loss": 1.8408,
+      "step": 3905
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7492526769638062,
+      "learning_rate": 4.762211286723321e-05,
+      "loss": 2.181,
+      "step": 3910
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6265992522239685,
+      "learning_rate": 4.761612450023526e-05,
+      "loss": 2.1916,
+      "step": 3915
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4639782905578613,
+      "learning_rate": 4.7610128979778883e-05,
+      "loss": 2.1112,
+      "step": 3920
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7836752533912659,
+      "learning_rate": 4.760412630776046e-05,
+      "loss": 1.8764,
+      "step": 3925
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8531836867332458,
+      "learning_rate": 4.759811648607862e-05,
+      "loss": 2.0257,
+      "step": 3930
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.540393054485321,
+      "learning_rate": 4.7592099516634294e-05,
+      "loss": 1.6083,
+      "step": 3935
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5300696492195129,
+      "learning_rate": 4.75860754013306e-05,
+      "loss": 1.8257,
+      "step": 3940
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6871602535247803,
+      "learning_rate": 4.758004414207299e-05,
+      "loss": 2.0006,
+      "step": 3945
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5659440159797668,
+      "learning_rate": 4.7574005740769136e-05,
+      "loss": 1.6773,
+      "step": 3950
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.466205358505249,
+      "learning_rate": 4.7567960199328984e-05,
+      "loss": 2.129,
+      "step": 3955
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5387719869613647,
+      "learning_rate": 4.7561907519664724e-05,
+      "loss": 1.8874,
+      "step": 3960
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7125961184501648,
+      "learning_rate": 4.755584770369081e-05,
+      "loss": 1.919,
+      "step": 3965
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.9576629400253296,
+      "learning_rate": 4.754978075332398e-05,
+      "loss": 1.9866,
+      "step": 3970
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6893014311790466,
+      "learning_rate": 4.754370667048317e-05,
+      "loss": 1.6943,
+      "step": 3975
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0120683908462524,
+      "learning_rate": 4.753762545708964e-05,
+      "loss": 1.7324,
+      "step": 3980
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.0890415906906128,
+      "learning_rate": 4.753153711506685e-05,
+      "loss": 1.8863,
+      "step": 3985
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5287376642227173,
+      "learning_rate": 4.752544164634054e-05,
+      "loss": 2.111,
+      "step": 3990
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.014571189880371,
+      "learning_rate": 4.7519339052838706e-05,
+      "loss": 1.9179,
+      "step": 3995
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6377423405647278,
+      "learning_rate": 4.7513229336491594e-05,
+      "loss": 1.736,
+      "step": 4000
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7745433449745178,
+      "learning_rate": 4.7507112499231696e-05,
+      "loss": 2.0101,
+      "step": 4005
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7528572678565979,
+      "learning_rate": 4.750098854299376e-05,
+      "loss": 1.9943,
+      "step": 4010
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8742436170578003,
+      "learning_rate": 4.7494857469714796e-05,
+      "loss": 1.9835,
+      "step": 4015
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.895379662513733,
+      "learning_rate": 4.748871928133405e-05,
+      "loss": 1.7243,
+      "step": 4020
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6951816082000732,
+      "learning_rate": 4.748257397979302e-05,
+      "loss": 1.7516,
+      "step": 4025
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9127728939056396,
+      "learning_rate": 4.7476421567035475e-05,
+      "loss": 2.2468,
+      "step": 4030
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.9135558605194092,
+      "learning_rate": 4.747026204500741e-05,
+      "loss": 1.8177,
+      "step": 4035
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3471672236919403,
+      "learning_rate": 4.746409541565706e-05,
+      "loss": 1.8712,
+      "step": 4040
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.621139407157898,
+      "learning_rate": 4.7457921680934946e-05,
+      "loss": 1.9565,
+      "step": 4045
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2294261455535889,
+      "learning_rate": 4.745174084279381e-05,
+      "loss": 2.0262,
+      "step": 4050
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7085707783699036,
+      "learning_rate": 4.744555290318864e-05,
+      "loss": 1.8709,
+      "step": 4055
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5060424208641052,
+      "learning_rate": 4.7439357864076674e-05,
+      "loss": 1.8826,
+      "step": 4060
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5154075622558594,
+      "learning_rate": 4.74331557274174e-05,
+      "loss": 2.1846,
+      "step": 4065
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9464437961578369,
+      "learning_rate": 4.7426946495172545e-05,
+      "loss": 2.0822,
+      "step": 4070
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2713757753372192,
+      "learning_rate": 4.7420730169306084e-05,
+      "loss": 1.9661,
+      "step": 4075
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.036674976348877,
+      "learning_rate": 4.7414506751784236e-05,
+      "loss": 1.8481,
+      "step": 4080
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6142359972000122,
+      "learning_rate": 4.740827624457547e-05,
+      "loss": 1.8596,
+      "step": 4085
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9795764088630676,
+      "learning_rate": 4.740203864965048e-05,
+      "loss": 2.054,
+      "step": 4090
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9507243633270264,
+      "learning_rate": 4.739579396898222e-05,
+      "loss": 1.9746,
+      "step": 4095
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5483179092407227,
+      "learning_rate": 4.738954220454586e-05,
+      "loss": 2.0328,
+      "step": 4100
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.9164968729019165,
+      "learning_rate": 4.738328335831883e-05,
+      "loss": 2.0429,
+      "step": 4105
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.672351062297821,
+      "learning_rate": 4.7377017432280815e-05,
+      "loss": 1.8414,
+      "step": 4110
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5230313539505005,
+      "learning_rate": 4.7370744428413704e-05,
+      "loss": 2.2537,
+      "step": 4115
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.480780690908432,
+      "learning_rate": 4.736446434870166e-05,
+      "loss": 1.8738,
+      "step": 4120
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4368326663970947,
+      "learning_rate": 4.7358177195131035e-05,
+      "loss": 1.8399,
+      "step": 4125
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.0797245502471924,
+      "learning_rate": 4.735188296969048e-05,
+      "loss": 1.8166,
+      "step": 4130
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6688716411590576,
+      "learning_rate": 4.734558167437084e-05,
+      "loss": 1.8741,
+      "step": 4135
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3694940507411957,
+      "learning_rate": 4.733927331116521e-05,
+      "loss": 2.2319,
+      "step": 4140
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4561655521392822,
+      "learning_rate": 4.7332957882068925e-05,
+      "loss": 1.6936,
+      "step": 4145
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8536581993103027,
+      "learning_rate": 4.732663538907954e-05,
+      "loss": 1.6583,
+      "step": 4150
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9004095196723938,
+      "learning_rate": 4.7320305834196856e-05,
+      "loss": 1.9843,
+      "step": 4155
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3630378246307373,
+      "learning_rate": 4.7313969219422906e-05,
+      "loss": 1.6481,
+      "step": 4160
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9355126023292542,
+      "learning_rate": 4.7307625546761945e-05,
+      "loss": 1.9516,
+      "step": 4165
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.281752347946167,
+      "learning_rate": 4.73012748182205e-05,
+      "loss": 1.5777,
+      "step": 4170
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5159614086151123,
+      "learning_rate": 4.729491703580727e-05,
+      "loss": 2.0937,
+      "step": 4175
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5500282049179077,
+      "learning_rate": 4.728855220153322e-05,
+      "loss": 1.8505,
+      "step": 4180
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.47913625836372375,
+      "learning_rate": 4.728218031741156e-05,
+      "loss": 2.1684,
+      "step": 4185
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0718451738357544,
+      "learning_rate": 4.72758013854577e-05,
+      "loss": 1.9026,
+      "step": 4190
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.538801908493042,
+      "learning_rate": 4.7269415407689276e-05,
+      "loss": 1.8334,
+      "step": 4195
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.0258709192276,
+      "learning_rate": 4.726302238612619e-05,
+      "loss": 1.8611,
+      "step": 4200
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7120585441589355,
+      "learning_rate": 4.725662232279053e-05,
+      "loss": 1.8925,
+      "step": 4205
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7516504526138306,
+      "learning_rate": 4.7250215219706636e-05,
+      "loss": 2.3299,
+      "step": 4210
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.125640630722046,
+      "learning_rate": 4.7243801078901084e-05,
+      "loss": 2.3689,
+      "step": 4215
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7650467157363892,
+      "learning_rate": 4.7237379902402636e-05,
+      "loss": 1.7783,
+      "step": 4220
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7223068475723267,
+      "learning_rate": 4.723095169224231e-05,
+      "loss": 2.2361,
+      "step": 4225
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.856893002986908,
+      "learning_rate": 4.722451645045336e-05,
+      "loss": 1.9778,
+      "step": 4230
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5695852041244507,
+      "learning_rate": 4.7218074179071225e-05,
+      "loss": 1.8546,
+      "step": 4235
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7213203310966492,
+      "learning_rate": 4.72116248801336e-05,
+      "loss": 1.9171,
+      "step": 4240
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4641231298446655,
+      "learning_rate": 4.720516855568039e-05,
+      "loss": 2.0134,
+      "step": 4245
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.9336233139038086,
+      "learning_rate": 4.719870520775372e-05,
+      "loss": 1.7665,
+      "step": 4250
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6485480070114136,
+      "learning_rate": 4.719223483839795e-05,
+      "loss": 1.6734,
+      "step": 4255
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.33693981170654297,
+      "learning_rate": 4.718575744965965e-05,
+      "loss": 1.7652,
+      "step": 4260
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7903716564178467,
+      "learning_rate": 4.717927304358761e-05,
+      "loss": 2.0551,
+      "step": 4265
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.736282467842102,
+      "learning_rate": 4.7172781622232834e-05,
+      "loss": 1.9317,
+      "step": 4270
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5668120384216309,
+      "learning_rate": 4.716628318764856e-05,
+      "loss": 1.8764,
+      "step": 4275
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4904122054576874,
+      "learning_rate": 4.715977774189023e-05,
+      "loss": 1.8657,
+      "step": 4280
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4836113154888153,
+      "learning_rate": 4.715326528701553e-05,
+      "loss": 2.1163,
+      "step": 4285
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0537573099136353,
+      "learning_rate": 4.7146745825084315e-05,
+      "loss": 1.8157,
+      "step": 4290
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7334871888160706,
+      "learning_rate": 4.71402193581587e-05,
+      "loss": 1.9083,
+      "step": 4295
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6025909185409546,
+      "learning_rate": 4.7133685888303e-05,
+      "loss": 1.6309,
+      "step": 4300
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5878241658210754,
+      "learning_rate": 4.712714541758374e-05,
+      "loss": 1.7923,
+      "step": 4305
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9317494630813599,
+      "learning_rate": 4.7120597948069676e-05,
+      "loss": 1.6716,
+      "step": 4310
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2589433193206787,
+      "learning_rate": 4.711404348183175e-05,
+      "loss": 1.9239,
+      "step": 4315
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.0758674144744873,
+      "learning_rate": 4.710748202094315e-05,
+      "loss": 1.8766,
+      "step": 4320
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7382737994194031,
+      "learning_rate": 4.710091356747924e-05,
+      "loss": 1.8116,
+      "step": 4325
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.790633499622345,
+      "learning_rate": 4.709433812351764e-05,
+      "loss": 2.1118,
+      "step": 4330
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": Infinity,
+      "learning_rate": 4.708907273658756e-05,
+      "loss": 1.9079,
+      "step": 4335
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3426882028579712,
+      "learning_rate": 4.708248471497269e-05,
+      "loss": 1.958,
+      "step": 4340
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6428537368774414,
+      "learning_rate": 4.707588970868914e-05,
+      "loss": 1.9855,
+      "step": 4345
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4597220718860626,
+      "learning_rate": 4.7069287719822915e-05,
+      "loss": 1.875,
+      "step": 4350
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8507035970687866,
+      "learning_rate": 4.7062678750462205e-05,
+      "loss": 1.7406,
+      "step": 4355
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8498551845550537,
+      "learning_rate": 4.7056062802697435e-05,
+      "loss": 1.8201,
+      "step": 4360
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.037539005279541,
+      "learning_rate": 4.704943987862121e-05,
+      "loss": 2.0407,
+      "step": 4365
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6955700516700745,
+      "learning_rate": 4.704280998032836e-05,
+      "loss": 1.8599,
+      "step": 4370
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.9754367470741272,
+      "learning_rate": 4.703617310991592e-05,
+      "loss": 2.0508,
+      "step": 4375
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6729751825332642,
+      "learning_rate": 4.702952926948312e-05,
+      "loss": 1.8073,
+      "step": 4380
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.47259873151779175,
+      "learning_rate": 4.7022878461131404e-05,
+      "loss": 1.7583,
+      "step": 4385
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3793861865997314,
+      "learning_rate": 4.701622068696441e-05,
+      "loss": 1.8839,
+      "step": 4390
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.0509848594665527,
+      "learning_rate": 4.700955594908799e-05,
+      "loss": 1.8347,
+      "step": 4395
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6066625118255615,
+      "learning_rate": 4.700288424961019e-05,
+      "loss": 1.9021,
+      "step": 4400
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6687719821929932,
+      "learning_rate": 4.6996205590641274e-05,
+      "loss": 1.9797,
+      "step": 4405
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5127503871917725,
+      "learning_rate": 4.6989519974293684e-05,
+      "loss": 1.7509,
+      "step": 4410
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6416090130805969,
+      "learning_rate": 4.698282740268208e-05,
+      "loss": 1.8623,
+      "step": 4415
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7178260087966919,
+      "learning_rate": 4.697612787792331e-05,
+      "loss": 1.8798,
+      "step": 4420
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5021282434463501,
+      "learning_rate": 4.696942140213643e-05,
+      "loss": 2.0592,
+      "step": 4425
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5925451517105103,
+      "learning_rate": 4.696270797744269e-05,
+      "loss": 1.859,
+      "step": 4430
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7658485770225525,
+      "learning_rate": 4.6955987605965555e-05,
+      "loss": 1.9042,
+      "step": 4435
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7695988416671753,
+      "learning_rate": 4.6949260289830644e-05,
+      "loss": 1.482,
+      "step": 4440
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6327041387557983,
+      "learning_rate": 4.694252603116582e-05,
+      "loss": 2.0297,
+      "step": 4445
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7171325087547302,
+      "learning_rate": 4.693578483210113e-05,
+      "loss": 1.9515,
+      "step": 4450
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.562306821346283,
+      "learning_rate": 4.692903669476878e-05,
+      "loss": 1.9341,
+      "step": 4455
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8432119488716125,
+      "learning_rate": 4.692228162130322e-05,
+      "loss": 2.1466,
+      "step": 4460
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.496180921792984,
+      "learning_rate": 4.691551961384108e-05,
+      "loss": 2.0088,
+      "step": 4465
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1345199346542358,
+      "learning_rate": 4.690875067452116e-05,
+      "loss": 2.2793,
+      "step": 4470
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6553413271903992,
+      "learning_rate": 4.690197480548447e-05,
+      "loss": 2.2163,
+      "step": 4475
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8875722289085388,
+      "learning_rate": 4.6895192008874225e-05,
+      "loss": 1.7843,
+      "step": 4480
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.8258360624313354,
+      "learning_rate": 4.6888402286835804e-05,
+      "loss": 1.8453,
+      "step": 4485
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.9482008814811707,
+      "learning_rate": 4.688160564151679e-05,
+      "loss": 1.6878,
+      "step": 4490
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.45167991518974304,
+      "learning_rate": 4.687480207506697e-05,
+      "loss": 2.0793,
+      "step": 4495
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2857486605644226,
+      "learning_rate": 4.6867991589638284e-05,
+      "loss": 2.0167,
+      "step": 4500
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6648562550544739,
+      "learning_rate": 4.686117418738489e-05,
+      "loss": 1.7389,
+      "step": 4505
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9790322780609131,
+      "learning_rate": 4.685434987046314e-05,
+      "loss": 2.0685,
+      "step": 4510
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4371451437473297,
+      "learning_rate": 4.6847518641031544e-05,
+      "loss": 1.9618,
+      "step": 4515
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7518529891967773,
+      "learning_rate": 4.6840680501250814e-05,
+      "loss": 1.7586,
+      "step": 4520
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6136543154716492,
+      "learning_rate": 4.683383545328385e-05,
+      "loss": 1.9332,
+      "step": 4525
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8958885073661804,
+      "learning_rate": 4.682698349929573e-05,
+      "loss": 1.6817,
+      "step": 4530
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6221014857292175,
+      "learning_rate": 4.682012464145373e-05,
+      "loss": 2.2389,
+      "step": 4535
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.043770432472229,
+      "learning_rate": 4.681325888192729e-05,
+      "loss": 1.7041,
+      "step": 4540
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.9577236175537109,
+      "learning_rate": 4.680638622288804e-05,
+      "loss": 1.9666,
+      "step": 4545
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5839371681213379,
+      "learning_rate": 4.679950666650982e-05,
+      "loss": 1.9035,
+      "step": 4550
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.314014196395874,
+      "learning_rate": 4.67926202149686e-05,
+      "loss": 1.9786,
+      "step": 4555
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.44664058089256287,
+      "learning_rate": 4.678572687044257e-05,
+      "loss": 1.9205,
+      "step": 4560
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5979194641113281,
+      "learning_rate": 4.6778826635112085e-05,
+      "loss": 1.8723,
+      "step": 4565
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8216999173164368,
+      "learning_rate": 4.677191951115968e-05,
+      "loss": 1.9851,
+      "step": 4570
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9313204288482666,
+      "learning_rate": 4.676638885365804e-05,
+      "loss": 1.9628,
+      "step": 4575
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9604901075363159,
+      "learning_rate": 4.675946933569314e-05,
+      "loss": 1.6638,
+      "step": 4580
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9496554732322693,
+      "learning_rate": 4.675254293522902e-05,
+      "loss": 1.7364,
+      "step": 4585
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6214568018913269,
+      "learning_rate": 4.674560965445649e-05,
+      "loss": 1.7255,
+      "step": 4590
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6437748670578003,
+      "learning_rate": 4.673866949556854e-05,
+      "loss": 2.1914,
+      "step": 4595
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4943290948867798,
+      "learning_rate": 4.6731722460760355e-05,
+      "loss": 1.9668,
+      "step": 4600
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.41138797998428345,
+      "learning_rate": 4.672476855222924e-05,
+      "loss": 1.8255,
+      "step": 4605
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6955471634864807,
+      "learning_rate": 4.671780777217474e-05,
+      "loss": 1.9333,
+      "step": 4610
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1246094703674316,
+      "learning_rate": 4.671084012279853e-05,
+      "loss": 2.0143,
+      "step": 4615
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7304584980010986,
+      "learning_rate": 4.6703865606304465e-05,
+      "loss": 1.9142,
+      "step": 4620
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.188251256942749,
+      "learning_rate": 4.6696884224898584e-05,
+      "loss": 1.8847,
+      "step": 4625
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9002482295036316,
+      "learning_rate": 4.66898959807891e-05,
+      "loss": 2.1051,
+      "step": 4630
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9018658995628357,
+      "learning_rate": 4.668290087618638e-05,
+      "loss": 1.8925,
+      "step": 4635
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6331837177276611,
+      "learning_rate": 4.667589891330297e-05,
+      "loss": 2.0296,
+      "step": 4640
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5853480696678162,
+      "learning_rate": 4.666889009435358e-05,
+      "loss": 1.9348,
+      "step": 4645
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.261925458908081,
+      "learning_rate": 4.6661874421555094e-05,
+      "loss": 2.1714,
+      "step": 4650
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5459082126617432,
+      "learning_rate": 4.665485189712657e-05,
+      "loss": 1.7488,
+      "step": 4655
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5560827851295471,
+      "learning_rate": 4.664782252328922e-05,
+      "loss": 1.8829,
+      "step": 4660
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 12.994915008544922,
+      "learning_rate": 4.664078630226643e-05,
+      "loss": 2.2486,
+      "step": 4665
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5184732675552368,
+      "learning_rate": 4.663374323628374e-05,
+      "loss": 1.5356,
+      "step": 4670
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5672849416732788,
+      "learning_rate": 4.662669332756887e-05,
+      "loss": 1.6331,
+      "step": 4675
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5396742224693298,
+      "learning_rate": 4.661963657835172e-05,
+      "loss": 2.1822,
+      "step": 4680
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.557628333568573,
+      "learning_rate": 4.6612572990864303e-05,
+      "loss": 1.9965,
+      "step": 4685
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8263502717018127,
+      "learning_rate": 4.6605502567340844e-05,
+      "loss": 2.1538,
+      "step": 4690
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.053760051727295,
+      "learning_rate": 4.6598425310017704e-05,
+      "loss": 2.0905,
+      "step": 4695
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3697140216827393,
+      "learning_rate": 4.659134122113341e-05,
+      "loss": 1.7339,
+      "step": 4700
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7235055565834045,
+      "learning_rate": 4.658425030292866e-05,
+      "loss": 1.8672,
+      "step": 4705
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7260095477104187,
+      "learning_rate": 4.657715255764629e-05,
+      "loss": 1.7353,
+      "step": 4710
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5732203125953674,
+      "learning_rate": 4.657004798753133e-05,
+      "loss": 2.0318,
+      "step": 4715
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.018949508666992,
+      "learning_rate": 4.656293659483093e-05,
+      "loss": 2.1095,
+      "step": 4720
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6857671737670898,
+      "learning_rate": 4.655581838179444e-05,
+      "loss": 1.8601,
+      "step": 4725
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5136845707893372,
+      "learning_rate": 4.6548693350673325e-05,
+      "loss": 1.9605,
+      "step": 4730
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.44918909668922424,
+      "learning_rate": 4.654156150372123e-05,
+      "loss": 1.9067,
+      "step": 4735
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4144088327884674,
+      "learning_rate": 4.653442284319395e-05,
+      "loss": 2.0102,
+      "step": 4740
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.849292516708374,
+      "learning_rate": 4.652727737134944e-05,
+      "loss": 2.1593,
+      "step": 4745
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5621273517608643,
+      "learning_rate": 4.65201250904478e-05,
+      "loss": 2.0483,
+      "step": 4750
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3969671726226807,
+      "learning_rate": 4.65129660027513e-05,
+      "loss": 1.9204,
+      "step": 4755
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.598132312297821,
+      "learning_rate": 4.6505800110524347e-05,
+      "loss": 2.0326,
+      "step": 4760
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7579675316810608,
+      "learning_rate": 4.649862741603349e-05,
+      "loss": 1.7728,
+      "step": 4765
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7204775810241699,
+      "learning_rate": 4.649144792154747e-05,
+      "loss": 1.9556,
+      "step": 4770
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0572607517242432,
+      "learning_rate": 4.648426162933716e-05,
+      "loss": 2.1159,
+      "step": 4775
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5266518592834473,
+      "learning_rate": 4.647706854167554e-05,
+      "loss": 1.9842,
+      "step": 4780
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8113346695899963,
+      "learning_rate": 4.6469868660837805e-05,
+      "loss": 1.7724,
+      "step": 4785
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8336856961250305,
+      "learning_rate": 4.646266198910126e-05,
+      "loss": 1.8038,
+      "step": 4790
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7910550832748413,
+      "learning_rate": 4.645544852874538e-05,
+      "loss": 1.8405,
+      "step": 4795
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7810003161430359,
+      "learning_rate": 4.644822828205176e-05,
+      "loss": 1.825,
+      "step": 4800
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8987298607826233,
+      "learning_rate": 4.644100125130418e-05,
+      "loss": 1.9643,
+      "step": 4805
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5562641620635986,
+      "learning_rate": 4.643376743878852e-05,
+      "loss": 1.7666,
+      "step": 4810
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8323341012001038,
+      "learning_rate": 4.642652684679283e-05,
+      "loss": 1.6686,
+      "step": 4815
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.0062700510025024,
+      "learning_rate": 4.6419279477607314e-05,
+      "loss": 1.9944,
+      "step": 4820
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6746718883514404,
+      "learning_rate": 4.64120253335243e-05,
+      "loss": 1.91,
+      "step": 4825
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5174027681350708,
+      "learning_rate": 4.6404764416838264e-05,
+      "loss": 1.8239,
+      "step": 4830
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3364204466342926,
+      "learning_rate": 4.6397496729845844e-05,
+      "loss": 1.7488,
+      "step": 4835
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2048060894012451,
+      "learning_rate": 4.639022227484578e-05,
+      "loss": 1.7477,
+      "step": 4840
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4208517372608185,
+      "learning_rate": 4.638294105413898e-05,
+      "loss": 1.9285,
+      "step": 4845
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.080003261566162,
+      "learning_rate": 4.6375653070028505e-05,
+      "loss": 1.8835,
+      "step": 4850
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.1424204111099243,
+      "learning_rate": 4.6368358324819524e-05,
+      "loss": 1.6948,
+      "step": 4855
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.866698682308197,
+      "learning_rate": 4.636105682081935e-05,
+      "loss": 1.8874,
+      "step": 4860
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.8074659705162048,
+      "learning_rate": 4.6353748560337456e-05,
+      "loss": 1.7727,
+      "step": 4865
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.238422393798828,
+      "learning_rate": 4.634643354568543e-05,
+      "loss": 1.9305,
+      "step": 4870
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5514354109764099,
+      "learning_rate": 4.633911177917701e-05,
+      "loss": 1.9675,
+      "step": 4875
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.41248565912246704,
+      "learning_rate": 4.633178326312806e-05,
+      "loss": 1.7863,
+      "step": 4880
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5374939441680908,
+      "learning_rate": 4.632444799985657e-05,
+      "loss": 1.8905,
+      "step": 4885
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9222282767295837,
+      "learning_rate": 4.63171059916827e-05,
+      "loss": 1.9385,
+      "step": 4890
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0097501277923584,
+      "learning_rate": 4.630975724092871e-05,
+      "loss": 1.772,
+      "step": 4895
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.815949559211731,
+      "learning_rate": 4.6302401749919e-05,
+      "loss": 1.9147,
+      "step": 4900
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4607148766517639,
+      "learning_rate": 4.629503952098011e-05,
+      "loss": 1.8251,
+      "step": 4905
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6612110137939453,
+      "learning_rate": 4.62876705564407e-05,
+      "loss": 2.0799,
+      "step": 4910
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.9041725993156433,
+      "learning_rate": 4.628029485863157e-05,
+      "loss": 1.8727,
+      "step": 4915
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5865108370780945,
+      "learning_rate": 4.627291242988564e-05,
+      "loss": 2.0539,
+      "step": 4920
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.646685779094696,
+      "learning_rate": 4.6265523272537976e-05,
+      "loss": 2.1545,
+      "step": 4925
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8874340057373047,
+      "learning_rate": 4.625812738892575e-05,
+      "loss": 2.0838,
+      "step": 4930
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.48262983560562134,
+      "learning_rate": 4.625072478138828e-05,
+      "loss": 1.7378,
+      "step": 4935
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4273892641067505,
+      "learning_rate": 4.6243315452267014e-05,
+      "loss": 1.6208,
+      "step": 4940
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.860099732875824,
+      "learning_rate": 4.623589940390549e-05,
+      "loss": 1.5971,
+      "step": 4945
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6735519170761108,
+      "learning_rate": 4.622847663864941e-05,
+      "loss": 1.8806,
+      "step": 4950
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.47200706601142883,
+      "learning_rate": 4.6221047158846594e-05,
+      "loss": 1.8463,
+      "step": 4955
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0925625562667847,
+      "learning_rate": 4.621361096684698e-05,
+      "loss": 1.9257,
+      "step": 4960
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.9255615472793579,
+      "learning_rate": 4.6206168065002613e-05,
+      "loss": 2.2559,
+      "step": 4965
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8095393776893616,
+      "learning_rate": 4.619871845566769e-05,
+      "loss": 1.8567,
+      "step": 4970
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8361244201660156,
+      "learning_rate": 4.619126214119851e-05,
+      "loss": 1.4803,
+      "step": 4975
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6798988580703735,
+      "learning_rate": 4.618379912395349e-05,
+      "loss": 2.0151,
+      "step": 4980
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3742920756340027,
+      "learning_rate": 4.617632940629319e-05,
+      "loss": 2.137,
+      "step": 4985
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6546251177787781,
+      "learning_rate": 4.616885299058027e-05,
+      "loss": 2.2428,
+      "step": 4990
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6916753649711609,
+      "learning_rate": 4.6161369879179504e-05,
+      "loss": 2.0984,
+      "step": 4995
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0272783041000366,
+      "learning_rate": 4.61538800744578e-05,
+      "loss": 1.6963,
+      "step": 5000
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4335922598838806,
+      "learning_rate": 4.614638357878418e-05,
+      "loss": 2.0127,
+      "step": 5005
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4880688190460205,
+      "learning_rate": 4.613888039452978e-05,
+      "loss": 1.786,
+      "step": 5010
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.689954400062561,
+      "learning_rate": 4.613137052406784e-05,
+      "loss": 1.6233,
+      "step": 5015
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.206842303276062,
+      "learning_rate": 4.6123853969773726e-05,
+      "loss": 1.8923,
+      "step": 5020
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6934424042701721,
+      "learning_rate": 4.611633073402492e-05,
+      "loss": 1.8961,
+      "step": 5025
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.48485127091407776,
+      "learning_rate": 4.610880081920101e-05,
+      "loss": 1.6494,
+      "step": 5030
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7396721243858337,
+      "learning_rate": 4.610126422768372e-05,
+      "loss": 1.9127,
+      "step": 5035
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.117673873901367,
+      "learning_rate": 4.609372096185683e-05,
+      "loss": 2.0358,
+      "step": 5040
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5403966307640076,
+      "learning_rate": 4.608617102410631e-05,
+      "loss": 1.7811,
+      "step": 5045
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5867857933044434,
+      "learning_rate": 4.6078614416820164e-05,
+      "loss": 1.9641,
+      "step": 5050
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7298117876052856,
+      "learning_rate": 4.6071051142388555e-05,
+      "loss": 1.7933,
+      "step": 5055
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.0257443189620972,
+      "learning_rate": 4.606348120320374e-05,
+      "loss": 1.8427,
+      "step": 5060
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8619400262832642,
+      "learning_rate": 4.6055904601660084e-05,
+      "loss": 1.9523,
+      "step": 5065
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7043263912200928,
+      "learning_rate": 4.6048321340154054e-05,
+      "loss": 1.9647,
+      "step": 5070
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8900740742683411,
+      "learning_rate": 4.604073142108423e-05,
+      "loss": 1.8617,
+      "step": 5075
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5621470808982849,
+      "learning_rate": 4.6033134846851294e-05,
+      "loss": 2.0817,
+      "step": 5080
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5144213438034058,
+      "learning_rate": 4.602553161985804e-05,
+      "loss": 1.9585,
+      "step": 5085
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5859657526016235,
+      "learning_rate": 4.601792174250935e-05,
+      "loss": 1.9843,
+      "step": 5090
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.48220011591911316,
+      "learning_rate": 4.6010305217212244e-05,
+      "loss": 2.3189,
+      "step": 5095
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6994110941886902,
+      "learning_rate": 4.60026820463758e-05,
+      "loss": 1.699,
+      "step": 5100
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.361588478088379,
+      "learning_rate": 4.5995052232411216e-05,
+      "loss": 1.8276,
+      "step": 5105
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.627723753452301,
+      "learning_rate": 4.598741577773182e-05,
+      "loss": 2.045,
+      "step": 5110
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4922670125961304,
+      "learning_rate": 4.5979772684752995e-05,
+      "loss": 1.7747,
+      "step": 5115
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6467026472091675,
+      "learning_rate": 4.597212295589225e-05,
+      "loss": 1.7939,
+      "step": 5120
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6912665367126465,
+      "learning_rate": 4.596446659356919e-05,
+      "loss": 1.8892,
+      "step": 5125
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.146193265914917,
+      "learning_rate": 4.595680360020551e-05,
+      "loss": 1.7835,
+      "step": 5130
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.549444317817688,
+      "learning_rate": 4.594913397822501e-05,
+      "loss": 1.8231,
+      "step": 5135
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6164882779121399,
+      "learning_rate": 4.594145773005358e-05,
+      "loss": 1.8279,
+      "step": 5140
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3416946232318878,
+      "learning_rate": 4.593377485811922e-05,
+      "loss": 2.243,
+      "step": 5145
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5608302354812622,
+      "learning_rate": 4.5926085364852e-05,
+      "loss": 1.7795,
+      "step": 5150
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5376662015914917,
+      "learning_rate": 4.591838925268411e-05,
+      "loss": 1.7712,
+      "step": 5155
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6536192893981934,
+      "learning_rate": 4.591068652404982e-05,
+      "loss": 2.2747,
+      "step": 5160
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.40830197930336,
+      "learning_rate": 4.59029771813855e-05,
+      "loss": 1.9315,
+      "step": 5165
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8668363094329834,
+      "learning_rate": 4.58952612271296e-05,
+      "loss": 2.0715,
+      "step": 5170
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6481930613517761,
+      "learning_rate": 4.588753866372267e-05,
+      "loss": 1.8249,
+      "step": 5175
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3583790063858032,
+      "learning_rate": 4.5879809493607364e-05,
+      "loss": 1.8025,
+      "step": 5180
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5957528352737427,
+      "learning_rate": 4.587207371922838e-05,
+      "loss": 2.1159,
+      "step": 5185
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6677088141441345,
+      "learning_rate": 4.586433134303257e-05,
+      "loss": 1.9156,
+      "step": 5190
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8160932660102844,
+      "learning_rate": 4.585658236746881e-05,
+      "loss": 2.0241,
+      "step": 5195
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6950777769088745,
+      "learning_rate": 4.584882679498813e-05,
+      "loss": 1.9214,
+      "step": 5200
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.726952314376831,
+      "learning_rate": 4.5841064628043575e-05,
+      "loss": 1.7244,
+      "step": 5205
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7783024907112122,
+      "learning_rate": 4.583329586909033e-05,
+      "loss": 1.8253,
+      "step": 5210
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8075493574142456,
+      "learning_rate": 4.5825520520585635e-05,
+      "loss": 1.7076,
+      "step": 5215
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0622972249984741,
+      "learning_rate": 4.5817738584988835e-05,
+      "loss": 1.9551,
+      "step": 5220
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6024557948112488,
+      "learning_rate": 4.580995006476134e-05,
+      "loss": 1.8075,
+      "step": 5225
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5445288419723511,
+      "learning_rate": 4.580215496236666e-05,
+      "loss": 1.7501,
+      "step": 5230
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8697842359542847,
+      "learning_rate": 4.5794353280270376e-05,
+      "loss": 1.7579,
+      "step": 5235
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6215696334838867,
+      "learning_rate": 4.5786545020940155e-05,
+      "loss": 1.6428,
+      "step": 5240
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6145104169845581,
+      "learning_rate": 4.577873018684573e-05,
+      "loss": 1.7593,
+      "step": 5245
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7944318056106567,
+      "learning_rate": 4.577090878045893e-05,
+      "loss": 2.1836,
+      "step": 5250
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7711953520774841,
+      "learning_rate": 4.576308080425367e-05,
+      "loss": 1.9731,
+      "step": 5255
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5569404363632202,
+      "learning_rate": 4.575524626070592e-05,
+      "loss": 2.1212,
+      "step": 5260
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6812270283699036,
+      "learning_rate": 4.574740515229374e-05,
+      "loss": 2.0945,
+      "step": 5265
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8699807524681091,
+      "learning_rate": 4.5739557481497275e-05,
+      "loss": 2.0372,
+      "step": 5270
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.8972649574279785,
+      "learning_rate": 4.573170325079872e-05,
+      "loss": 2.3673,
+      "step": 5275
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6763632297515869,
+      "learning_rate": 4.5723842462682375e-05,
+      "loss": 1.907,
+      "step": 5280
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.44302767515182495,
+      "learning_rate": 4.571597511963459e-05,
+      "loss": 1.939,
+      "step": 5285
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5313250422477722,
+      "learning_rate": 4.5708101224143796e-05,
+      "loss": 1.8028,
+      "step": 5290
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4340111315250397,
+      "learning_rate": 4.5700220778700504e-05,
+      "loss": 1.7074,
+      "step": 5295
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5250926613807678,
+      "learning_rate": 4.569233378579729e-05,
+      "loss": 1.956,
+      "step": 5300
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8783259391784668,
+      "learning_rate": 4.568444024792881e-05,
+      "loss": 2.1352,
+      "step": 5305
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6077532172203064,
+      "learning_rate": 4.5676540167591776e-05,
+      "loss": 1.6879,
+      "step": 5310
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.854468584060669,
+      "learning_rate": 4.5668633547284964e-05,
+      "loss": 2.0933,
+      "step": 5315
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.812654972076416,
+      "learning_rate": 4.566072038950925e-05,
+      "loss": 1.6729,
+      "step": 5320
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6669471859931946,
+      "learning_rate": 4.565280069676755e-05,
+      "loss": 2.0241,
+      "step": 5325
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4599258303642273,
+      "learning_rate": 4.5644874471564854e-05,
+      "loss": 1.802,
+      "step": 5330
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0719258785247803,
+      "learning_rate": 4.5636941716408224e-05,
+      "loss": 2.0428,
+      "step": 5335
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8084414601325989,
+      "learning_rate": 4.5629002433806764e-05,
+      "loss": 1.8502,
+      "step": 5340
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.715983510017395,
+      "learning_rate": 4.5621056626271694e-05,
+      "loss": 1.9962,
+      "step": 5345
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.682518720626831,
+      "learning_rate": 4.5613104296316245e-05,
+      "loss": 1.5369,
+      "step": 5350
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.020588994026184,
+      "learning_rate": 4.560514544645573e-05,
+      "loss": 1.8893,
+      "step": 5355
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0426610708236694,
+      "learning_rate": 4.559718007920753e-05,
+      "loss": 1.7818,
+      "step": 5360
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8958699703216553,
+      "learning_rate": 4.5589208197091084e-05,
+      "loss": 1.9423,
+      "step": 5365
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6348686218261719,
+      "learning_rate": 4.558122980262789e-05,
+      "loss": 1.9759,
+      "step": 5370
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.073533058166504,
+      "learning_rate": 4.557324489834151e-05,
+      "loss": 1.9607,
+      "step": 5375
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9768711924552917,
+      "learning_rate": 4.556525348675755e-05,
+      "loss": 1.8728,
+      "step": 5380
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6651968359947205,
+      "learning_rate": 4.55572555704037e-05,
+      "loss": 1.8073,
+      "step": 5385
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7141188979148865,
+      "learning_rate": 4.554925115180968e-05,
+      "loss": 1.7795,
+      "step": 5390
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1313745975494385,
+      "learning_rate": 4.5541240233507285e-05,
+      "loss": 1.8481,
+      "step": 5395
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4851437211036682,
+      "learning_rate": 4.5533222818030376e-05,
+      "loss": 1.9045,
+      "step": 5400
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3471333384513855,
+      "learning_rate": 4.5525198907914826e-05,
+      "loss": 2.006,
+      "step": 5405
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5303329229354858,
+      "learning_rate": 4.5517168505698616e-05,
+      "loss": 2.0183,
+      "step": 5410
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.40987589955329895,
+      "learning_rate": 4.550913161392173e-05,
+      "loss": 2.2088,
+      "step": 5415
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5597338676452637,
+      "learning_rate": 4.5501088235126254e-05,
+      "loss": 1.9469,
+      "step": 5420
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5752057433128357,
+      "learning_rate": 4.549303837185628e-05,
+      "loss": 1.9666,
+      "step": 5425
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7234611511230469,
+      "learning_rate": 4.548498202665798e-05,
+      "loss": 1.6851,
+      "step": 5430
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3815643787384033,
+      "learning_rate": 4.5476919202079574e-05,
+      "loss": 2.0537,
+      "step": 5435
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5907308459281921,
+      "learning_rate": 4.5468849900671324e-05,
+      "loss": 1.8174,
+      "step": 5440
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5893926024436951,
+      "learning_rate": 4.546077412498553e-05,
+      "loss": 1.6955,
+      "step": 5445
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5866884589195251,
+      "learning_rate": 4.545269187757657e-05,
+      "loss": 1.7888,
+      "step": 5450
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.43314459919929504,
+      "learning_rate": 4.5444603161000834e-05,
+      "loss": 1.848,
+      "step": 5455
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9520348906517029,
+      "learning_rate": 4.543650797781679e-05,
+      "loss": 2.004,
+      "step": 5460
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.312831401824951,
+      "learning_rate": 4.542840633058493e-05,
+      "loss": 2.1996,
+      "step": 5465
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.944259524345398,
+      "learning_rate": 4.542029822186779e-05,
+      "loss": 1.7677,
+      "step": 5470
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5071392059326172,
+      "learning_rate": 4.5412183654229965e-05,
+      "loss": 1.67,
+      "step": 5475
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7843353152275085,
+      "learning_rate": 4.5404062630238094e-05,
+      "loss": 2.2598,
+      "step": 5480
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4275512397289276,
+      "learning_rate": 4.5395935152460834e-05,
+      "loss": 1.7225,
+      "step": 5485
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9836548566818237,
+      "learning_rate": 4.53878012234689e-05,
+      "loss": 1.9897,
+      "step": 5490
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.572413206100464,
+      "learning_rate": 4.5379660845835045e-05,
+      "loss": 1.7828,
+      "step": 5495
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5451618432998657,
+      "learning_rate": 4.5371514022134085e-05,
+      "loss": 2.0577,
+      "step": 5500
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5373215675354004,
+      "learning_rate": 4.536336075494282e-05,
+      "loss": 1.6858,
+      "step": 5505
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5195179581642151,
+      "learning_rate": 4.535520104684014e-05,
+      "loss": 1.8985,
+      "step": 5510
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5343700647354126,
+      "learning_rate": 4.534703490040695e-05,
+      "loss": 2.0989,
+      "step": 5515
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7476509213447571,
+      "learning_rate": 4.533886231822619e-05,
+      "loss": 1.7542,
+      "step": 5520
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7178905010223389,
+      "learning_rate": 4.533068330288284e-05,
+      "loss": 2.0403,
+      "step": 5525
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4951605796813965,
+      "learning_rate": 4.532249785696392e-05,
+      "loss": 1.9205,
+      "step": 5530
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.46740642189979553,
+      "learning_rate": 4.531430598305848e-05,
+      "loss": 2.0617,
+      "step": 5535
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.046222448348999,
+      "learning_rate": 4.5306107683757584e-05,
+      "loss": 1.6893,
+      "step": 5540
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4250253736972809,
+      "learning_rate": 4.529790296165437e-05,
+      "loss": 2.0393,
+      "step": 5545
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2679554224014282,
+      "learning_rate": 4.528969181934397e-05,
+      "loss": 1.7908,
+      "step": 5550
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8419818878173828,
+      "learning_rate": 4.5281474259423565e-05,
+      "loss": 1.8416,
+      "step": 5555
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6362389922142029,
+      "learning_rate": 4.527325028449236e-05,
+      "loss": 2.0143,
+      "step": 5560
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8798420429229736,
+      "learning_rate": 4.5265019897151595e-05,
+      "loss": 1.7087,
+      "step": 5565
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9879406690597534,
+      "learning_rate": 4.525678310000452e-05,
+      "loss": 2.0423,
+      "step": 5570
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0143115520477295,
+      "learning_rate": 4.524853989565644e-05,
+      "loss": 1.9527,
+      "step": 5575
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6878682971000671,
+      "learning_rate": 4.524029028671467e-05,
+      "loss": 1.9395,
+      "step": 5580
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6117766499519348,
+      "learning_rate": 4.523203427578855e-05,
+      "loss": 1.7447,
+      "step": 5585
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.42814651131629944,
+      "learning_rate": 4.522377186548944e-05,
+      "loss": 1.8254,
+      "step": 5590
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6543037295341492,
+      "learning_rate": 4.5215503058430754e-05,
+      "loss": 2.0367,
+      "step": 5595
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5894500017166138,
+      "learning_rate": 4.520722785722789e-05,
+      "loss": 1.8911,
+      "step": 5600
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7925021052360535,
+      "learning_rate": 4.519894626449829e-05,
+      "loss": 1.7694,
+      "step": 5605
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6827127933502197,
+      "learning_rate": 4.519065828286142e-05,
+      "loss": 1.7906,
+      "step": 5610
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5993368029594421,
+      "learning_rate": 4.5182363914938754e-05,
+      "loss": 2.142,
+      "step": 5615
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5092318654060364,
+      "learning_rate": 4.517406316335379e-05,
+      "loss": 2.0453,
+      "step": 5620
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5682955980300903,
+      "learning_rate": 4.5165756030732056e-05,
+      "loss": 1.866,
+      "step": 5625
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5514644980430603,
+      "learning_rate": 4.515744251970109e-05,
+      "loss": 1.9906,
+      "step": 5630
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.839066207408905,
+      "learning_rate": 4.514912263289044e-05,
+      "loss": 1.9757,
+      "step": 5635
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.632759153842926,
+      "learning_rate": 4.514079637293168e-05,
+      "loss": 1.8389,
+      "step": 5640
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.5392670631408691,
+      "learning_rate": 4.513246374245841e-05,
+      "loss": 2.0059,
+      "step": 5645
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.7347068190574646,
+      "learning_rate": 4.5124124744106216e-05,
+      "loss": 1.8033,
+      "step": 5650
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.1633368730545044,
+      "learning_rate": 4.511577938051272e-05,
+      "loss": 2.0502,
+      "step": 5655
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.4993457794189453,
+      "learning_rate": 4.510742765431756e-05,
+      "loss": 2.0386,
+      "step": 5660
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.9043493270874023,
+      "learning_rate": 4.509906956816238e-05,
+      "loss": 2.1084,
+      "step": 5665
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 0.4216392934322357,
+      "learning_rate": 4.509070512469081e-05,
+      "loss": 1.8757,
+      "step": 5670
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5536708235740662,
+      "learning_rate": 4.508233432654855e-05,
+      "loss": 1.8147,
+      "step": 5675
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.015234112739563,
+      "learning_rate": 4.507395717638325e-05,
+      "loss": 1.9113,
+      "step": 5680
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.9013848900794983,
+      "learning_rate": 4.506557367684461e-05,
+      "loss": 1.6582,
+      "step": 5685
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.41156238317489624,
+      "learning_rate": 4.505718383058431e-05,
+      "loss": 1.8162,
+      "step": 5690
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8023267984390259,
+      "learning_rate": 4.504878764025606e-05,
+      "loss": 1.9033,
+      "step": 5695
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.4735568165779114,
+      "learning_rate": 4.504038510851557e-05,
+      "loss": 1.604,
+      "step": 5700
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.373860478401184,
+      "learning_rate": 4.503197623802054e-05,
+      "loss": 1.9613,
+      "step": 5705
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.47717657685279846,
+      "learning_rate": 4.5023561031430696e-05,
+      "loss": 1.8763,
+      "step": 5710
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.9874609708786011,
+      "learning_rate": 4.501513949140775e-05,
+      "loss": 1.7237,
+      "step": 5715
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.4677652418613434,
+      "learning_rate": 4.500671162061545e-05,
+      "loss": 2.2116,
+      "step": 5720
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.7714501023292542,
+      "learning_rate": 4.4998277421719506e-05,
+      "loss": 1.9144,
+      "step": 5725
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6441267132759094,
+      "learning_rate": 4.498983689738764e-05,
+      "loss": 1.7153,
+      "step": 5730
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.421938419342041,
+      "learning_rate": 4.4981390050289595e-05,
+      "loss": 1.882,
+      "step": 5735
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5032657980918884,
+      "learning_rate": 4.497293688309709e-05,
+      "loss": 1.9654,
+      "step": 5740
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5165999531745911,
+      "learning_rate": 4.4964477398483875e-05,
+      "loss": 1.7969,
+      "step": 5745
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.9800401329994202,
+      "learning_rate": 4.495601159912566e-05,
+      "loss": 1.7086,
+      "step": 5750
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.6153430342674255,
+      "learning_rate": 4.494753948770016e-05,
+      "loss": 1.641,
+      "step": 5755
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.5942739844322205,
+      "learning_rate": 4.493906106688712e-05,
+      "loss": 1.9742,
+      "step": 5760
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.4884500205516815,
+      "learning_rate": 4.493057633936824e-05,
+      "loss": 2.0754,
+      "step": 5765
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.8295217156410217,
+      "learning_rate": 4.492208530782722e-05,
+      "loss": 1.7546,
+      "step": 5770
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7193713784217834,
+      "learning_rate": 4.49135879749498e-05,
+      "loss": 1.9467,
+      "step": 5775
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 0.7422948479652405,
+      "learning_rate": 4.490508434342364e-05,
+      "loss": 1.8656,
+      "step": 5780
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.331986665725708,
+      "learning_rate": 4.4896574415938465e-05,
+      "loss": 1.7584,
+      "step": 5785
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6991448402404785,
+      "learning_rate": 4.4888058195185935e-05,
+      "loss": 1.6829,
+      "step": 5790
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7543762922286987,
+      "learning_rate": 4.487953568385972e-05,
+      "loss": 1.7909,
+      "step": 5795
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5476166605949402,
+      "learning_rate": 4.4871006884655495e-05,
+      "loss": 2.1132,
+      "step": 5800
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.785237729549408,
+      "learning_rate": 4.486247180027091e-05,
+      "loss": 1.6932,
+      "step": 5805
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8334198594093323,
+      "learning_rate": 4.485393043340559e-05,
+      "loss": 1.9682,
+      "step": 5810
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.4550981521606445,
+      "learning_rate": 4.484538278676118e-05,
+      "loss": 1.4896,
+      "step": 5815
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.184747576713562,
+      "learning_rate": 4.483682886304129e-05,
+      "loss": 1.6799,
+      "step": 5820
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.6847954392433167,
+      "learning_rate": 4.48282686649515e-05,
+      "loss": 1.8183,
+      "step": 5825
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5142964124679565,
+      "learning_rate": 4.4819702195199406e-05,
+      "loss": 2.0018,
+      "step": 5830
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.4494707584381104,
+      "learning_rate": 4.481112945649457e-05,
+      "loss": 1.7658,
+      "step": 5835
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.7955223321914673,
+      "learning_rate": 4.480255045154854e-05,
+      "loss": 1.7644,
+      "step": 5840
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.46225401759147644,
+      "learning_rate": 4.479396518307486e-05,
+      "loss": 1.846,
+      "step": 5845
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8068023920059204,
+      "learning_rate": 4.478537365378902e-05,
+      "loss": 1.9631,
+      "step": 5850
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.9534473419189453,
+      "learning_rate": 4.477677586640854e-05,
+      "loss": 1.9243,
+      "step": 5855
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.4949363470077515,
+      "learning_rate": 4.476817182365286e-05,
+      "loss": 1.7548,
+      "step": 5860
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.6866559386253357,
+      "learning_rate": 4.4759561528243455e-05,
+      "loss": 1.7041,
+      "step": 5865
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.4342908561229706,
+      "learning_rate": 4.475094498290374e-05,
+      "loss": 1.9776,
+      "step": 5870
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.4420872628688812,
+      "learning_rate": 4.474232219035913e-05,
+      "loss": 2.1547,
+      "step": 5875
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.4919748902320862,
+      "learning_rate": 4.4733693153336995e-05,
+      "loss": 1.8431,
+      "step": 5880
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8170517683029175,
+      "learning_rate": 4.472505787456671e-05,
+      "loss": 1.7578,
+      "step": 5885
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 0.8107596039772034,
+      "learning_rate": 4.4716416356779576e-05,
+      "loss": 1.8205,
+      "step": 5890
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5859491229057312,
+      "learning_rate": 4.470776860270892e-05,
+      "loss": 1.5296,
+      "step": 5895
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.41586703062057495,
+      "learning_rate": 4.469911461509002e-05,
+      "loss": 2.2296,
+      "step": 5900
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.7300328016281128,
+      "learning_rate": 4.4690454396660104e-05,
+      "loss": 1.9785,
+      "step": 5905
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.0052844285964966,
+      "learning_rate": 4.46817879501584e-05,
+      "loss": 1.8619,
+      "step": 5910
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6375717520713806,
+      "learning_rate": 4.467311527832611e-05,
+      "loss": 1.8155,
+      "step": 5915
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5199123620986938,
+      "learning_rate": 4.466443638390637e-05,
+      "loss": 1.8485,
+      "step": 5920
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.6076669692993164,
+      "learning_rate": 4.465575126964433e-05,
+      "loss": 1.644,
+      "step": 5925
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.4569863975048065,
+      "learning_rate": 4.464705993828706e-05,
+      "loss": 1.9771,
+      "step": 5930
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.5174528360366821,
+      "learning_rate": 4.4638362392583636e-05,
+      "loss": 1.8152,
+      "step": 5935
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.4930440187454224,
+      "learning_rate": 4.462965863528507e-05,
+      "loss": 1.7984,
+      "step": 5940
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.42374441027641296,
+      "learning_rate": 4.4620948669144356e-05,
+      "loss": 1.7553,
+      "step": 5945
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7552632093429565,
+      "learning_rate": 4.461223249691645e-05,
+      "loss": 2.0744,
+      "step": 5950
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7354189157485962,
+      "learning_rate": 4.460351012135828e-05,
+      "loss": 1.9068,
+      "step": 5955
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.44200262427330017,
+      "learning_rate": 4.4594781545228715e-05,
+      "loss": 1.8694,
+      "step": 5960
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.45945167541503906,
+      "learning_rate": 4.4586046771288593e-05,
+      "loss": 2.0557,
+      "step": 5965
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.506199061870575,
+      "learning_rate": 4.4577305802300715e-05,
+      "loss": 1.8714,
+      "step": 5970
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6749559640884399,
+      "learning_rate": 4.456855864102984e-05,
+      "loss": 2.0554,
+      "step": 5975
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.7233107686042786,
+      "learning_rate": 4.455980529024269e-05,
+      "loss": 1.9416,
+      "step": 5980
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6712331771850586,
+      "learning_rate": 4.4551045752707945e-05,
+      "loss": 1.9445,
+      "step": 5985
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.0653775930404663,
+      "learning_rate": 4.4542280031196224e-05,
+      "loss": 1.9905,
+      "step": 5990
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.2245427370071411,
+      "learning_rate": 4.453350812848014e-05,
+      "loss": 1.5629,
+      "step": 5995
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 0.6561611890792847,
+      "learning_rate": 4.452473004733422e-05,
+      "loss": 2.0267,
+      "step": 6000
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.8881734609603882,
+      "learning_rate": 4.4515945790534964e-05,
+      "loss": 1.7464,
+      "step": 6005
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6296302080154419,
+      "learning_rate": 4.450715536086082e-05,
+      "loss": 1.8395,
+      "step": 6010
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5000322461128235,
+      "learning_rate": 4.4498358761092204e-05,
+      "loss": 1.8501,
+      "step": 6015
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5313418507575989,
+      "learning_rate": 4.448955599401147e-05,
+      "loss": 1.879,
+      "step": 6020
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8610731363296509,
+      "learning_rate": 4.448074706240292e-05,
+      "loss": 1.8825,
+      "step": 6025
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.30168166756629944,
+      "learning_rate": 4.4471931969052816e-05,
+      "loss": 1.7678,
+      "step": 6030
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8175820708274841,
+      "learning_rate": 4.446311071674936e-05,
+      "loss": 1.8989,
+      "step": 6035
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7665614485740662,
+      "learning_rate": 4.4454283308282695e-05,
+      "loss": 1.8553,
+      "step": 6040
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.42481526732444763,
+      "learning_rate": 4.444544974644493e-05,
+      "loss": 2.1974,
+      "step": 6045
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.5651946067810059,
+      "learning_rate": 4.443661003403013e-05,
+      "loss": 2.1612,
+      "step": 6050
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.8740639686584473,
+      "learning_rate": 4.4427764173834264e-05,
+      "loss": 1.6421,
+      "step": 6055
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.6985417604446411,
+      "learning_rate": 4.441891216865528e-05,
+      "loss": 2.0725,
+      "step": 6060
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.8365675210952759,
+      "learning_rate": 4.441005402129306e-05,
+      "loss": 1.9568,
+      "step": 6065
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5997393727302551,
+      "learning_rate": 4.4401189734549405e-05,
+      "loss": 2.0727,
+      "step": 6070
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.2542804479599,
+      "learning_rate": 4.439231931122811e-05,
+      "loss": 2.039,
+      "step": 6075
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.8525872230529785,
+      "learning_rate": 4.4383442754134874e-05,
+      "loss": 1.6291,
+      "step": 6080
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5538986325263977,
+      "learning_rate": 4.4374560066077336e-05,
+      "loss": 2.0365,
+      "step": 6085
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.48354509472846985,
+      "learning_rate": 4.436567124986508e-05,
+      "loss": 2.0723,
+      "step": 6090
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.0979423522949219,
+      "learning_rate": 4.435677630830964e-05,
+      "loss": 1.984,
+      "step": 6095
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6961491107940674,
+      "learning_rate": 4.434787524422448e-05,
+      "loss": 1.7996,
+      "step": 6100
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.0703699588775635,
+      "learning_rate": 4.433896806042498e-05,
+      "loss": 1.904,
+      "step": 6105
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.8863664269447327,
+      "learning_rate": 4.433005475972849e-05,
+      "loss": 1.7778,
+      "step": 6110
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.995624303817749,
+      "learning_rate": 4.432113534495427e-05,
+      "loss": 2.0512,
+      "step": 6115
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.5747253894805908,
+      "learning_rate": 4.4312209818923526e-05,
+      "loss": 1.7252,
+      "step": 6120
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.899549663066864,
+      "learning_rate": 4.430327818445939e-05,
+      "loss": 1.8408,
+      "step": 6125
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6088781356811523,
+      "learning_rate": 4.429434044438693e-05,
+      "loss": 1.7636,
+      "step": 6130
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.7908390164375305,
+      "learning_rate": 4.428539660153315e-05,
+      "loss": 1.7863,
+      "step": 6135
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.4704549312591553,
+      "learning_rate": 4.427644665872697e-05,
+      "loss": 1.8403,
+      "step": 6140
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.4836312532424927,
+      "learning_rate": 4.426749061879926e-05,
+      "loss": 1.8234,
+      "step": 6145
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6054650545120239,
+      "learning_rate": 4.425852848458279e-05,
+      "loss": 1.9921,
+      "step": 6150
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.598057210445404,
+      "learning_rate": 4.42495602589123e-05,
+      "loss": 1.9285,
+      "step": 6155
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.6998239755630493,
+      "learning_rate": 4.42405859446244e-05,
+      "loss": 1.7761,
+      "step": 6160
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.4462050199508667,
+      "learning_rate": 4.4231605544557684e-05,
+      "loss": 1.7563,
+      "step": 6165
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.1818965673446655,
+      "learning_rate": 4.422261906155263e-05,
+      "loss": 2.1757,
+      "step": 6170
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.3994632959365845,
+      "learning_rate": 4.421362649845165e-05,
+      "loss": 1.8819,
+      "step": 6175
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.699133038520813,
+      "learning_rate": 4.42046278580991e-05,
+      "loss": 1.9893,
+      "step": 6180
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.4134710431098938,
+      "learning_rate": 4.419562314334123e-05,
+      "loss": 2.0212,
+      "step": 6185
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.8407548666000366,
+      "learning_rate": 4.4186612357026225e-05,
+      "loss": 1.8084,
+      "step": 6190
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.4963303506374359,
+      "learning_rate": 4.4177595502004175e-05,
+      "loss": 1.9406,
+      "step": 6195
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6621739864349365,
+      "learning_rate": 4.416857258112712e-05,
+      "loss": 1.9083,
+      "step": 6200
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6170079708099365,
+      "learning_rate": 4.4159543597248996e-05,
+      "loss": 1.9234,
+      "step": 6205
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6613223552703857,
+      "learning_rate": 4.415050855322567e-05,
+      "loss": 2.0523,
+      "step": 6210
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.43656325340271,
+      "learning_rate": 4.41414674519149e-05,
+      "loss": 1.9491,
+      "step": 6215
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.7505949139595032,
+      "learning_rate": 4.4132420296176386e-05,
+      "loss": 1.9439,
+      "step": 6220
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.665459394454956,
+      "learning_rate": 4.4123367088871735e-05,
+      "loss": 1.7187,
+      "step": 6225
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3662698566913605,
+      "learning_rate": 4.411430783286447e-05,
+      "loss": 1.6875,
+      "step": 6230
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.3431253731250763,
+      "learning_rate": 4.410524253102003e-05,
+      "loss": 1.9651,
+      "step": 6235
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.5281440615653992,
+      "learning_rate": 4.409617118620574e-05,
+      "loss": 1.9109,
+      "step": 6240
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.659487247467041,
+      "learning_rate": 4.408709380129088e-05,
+      "loss": 1.9324,
+      "step": 6245
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7130793333053589,
+      "learning_rate": 4.407801037914662e-05,
+      "loss": 1.813,
+      "step": 6250
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6839697360992432,
+      "learning_rate": 4.4068920922646015e-05,
+      "loss": 1.8927,
+      "step": 6255
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.225645661354065,
+      "learning_rate": 4.405982543466406e-05,
+      "loss": 1.9195,
+      "step": 6260
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6370910406112671,
+      "learning_rate": 4.4050723918077664e-05,
+      "loss": 2.0735,
+      "step": 6265
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.8523099422454834,
+      "learning_rate": 4.4041616375765614e-05,
+      "loss": 1.8858,
+      "step": 6270
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.1719610691070557,
+      "learning_rate": 4.4032502810608614e-05,
+      "loss": 1.7128,
+      "step": 6275
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.9267966747283936,
+      "learning_rate": 4.402338322548929e-05,
+      "loss": 1.6958,
+      "step": 6280
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.4360824525356293,
+      "learning_rate": 4.401425762329214e-05,
+      "loss": 1.9678,
+      "step": 6285
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7322136163711548,
+      "learning_rate": 4.400512600690361e-05,
+      "loss": 1.8844,
+      "step": 6290
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.1388806104660034,
+      "learning_rate": 4.399598837921199e-05,
+      "loss": 1.645,
+      "step": 6295
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.42380547523498535,
+      "learning_rate": 4.3986844743107525e-05,
+      "loss": 1.9983,
+      "step": 6300
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.9260616302490234,
+      "learning_rate": 4.397769510148233e-05,
+      "loss": 1.6449,
+      "step": 6305
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5340147614479065,
+      "learning_rate": 4.396853945723043e-05,
+      "loss": 1.7605,
+      "step": 6310
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.5331189632415771,
+      "learning_rate": 4.395937781324774e-05,
+      "loss": 1.8059,
+      "step": 6315
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6908158659934998,
+      "learning_rate": 4.395021017243208e-05,
+      "loss": 1.9612,
+      "step": 6320
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8586407899856567,
+      "learning_rate": 4.3941036537683176e-05,
+      "loss": 1.9027,
+      "step": 6325
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7058882713317871,
+      "learning_rate": 4.393185691190264e-05,
+      "loss": 1.9719,
+      "step": 6330
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.7748107314109802,
+      "learning_rate": 4.392267129799397e-05,
+      "loss": 2.0075,
+      "step": 6335
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.6403058767318726,
+      "learning_rate": 4.391347969886256e-05,
+      "loss": 2.2573,
+      "step": 6340
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.7095941305160522,
+      "learning_rate": 4.3904282117415724e-05,
+      "loss": 1.8924,
+      "step": 6345
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.880134642124176,
+      "learning_rate": 4.389507855656263e-05,
+      "loss": 1.6801,
+      "step": 6350
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.8959532976150513,
+      "learning_rate": 4.3885869019214374e-05,
+      "loss": 1.6508,
+      "step": 6355
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.0066273212432861,
+      "learning_rate": 4.38766535082839e-05,
+      "loss": 1.9506,
+      "step": 6360
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.32858455181121826,
+      "learning_rate": 4.386743202668609e-05,
+      "loss": 2.0022,
+      "step": 6365
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.6246061325073242,
+      "learning_rate": 4.3858204577337674e-05,
+      "loss": 1.8174,
+      "step": 6370
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.4996178448200226,
+      "learning_rate": 4.3848971163157296e-05,
+      "loss": 1.7198,
+      "step": 6375
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.6967160105705261,
+      "learning_rate": 4.383973178706545e-05,
+      "loss": 1.9285,
+      "step": 6380
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.0754936933517456,
+      "learning_rate": 4.383048645198458e-05,
+      "loss": 2.0293,
+      "step": 6385
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.49468663334846497,
+      "learning_rate": 4.3821235160838955e-05,
+      "loss": 1.8142,
+      "step": 6390
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 0.9518940448760986,
+      "learning_rate": 4.3811977916554746e-05,
+      "loss": 2.1539,
+      "step": 6395
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5778300166130066,
+      "learning_rate": 4.3802714722060024e-05,
+      "loss": 1.8289,
+      "step": 6400
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8983391523361206,
+      "learning_rate": 4.379344558028471e-05,
+      "loss": 1.6746,
+      "step": 6405
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.6165648102760315,
+      "learning_rate": 4.3784170494160646e-05,
+      "loss": 2.0491,
+      "step": 6410
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.9584977626800537,
+      "learning_rate": 4.377488946662152e-05,
+      "loss": 1.8224,
+      "step": 6415
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.769154667854309,
+      "learning_rate": 4.376560250060292e-05,
+      "loss": 1.8792,
+      "step": 6420
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5984408855438232,
+      "learning_rate": 4.375630959904229e-05,
+      "loss": 1.9362,
+      "step": 6425
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.7467585206031799,
+      "learning_rate": 4.3747010764878985e-05,
+      "loss": 1.4512,
+      "step": 6430
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.2350887060165405,
+      "learning_rate": 4.373770600105419e-05,
+      "loss": 1.8879,
+      "step": 6435
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 3.794597625732422,
+      "learning_rate": 4.372839531051103e-05,
+      "loss": 1.7501,
+      "step": 6440
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.3860040605068207,
+      "learning_rate": 4.371907869619444e-05,
+      "loss": 1.7878,
+      "step": 6445
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.4788179099559784,
+      "learning_rate": 4.3709756161051266e-05,
+      "loss": 2.0143,
+      "step": 6450
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5330110192298889,
+      "learning_rate": 4.3700427708030205e-05,
+      "loss": 1.6505,
+      "step": 6455
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.7788562178611755,
+      "learning_rate": 4.369109334008186e-05,
+      "loss": 1.9456,
+      "step": 6460
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.4161795675754547,
+      "learning_rate": 4.3681753060158656e-05,
+      "loss": 2.1514,
+      "step": 6465
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8420090079307556,
+      "learning_rate": 4.367240687121495e-05,
+      "loss": 1.8194,
+      "step": 6470
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.37459495663642883,
+      "learning_rate": 4.366305477620689e-05,
+      "loss": 1.8492,
+      "step": 6475
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8855760097503662,
+      "learning_rate": 4.365369677809257e-05,
+      "loss": 1.6327,
+      "step": 6480
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.2534199953079224,
+      "learning_rate": 4.36443328798319e-05,
+      "loss": 1.9092,
+      "step": 6485
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.44040653109550476,
+      "learning_rate": 4.363496308438666e-05,
+      "loss": 1.9596,
+      "step": 6490
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6483808755874634,
+      "learning_rate": 4.3625587394720534e-05,
+      "loss": 1.8081,
+      "step": 6495
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.1044362783432007,
+      "learning_rate": 4.3616205813799025e-05,
+      "loss": 1.8881,
+      "step": 6500
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.2582200765609741,
+      "learning_rate": 4.360681834458953e-05,
+      "loss": 2.0978,
+      "step": 6505
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6778598427772522,
+      "learning_rate": 4.359742499006129e-05,
+      "loss": 2.1348,
+      "step": 6510
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.45896151661872864,
+      "learning_rate": 4.35880257531854e-05,
+      "loss": 1.8157,
+      "step": 6515
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8110174536705017,
+      "learning_rate": 4.357862063693486e-05,
+      "loss": 1.9552,
+      "step": 6520
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5105950236320496,
+      "learning_rate": 4.356920964428448e-05,
+      "loss": 2.0259,
+      "step": 6525
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7677923440933228,
+      "learning_rate": 4.3559792778210945e-05,
+      "loss": 1.9829,
+      "step": 6530
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.0216026306152344,
+      "learning_rate": 4.3550370041692805e-05,
+      "loss": 1.8117,
+      "step": 6535
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.8491010069847107,
+      "learning_rate": 4.354094143771047e-05,
+      "loss": 1.9182,
+      "step": 6540
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7707613706588745,
+      "learning_rate": 4.353150696924619e-05,
+      "loss": 1.9682,
+      "step": 6545
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6204885840415955,
+      "learning_rate": 4.3522066639284085e-05,
+      "loss": 1.641,
+      "step": 6550
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7384446263313293,
+      "learning_rate": 4.3512620450810114e-05,
+      "loss": 1.9758,
+      "step": 6555
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.7315991520881653,
+      "learning_rate": 4.35031684068121e-05,
+      "loss": 1.6545,
+      "step": 6560
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5005430579185486,
+      "learning_rate": 4.349371051027972e-05,
+      "loss": 1.9201,
+      "step": 6565
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.0951027870178223,
+      "learning_rate": 4.348424676420449e-05,
+      "loss": 1.9609,
+      "step": 6570
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5726012587547302,
+      "learning_rate": 4.3474777171579794e-05,
+      "loss": 1.835,
+      "step": 6575
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7009186744689941,
+      "learning_rate": 4.3465301735400844e-05,
+      "loss": 2.0274,
+      "step": 6580
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8929822444915771,
+      "learning_rate": 4.345582045866472e-05,
+      "loss": 1.8169,
+      "step": 6585
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.615864098072052,
+      "learning_rate": 4.3446333344370326e-05,
+      "loss": 1.7477,
+      "step": 6590
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7010105848312378,
+      "learning_rate": 4.343684039551845e-05,
+      "loss": 1.6822,
+      "step": 6595
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.7396606206893921,
+      "learning_rate": 4.3427341615111696e-05,
+      "loss": 1.7189,
+      "step": 6600
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.8250526189804077,
+      "learning_rate": 4.3417837006154505e-05,
+      "loss": 1.8397,
+      "step": 6605
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.417876660823822,
+      "learning_rate": 4.340832657165318e-05,
+      "loss": 1.7366,
+      "step": 6610
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.5691598653793335,
+      "learning_rate": 4.3398810314615876e-05,
+      "loss": 1.678,
+      "step": 6615
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.3957613706588745,
+      "learning_rate": 4.338928823805256e-05,
+      "loss": 2.2207,
+      "step": 6620
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5127105712890625,
+      "learning_rate": 4.337976034497507e-05,
+      "loss": 2.0833,
+      "step": 6625
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7685470581054688,
+      "learning_rate": 4.337022663839706e-05,
+      "loss": 1.6234,
+      "step": 6630
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.8764786124229431,
+      "learning_rate": 4.336068712133403e-05,
+      "loss": 1.8451,
+      "step": 6635
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5960032939910889,
+      "learning_rate": 4.335114179680333e-05,
+      "loss": 1.9596,
+      "step": 6640
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.9183564782142639,
+      "learning_rate": 4.334159066782413e-05,
+      "loss": 1.9834,
+      "step": 6645
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.964171051979065,
+      "learning_rate": 4.333203373741744e-05,
+      "loss": 1.8609,
+      "step": 6650
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5139592885971069,
+      "learning_rate": 4.332247100860612e-05,
+      "loss": 1.8969,
+      "step": 6655
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.40804627537727356,
+      "learning_rate": 4.331290248441483e-05,
+      "loss": 1.6632,
+      "step": 6660
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5056566596031189,
+      "learning_rate": 4.3303328167870105e-05,
+      "loss": 1.7616,
+      "step": 6665
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.207601547241211,
+      "learning_rate": 4.32937480620003e-05,
+      "loss": 1.9739,
+      "step": 6670
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.5137903094291687,
+      "learning_rate": 4.3284162169835565e-05,
+      "loss": 1.8736,
+      "step": 6675
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5918135046958923,
+      "learning_rate": 4.327457049440792e-05,
+      "loss": 2.1611,
+      "step": 6680
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.4136945903301239,
+      "learning_rate": 4.326497303875121e-05,
+      "loss": 2.0066,
+      "step": 6685
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7371619343757629,
+      "learning_rate": 4.3255369805901096e-05,
+      "loss": 1.5216,
+      "step": 6690
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 7.495757579803467,
+      "learning_rate": 4.324576079889507e-05,
+      "loss": 1.752,
+      "step": 6695
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8859977722167969,
+      "learning_rate": 4.323614602077246e-05,
+      "loss": 1.7231,
+      "step": 6700
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.48695892095565796,
+      "learning_rate": 4.3226525474574405e-05,
+      "loss": 1.612,
+      "step": 6705
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5769939422607422,
+      "learning_rate": 4.321689916334387e-05,
+      "loss": 1.9836,
+      "step": 6710
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7489534616470337,
+      "learning_rate": 4.3207267090125646e-05,
+      "loss": 1.6957,
+      "step": 6715
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.0079684257507324,
+      "learning_rate": 4.3197629257966365e-05,
+      "loss": 1.89,
+      "step": 6720
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.8229489326477051,
+      "learning_rate": 4.3187985669914454e-05,
+      "loss": 1.929,
+      "step": 6725
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.6666339039802551,
+      "learning_rate": 4.317833632902016e-05,
+      "loss": 1.823,
+      "step": 6730
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8089262247085571,
+      "learning_rate": 4.316868123833558e-05,
+      "loss": 1.9112,
+      "step": 6735
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6247695684432983,
+      "learning_rate": 4.31590204009146e-05,
+      "loss": 1.637,
+      "step": 6740
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.4710233211517334,
+      "learning_rate": 4.314935381981293e-05,
+      "loss": 2.0317,
+      "step": 6745
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6251475214958191,
+      "learning_rate": 4.31396814980881e-05,
+      "loss": 2.1173,
+      "step": 6750
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8826267719268799,
+      "learning_rate": 4.3130003438799466e-05,
+      "loss": 2.0483,
+      "step": 6755
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6971041560173035,
+      "learning_rate": 4.312031964500818e-05,
+      "loss": 1.921,
+      "step": 6760
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.6338077187538147,
+      "learning_rate": 4.311063011977723e-05,
+      "loss": 1.7711,
+      "step": 6765
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.1039173603057861,
+      "learning_rate": 4.3100934866171385e-05,
+      "loss": 1.5759,
+      "step": 6770
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.1805471181869507,
+      "learning_rate": 4.3091233887257256e-05,
+      "loss": 1.7043,
+      "step": 6775
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.7616626024246216,
+      "learning_rate": 4.308152718610324e-05,
+      "loss": 1.739,
+      "step": 6780
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.3182941675186157,
+      "learning_rate": 4.307181476577957e-05,
+      "loss": 1.8309,
+      "step": 6785
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.0036624670028687,
+      "learning_rate": 4.3062096629358285e-05,
+      "loss": 1.7662,
+      "step": 6790
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 4.941795349121094,
+      "learning_rate": 4.3052372779913194e-05,
+      "loss": 1.7195,
+      "step": 6795
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5514867901802063,
+      "learning_rate": 4.304264322051997e-05,
+      "loss": 1.899,
+      "step": 6800
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.600469708442688,
+      "learning_rate": 4.303290795425605e-05,
+      "loss": 2.0107,
+      "step": 6805
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.567573070526123,
+      "learning_rate": 4.302316698420068e-05,
+      "loss": 1.89,
+      "step": 6810
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.5658138394355774,
+      "learning_rate": 4.301342031343493e-05,
+      "loss": 2.2113,
+      "step": 6815
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.3791340589523315,
+      "learning_rate": 4.300366794504167e-05,
+      "loss": 2.0512,
+      "step": 6820
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7716684937477112,
+      "learning_rate": 4.2993909882105546e-05,
+      "loss": 1.9361,
+      "step": 6825
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.4616861939430237,
+      "learning_rate": 4.2984146127713046e-05,
+      "loss": 2.2654,
+      "step": 6830
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.1081273555755615,
+      "learning_rate": 4.2974376684952414e-05,
+      "loss": 2.059,
+      "step": 6835
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.547261118888855,
+      "learning_rate": 4.296460155691373e-05,
+      "loss": 1.8463,
+      "step": 6840
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.597197949886322,
+      "learning_rate": 4.295482074668885e-05,
+      "loss": 1.8631,
+      "step": 6845
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.4494568705558777,
+      "learning_rate": 4.2945034257371445e-05,
+      "loss": 1.693,
+      "step": 6850
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7037421464920044,
+      "learning_rate": 4.293524209205696e-05,
+      "loss": 1.7386,
+      "step": 6855
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.509117066860199,
+      "learning_rate": 4.292544425384265e-05,
+      "loss": 1.7397,
+      "step": 6860
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8861694931983948,
+      "learning_rate": 4.2915640745827576e-05,
+      "loss": 1.8453,
+      "step": 6865
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8716956973075867,
+      "learning_rate": 4.290583157111255e-05,
+      "loss": 1.8874,
+      "step": 6870
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.608605682849884,
+      "learning_rate": 4.2896016732800235e-05,
+      "loss": 1.6868,
+      "step": 6875
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.43817174434661865,
+      "learning_rate": 4.288619623399503e-05,
+      "loss": 1.7757,
+      "step": 6880
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.7979812622070312,
+      "learning_rate": 4.2876370077803174e-05,
+      "loss": 1.9506,
+      "step": 6885
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8355962634086609,
+      "learning_rate": 4.286653826733265e-05,
+      "loss": 1.8282,
+      "step": 6890
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.4820092022418976,
+      "learning_rate": 4.285670080569326e-05,
+      "loss": 1.7797,
+      "step": 6895
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.9699495434761047,
+      "learning_rate": 4.284685769599658e-05,
+      "loss": 1.8737,
+      "step": 6900
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6602120995521545,
+      "learning_rate": 4.283700894135597e-05,
+      "loss": 1.8891,
+      "step": 6905
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5803684592247009,
+      "learning_rate": 4.282715454488661e-05,
+      "loss": 2.1173,
+      "step": 6910
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8737105131149292,
+      "learning_rate": 4.2817294509705394e-05,
+      "loss": 1.9188,
+      "step": 6915
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.0067740678787231,
+      "learning_rate": 4.280742883893106e-05,
+      "loss": 1.9328,
+      "step": 6920
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7453013062477112,
+      "learning_rate": 4.279755753568413e-05,
+      "loss": 2.0628,
+      "step": 6925
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7748093008995056,
+      "learning_rate": 4.278768060308686e-05,
+      "loss": 1.8465,
+      "step": 6930
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.8429641723632812,
+      "learning_rate": 4.277779804426333e-05,
+      "loss": 1.6647,
+      "step": 6935
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.9702791571617126,
+      "learning_rate": 4.2767909862339374e-05,
+      "loss": 2.1712,
+      "step": 6940
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.3242080211639404,
+      "learning_rate": 4.275801606044262e-05,
+      "loss": 1.812,
+      "step": 6945
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.5527956485748291,
+      "learning_rate": 4.2748116641702466e-05,
+      "loss": 1.7873,
+      "step": 6950
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.6284675002098083,
+      "learning_rate": 4.27382116092501e-05,
+      "loss": 1.9656,
+      "step": 6955
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 10.499287605285645,
+      "learning_rate": 4.272830096621845e-05,
+      "loss": 1.724,
+      "step": 6960
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.9335476756095886,
+      "learning_rate": 4.271838471574227e-05,
+      "loss": 1.9242,
+      "step": 6965
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.3686840832233429,
+      "learning_rate": 4.270846286095803e-05,
+      "loss": 2.0095,
+      "step": 6970
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.7806186676025391,
+      "learning_rate": 4.2698535405004034e-05,
+      "loss": 1.7676,
+      "step": 6975
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.4177755117416382,
+      "learning_rate": 4.2688602351020315e-05,
+      "loss": 1.824,
+      "step": 6980
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5843412280082703,
+      "learning_rate": 4.267866370214868e-05,
+      "loss": 1.9222,
+      "step": 6985
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.8569367527961731,
+      "learning_rate": 4.266871946153272e-05,
+      "loss": 1.8849,
+      "step": 6990
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.5695585012435913,
+      "learning_rate": 4.265876963231779e-05,
+      "loss": 2.0957,
+      "step": 6995
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 6.371403694152832,
+      "learning_rate": 4.2648814217651013e-05,
+      "loss": 1.9637,
+      "step": 7000
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.9307239651679993,
+      "learning_rate": 4.2638853220681276e-05,
+      "loss": 2.0494,
+      "step": 7005
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.0416122674942017,
+      "learning_rate": 4.262888664455923e-05,
+      "loss": 1.9864,
+      "step": 7010
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.673014223575592,
+      "learning_rate": 4.261891449243729e-05,
+      "loss": 1.7286,
+      "step": 7015
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7358102202415466,
+      "learning_rate": 4.2608936767469655e-05,
+      "loss": 1.7191,
+      "step": 7020
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7470425367355347,
+      "learning_rate": 4.259895347281225e-05,
+      "loss": 1.6247,
+      "step": 7025
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.47723567485809326,
+      "learning_rate": 4.258896461162279e-05,
+      "loss": 1.7392,
+      "step": 7030
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.9144250154495239,
+      "learning_rate": 4.257897018706074e-05,
+      "loss": 1.8239,
+      "step": 7035
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.484836608171463,
+      "learning_rate": 4.2568970202287334e-05,
+      "loss": 1.6783,
+      "step": 7040
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.3341679573059082,
+      "learning_rate": 4.255896466046555e-05,
+      "loss": 1.8347,
+      "step": 7045
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.928821861743927,
+      "learning_rate": 4.254895356476013e-05,
+      "loss": 1.8622,
+      "step": 7050
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.6659140586853027,
+      "learning_rate": 4.253893691833758e-05,
+      "loss": 2.1149,
+      "step": 7055
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.630410373210907,
+      "learning_rate": 4.252891472436616e-05,
+      "loss": 1.8716,
+      "step": 7060
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.746567964553833,
+      "learning_rate": 4.251888698601587e-05,
+      "loss": 1.8274,
+      "step": 7065
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6550970673561096,
+      "learning_rate": 4.250885370645847e-05,
+      "loss": 1.9049,
+      "step": 7070
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8050583004951477,
+      "learning_rate": 4.249881488886749e-05,
+      "loss": 1.9531,
+      "step": 7075
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8060429692268372,
+      "learning_rate": 4.248877053641819e-05,
+      "loss": 1.84,
+      "step": 7080
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6672247648239136,
+      "learning_rate": 4.2478720652287596e-05,
+      "loss": 1.9365,
+      "step": 7085
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.9606338739395142,
+      "learning_rate": 4.246866523965446e-05,
+      "loss": 1.8334,
+      "step": 7090
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8220806121826172,
+      "learning_rate": 4.2458604301699324e-05,
+      "loss": 2.2147,
+      "step": 7095
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.5984706282615662,
+      "learning_rate": 4.2448537841604435e-05,
+      "loss": 1.9588,
+      "step": 7100
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6690345406532288,
+      "learning_rate": 4.243846586255381e-05,
+      "loss": 1.8757,
+      "step": 7105
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.075214385986328,
+      "learning_rate": 4.242838836773321e-05,
+      "loss": 1.7474,
+      "step": 7110
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8031553626060486,
+      "learning_rate": 4.2418305360330134e-05,
+      "loss": 1.6749,
+      "step": 7115
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6923268437385559,
+      "learning_rate": 4.240821684353382e-05,
+      "loss": 1.7302,
+      "step": 7120
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5841795206069946,
+      "learning_rate": 4.2398122820535266e-05,
+      "loss": 1.7478,
+      "step": 7125
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6711515188217163,
+      "learning_rate": 4.2388023294527204e-05,
+      "loss": 2.0452,
+      "step": 7130
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.6497505903244019,
+      "learning_rate": 4.237791826870409e-05,
+      "loss": 1.7706,
+      "step": 7135
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7864512801170349,
+      "learning_rate": 4.236780774626215e-05,
+      "loss": 2.1351,
+      "step": 7140
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.8343969583511353,
+      "learning_rate": 4.235769173039932e-05,
+      "loss": 1.7934,
+      "step": 7145
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.734727144241333,
+      "learning_rate": 4.234959496459615e-05,
+      "loss": 1.9032,
+      "step": 7150
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.37156343460083,
+      "learning_rate": 4.233946906864011e-05,
+      "loss": 2.0043,
+      "step": 7155
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.48946473002433777,
+      "learning_rate": 4.232933768822668e-05,
+      "loss": 2.1292,
+      "step": 7160
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.64715576171875,
+      "learning_rate": 4.231920082656041e-05,
+      "loss": 1.7692,
+      "step": 7165
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7342175245285034,
+      "learning_rate": 4.230905848684757e-05,
+      "loss": 1.8173,
+      "step": 7170
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.2754048109054565,
+      "learning_rate": 4.2298910672296186e-05,
+      "loss": 1.6126,
+      "step": 7175
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.9194117784500122,
+      "learning_rate": 4.2288757386115994e-05,
+      "loss": 1.9669,
+      "step": 7180
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 3.632533550262451,
+      "learning_rate": 4.227859863151847e-05,
+      "loss": 1.7307,
+      "step": 7185
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.4630647599697113,
+      "learning_rate": 4.226843441171682e-05,
+      "loss": 1.6671,
+      "step": 7190
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5320610404014587,
+      "learning_rate": 4.225826472992598e-05,
+      "loss": 1.8363,
+      "step": 7195
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6387033462524414,
+      "learning_rate": 4.2248089589362593e-05,
+      "loss": 2.0306,
+      "step": 7200
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5456752181053162,
+      "learning_rate": 4.223790899324507e-05,
+      "loss": 1.7391,
+      "step": 7205
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 6.061885833740234,
+      "learning_rate": 4.2227722944793515e-05,
+      "loss": 1.9154,
+      "step": 7210
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6948996186256409,
+      "learning_rate": 4.221753144722976e-05,
+      "loss": 1.8669,
+      "step": 7215
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.5726954936981201,
+      "learning_rate": 4.220733450377737e-05,
+      "loss": 2.0606,
+      "step": 7220
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.9554140567779541,
+      "learning_rate": 4.219713211766162e-05,
+      "loss": 1.8438,
+      "step": 7225
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 0.6293787360191345,
+      "learning_rate": 4.218692429210953e-05,
+      "loss": 1.825,
+      "step": 7230
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.0556727647781372,
+      "learning_rate": 4.217671103034982e-05,
+      "loss": 1.7117,
+      "step": 7235
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6158782243728638,
+      "learning_rate": 4.216649233561293e-05,
+      "loss": 1.9096,
+      "step": 7240
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.9434940218925476,
+      "learning_rate": 4.2156268211131027e-05,
+      "loss": 1.8082,
+      "step": 7245
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.7455072402954102,
+      "learning_rate": 4.214603866013799e-05,
+      "loss": 1.9597,
+      "step": 7250
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.912811279296875,
+      "learning_rate": 4.213580368586942e-05,
+      "loss": 1.8557,
+      "step": 7255
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6361580491065979,
+      "learning_rate": 4.212556329156263e-05,
+      "loss": 2.0177,
+      "step": 7260
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 6.6606597900390625,
+      "learning_rate": 4.211531748045664e-05,
+      "loss": 1.7674,
+      "step": 7265
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5496176481246948,
+      "learning_rate": 4.2105066255792185e-05,
+      "loss": 2.1907,
+      "step": 7270
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6689378023147583,
+      "learning_rate": 4.209480962081174e-05,
+      "loss": 1.7675,
+      "step": 7275
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.8176302909851074,
+      "learning_rate": 4.208454757875945e-05,
+      "loss": 1.6746,
+      "step": 7280
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.5315504670143127,
+      "learning_rate": 4.2074280132881184e-05,
+      "loss": 1.6911,
+      "step": 7285
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.6745172739028931,
+      "learning_rate": 4.2064007286424545e-05,
+      "loss": 1.786,
+      "step": 7290
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8373345136642456,
+      "learning_rate": 4.205372904263881e-05,
+      "loss": 2.0911,
+      "step": 7295
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.54158616065979,
+      "learning_rate": 4.204344540477499e-05,
+      "loss": 1.7611,
+      "step": 7300
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6230586767196655,
+      "learning_rate": 4.203315637608578e-05,
+      "loss": 1.7406,
+      "step": 7305
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.31952166557312,
+      "learning_rate": 4.202286195982559e-05,
+      "loss": 1.8494,
+      "step": 7310
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6334202885627747,
+      "learning_rate": 4.201256215925054e-05,
+      "loss": 1.6764,
+      "step": 7315
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.43206092715263367,
+      "learning_rate": 4.200225697761844e-05,
+      "loss": 1.9862,
+      "step": 7320
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6146365404129028,
+      "learning_rate": 4.1991946418188805e-05,
+      "loss": 2.0393,
+      "step": 7325
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.48394322395324707,
+      "learning_rate": 4.198163048422287e-05,
+      "loss": 1.8087,
+      "step": 7330
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8574031591415405,
+      "learning_rate": 4.197130917898355e-05,
+      "loss": 1.9837,
+      "step": 7335
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.4108412563800812,
+      "learning_rate": 4.196098250573545e-05,
+      "loss": 2.2631,
+      "step": 7340
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.6564179062843323,
+      "learning_rate": 4.195065046774491e-05,
+      "loss": 2.0216,
+      "step": 7345
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.3432505130767822,
+      "learning_rate": 4.194031306827993e-05,
+      "loss": 2.1613,
+      "step": 7350
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.463270366191864,
+      "learning_rate": 4.192997031061021e-05,
+      "loss": 1.6699,
+      "step": 7355
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.4397541284561157,
+      "learning_rate": 4.1919622198007176e-05,
+      "loss": 1.7908,
+      "step": 7360
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.724583089351654,
+      "learning_rate": 4.190926873374391e-05,
+      "loss": 1.9014,
+      "step": 7365
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.0059683322906494,
+      "learning_rate": 4.189890992109521e-05,
+      "loss": 1.6797,
+      "step": 7370
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.4795385599136353,
+      "learning_rate": 4.1888545763337554e-05,
+      "loss": 1.8065,
+      "step": 7375
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.9191558361053467,
+      "learning_rate": 4.187817626374912e-05,
+      "loss": 1.9115,
+      "step": 7380
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.3600243628025055,
+      "learning_rate": 4.186780142560977e-05,
+      "loss": 2.096,
+      "step": 7385
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8530924320220947,
+      "learning_rate": 4.185742125220106e-05,
+      "loss": 1.5549,
+      "step": 7390
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6997917890548706,
+      "learning_rate": 4.1847035746806205e-05,
+      "loss": 1.9537,
+      "step": 7395
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.5921128392219543,
+      "learning_rate": 4.1836644912710166e-05,
+      "loss": 2.0546,
+      "step": 7400
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.5461764335632324,
+      "learning_rate": 4.1826248753199526e-05,
+      "loss": 1.5721,
+      "step": 7405
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8535491228103638,
+      "learning_rate": 4.1815847271562594e-05,
+      "loss": 1.8103,
+      "step": 7410
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.9533465504646301,
+      "learning_rate": 4.180544047108934e-05,
+      "loss": 1.87,
+      "step": 7415
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.49469488859176636,
+      "learning_rate": 4.179502835507143e-05,
+      "loss": 1.8931,
+      "step": 7420
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.4135587215423584,
+      "learning_rate": 4.1784610926802214e-05,
+      "loss": 2.1653,
+      "step": 7425
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.6948500871658325,
+      "learning_rate": 4.1774188189576705e-05,
+      "loss": 1.8297,
+      "step": 7430
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.7468460202217102,
+      "learning_rate": 4.17637601466916e-05,
+      "loss": 1.9556,
+      "step": 7435
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8151171207427979,
+      "learning_rate": 4.1753326801445285e-05,
+      "loss": 1.6892,
+      "step": 7440
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8748107552528381,
+      "learning_rate": 4.174288815713783e-05,
+      "loss": 1.7057,
+      "step": 7445
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.40320098400115967,
+      "learning_rate": 4.1732444217070943e-05,
+      "loss": 1.8913,
+      "step": 7450
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 0.8582679629325867,
+      "learning_rate": 4.172199498454804e-05,
+      "loss": 1.9204,
+      "step": 7455
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.35141122341156006,
+      "learning_rate": 4.1711540462874224e-05,
+      "loss": 1.9048,
+      "step": 7460
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.42094314098358154,
+      "learning_rate": 4.170108065535622e-05,
+      "loss": 1.9885,
+      "step": 7465
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6282821297645569,
+      "learning_rate": 4.169061556530248e-05,
+      "loss": 1.7992,
+      "step": 7470
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7469441890716553,
+      "learning_rate": 4.168014519602309e-05,
+      "loss": 1.9007,
+      "step": 7475
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5168988704681396,
+      "learning_rate": 4.1669669550829815e-05,
+      "loss": 1.8918,
+      "step": 7480
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.7953056693077087,
+      "learning_rate": 4.165918863303609e-05,
+      "loss": 2.1881,
+      "step": 7485
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6245642900466919,
+      "learning_rate": 4.164870244595703e-05,
+      "loss": 1.9878,
+      "step": 7490
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.3559062480926514,
+      "learning_rate": 4.16382109929094e-05,
+      "loss": 2.1096,
+      "step": 7495
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.1444271802902222,
+      "learning_rate": 4.162771427721164e-05,
+      "loss": 2.078,
+      "step": 7500
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5032902359962463,
+      "learning_rate": 4.161721230218384e-05,
+      "loss": 1.814,
+      "step": 7505
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.2069705724716187,
+      "learning_rate": 4.1606705071147775e-05,
+      "loss": 1.8121,
+      "step": 7510
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5158296823501587,
+      "learning_rate": 4.1596192587426874e-05,
+      "loss": 1.7489,
+      "step": 7515
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6261261701583862,
+      "learning_rate": 4.158567485434622e-05,
+      "loss": 1.8973,
+      "step": 7520
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.6703084707260132,
+      "learning_rate": 4.1575151875232565e-05,
+      "loss": 2.0307,
+      "step": 7525
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.8123947381973267,
+      "learning_rate": 4.156462365341431e-05,
+      "loss": 1.6317,
+      "step": 7530
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.7036327123641968,
+      "learning_rate": 4.1554090192221525e-05,
+      "loss": 1.4194,
+      "step": 7535
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5714519023895264,
+      "learning_rate": 4.154355149498594e-05,
+      "loss": 2.0492,
+      "step": 7540
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5328961610794067,
+      "learning_rate": 4.153300756504093e-05,
+      "loss": 1.8173,
+      "step": 7545
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.6683387756347656,
+      "learning_rate": 4.152245840572153e-05,
+      "loss": 2.0118,
+      "step": 7550
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5071770548820496,
+      "learning_rate": 4.151190402036443e-05,
+      "loss": 1.9307,
+      "step": 7555
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.5669893622398376,
+      "learning_rate": 4.150134441230797e-05,
+      "loss": 2.0048,
+      "step": 7560
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7352867722511292,
+      "learning_rate": 4.149077958489214e-05,
+      "loss": 2.0041,
+      "step": 7565
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 0.7905187010765076,
+      "learning_rate": 4.148020954145859e-05,
+      "loss": 2.0104,
+      "step": 7570
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.7344765067100525,
+      "learning_rate": 4.146963428535062e-05,
+      "loss": 1.8857,
+      "step": 7575
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.9353702664375305,
+      "learning_rate": 4.1459053819913164e-05,
+      "loss": 1.9718,
+      "step": 7580
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8625378012657166,
+      "learning_rate": 4.144846814849282e-05,
+      "loss": 1.9804,
+      "step": 7585
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.47366541624069214,
+      "learning_rate": 4.1437877274437817e-05,
+      "loss": 1.7501,
+      "step": 7590
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.9116063117980957,
+      "learning_rate": 4.1427281201098046e-05,
+      "loss": 1.4518,
+      "step": 7595
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.8927298784255981,
+      "learning_rate": 4.141667993182503e-05,
+      "loss": 1.7712,
+      "step": 7600
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.8056049942970276,
+      "learning_rate": 4.140607346997194e-05,
+      "loss": 1.6893,
+      "step": 7605
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.4493560791015625,
+      "learning_rate": 4.13954618188936e-05,
+      "loss": 1.5022,
+      "step": 7610
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.4495934844017029,
+      "learning_rate": 4.138484498194645e-05,
+      "loss": 2.0982,
+      "step": 7615
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6876751780509949,
+      "learning_rate": 4.1374222962488594e-05,
+      "loss": 1.8517,
+      "step": 7620
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.509681224822998,
+      "learning_rate": 4.136359576387976e-05,
+      "loss": 1.9499,
+      "step": 7625
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.3566737771034241,
+      "learning_rate": 4.135296338948134e-05,
+      "loss": 2.1347,
+      "step": 7630
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.6193088889122009,
+      "learning_rate": 4.1342325842656315e-05,
+      "loss": 1.9541,
+      "step": 7635
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8318669199943542,
+      "learning_rate": 4.133168312676935e-05,
+      "loss": 1.6854,
+      "step": 7640
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.5070250034332275,
+      "learning_rate": 4.1321035245186726e-05,
+      "loss": 1.8134,
+      "step": 7645
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 6.577232360839844,
+      "learning_rate": 4.1310382201276354e-05,
+      "loss": 2.0211,
+      "step": 7650
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 1.0688321590423584,
+      "learning_rate": 4.1299723998407774e-05,
+      "loss": 1.5149,
+      "step": 7655
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.326978921890259,
+      "learning_rate": 4.128906063995217e-05,
+      "loss": 2.042,
+      "step": 7660
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7662460803985596,
+      "learning_rate": 4.127839212928236e-05,
+      "loss": 1.7589,
+      "step": 7665
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.7111654877662659,
+      "learning_rate": 4.126771846977277e-05,
+      "loss": 1.8719,
+      "step": 7670
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.8834118843078613,
+      "learning_rate": 4.125703966479948e-05,
+      "loss": 2.0589,
+      "step": 7675
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 0.599921464920044,
+      "learning_rate": 4.124635571774018e-05,
+      "loss": 1.9423,
+      "step": 7680
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6738287806510925,
+      "learning_rate": 4.12356666319742e-05,
+      "loss": 1.9728,
+      "step": 7685
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6264171004295349,
+      "learning_rate": 4.122497241088247e-05,
+      "loss": 1.961,
+      "step": 7690
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.831802487373352,
+      "learning_rate": 4.121427305784758e-05,
+      "loss": 2.0417,
+      "step": 7695
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.9367872476577759,
+      "learning_rate": 4.120356857625372e-05,
+      "loss": 1.8298,
+      "step": 7700
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6761164665222168,
+      "learning_rate": 4.11928589694867e-05,
+      "loss": 1.6786,
+      "step": 7705
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7731988430023193,
+      "learning_rate": 4.118214424093396e-05,
+      "loss": 1.8711,
+      "step": 7710
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.530742347240448,
+      "learning_rate": 4.117142439398457e-05,
+      "loss": 2.009,
+      "step": 7715
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6078718900680542,
+      "learning_rate": 4.116069943202919e-05,
+      "loss": 1.9871,
+      "step": 7720
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.6330639719963074,
+      "learning_rate": 4.114996935846014e-05,
+      "loss": 1.7457,
+      "step": 7725
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.661787211894989,
+      "learning_rate": 4.113923417667131e-05,
+      "loss": 1.7891,
+      "step": 7730
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.9162965416908264,
+      "learning_rate": 4.112849389005823e-05,
+      "loss": 1.9484,
+      "step": 7735
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8108046054840088,
+      "learning_rate": 4.1117748502018036e-05,
+      "loss": 1.5814,
+      "step": 7740
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5153409838676453,
+      "learning_rate": 4.110699801594951e-05,
+      "loss": 1.4343,
+      "step": 7745
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.30527204275131226,
+      "learning_rate": 4.1096242435253e-05,
+      "loss": 2.0052,
+      "step": 7750
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5223674178123474,
+      "learning_rate": 4.108548176333049e-05,
+      "loss": 2.1677,
+      "step": 7755
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5189133286476135,
+      "learning_rate": 4.107471600358557e-05,
+      "loss": 1.9642,
+      "step": 7760
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6149393320083618,
+      "learning_rate": 4.1063945159423433e-05,
+      "loss": 1.8314,
+      "step": 7765
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.6912689805030823,
+      "learning_rate": 4.105316923425091e-05,
+      "loss": 1.9248,
+      "step": 7770
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.5106381177902222,
+      "learning_rate": 4.1042388231476384e-05,
+      "loss": 2.0442,
+      "step": 7775
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.4784005880355835,
+      "learning_rate": 4.10316021545099e-05,
+      "loss": 1.7323,
+      "step": 7780
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 0.8782150149345398,
+      "learning_rate": 4.102081100676307e-05,
+      "loss": 1.8736,
+      "step": 7785
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.3147215843200684,
+      "learning_rate": 4.101001479164913e-05,
+      "loss": 1.7307,
+      "step": 7790
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5481510758399963,
+      "learning_rate": 4.099921351258292e-05,
+      "loss": 1.6104,
+      "step": 7795
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.617058515548706,
+      "learning_rate": 4.098840717298085e-05,
+      "loss": 1.9166,
+      "step": 7800
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.2154157161712646,
+      "learning_rate": 4.097759577626098e-05,
+      "loss": 1.8153,
+      "step": 7805
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7002790570259094,
+      "learning_rate": 4.096677932584293e-05,
+      "loss": 1.7273,
+      "step": 7810
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.020469069480896,
+      "learning_rate": 4.095595782514794e-05,
+      "loss": 2.0839,
+      "step": 7815
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.7469928860664368,
+      "learning_rate": 4.094513127759883e-05,
+      "loss": 1.7802,
+      "step": 7820
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.54106605052948,
+      "learning_rate": 4.093429968662005e-05,
+      "loss": 1.7512,
+      "step": 7825
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8594502210617065,
+      "learning_rate": 4.09234630556376e-05,
+      "loss": 1.9866,
+      "step": 7830
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5994930267333984,
+      "learning_rate": 4.09126213880791e-05,
+      "loss": 1.8757,
+      "step": 7835
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.0593057870864868,
+      "learning_rate": 4.090177468737375e-05,
+      "loss": 1.6099,
+      "step": 7840
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.458686888217926,
+      "learning_rate": 4.089092295695238e-05,
+      "loss": 2.0074,
+      "step": 7845
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5471193790435791,
+      "learning_rate": 4.0880066200247346e-05,
+      "loss": 2.0962,
+      "step": 7850
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5086942315101624,
+      "learning_rate": 4.0869204420692665e-05,
+      "loss": 1.5895,
+      "step": 7855
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7652096748352051,
+      "learning_rate": 4.085833762172387e-05,
+      "loss": 2.0743,
+      "step": 7860
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.0196499824523926,
+      "learning_rate": 4.084746580677816e-05,
+      "loss": 2.0508,
+      "step": 7865
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.541129469871521,
+      "learning_rate": 4.083658897929426e-05,
+      "loss": 1.857,
+      "step": 7870
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.4016308784484863,
+      "learning_rate": 4.082570714271248e-05,
+      "loss": 1.669,
+      "step": 7875
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5275091528892517,
+      "learning_rate": 4.081482030047476e-05,
+      "loss": 1.9097,
+      "step": 7880
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7216492295265198,
+      "learning_rate": 4.080392845602459e-05,
+      "loss": 1.9592,
+      "step": 7885
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5998548865318298,
+      "learning_rate": 4.079303161280706e-05,
+      "loss": 1.7357,
+      "step": 7890
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.7786138653755188,
+      "learning_rate": 4.0782129774268815e-05,
+      "loss": 2.0231,
+      "step": 7895
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.4815676510334015,
+      "learning_rate": 4.0771222943858104e-05,
+      "loss": 1.8351,
+      "step": 7900
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.5540018677711487,
+      "learning_rate": 4.076031112502474e-05,
+      "loss": 1.8817,
+      "step": 7905
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.1450433731079102,
+      "learning_rate": 4.074939432122014e-05,
+      "loss": 2.1311,
+      "step": 7910
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.612622857093811,
+      "learning_rate": 4.073847253589725e-05,
+      "loss": 1.9964,
+      "step": 7915
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6572409868240356,
+      "learning_rate": 4.0727545772510645e-05,
+      "loss": 1.8965,
+      "step": 7920
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8887383937835693,
+      "learning_rate": 4.071661403451643e-05,
+      "loss": 1.9419,
+      "step": 7925
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.8722497820854187,
+      "learning_rate": 4.070567732537232e-05,
+      "loss": 2.0361,
+      "step": 7930
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.48383402824401855,
+      "learning_rate": 4.0694735648537584e-05,
+      "loss": 1.7524,
+      "step": 7935
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5321690440177917,
+      "learning_rate": 4.0683789007473055e-05,
+      "loss": 1.8623,
+      "step": 7940
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6886895298957825,
+      "learning_rate": 4.067283740564115e-05,
+      "loss": 1.5339,
+      "step": 7945
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.0236868858337402,
+      "learning_rate": 4.066188084650584e-05,
+      "loss": 1.8861,
+      "step": 7950
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5006799697875977,
+      "learning_rate": 4.065091933353269e-05,
+      "loss": 2.1672,
+      "step": 7955
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.0537916421890259,
+      "learning_rate": 4.0639952870188815e-05,
+      "loss": 1.8532,
+      "step": 7960
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6310824155807495,
+      "learning_rate": 4.062898145994288e-05,
+      "loss": 1.9918,
+      "step": 7965
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.6223093271255493,
+      "learning_rate": 4.061800510626515e-05,
+      "loss": 1.8719,
+      "step": 7970
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.3217649459838867,
+      "learning_rate": 4.060702381262742e-05,
+      "loss": 1.9299,
+      "step": 7975
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.998820960521698,
+      "learning_rate": 4.0596037582503084e-05,
+      "loss": 1.8601,
+      "step": 7980
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.5109826922416687,
+      "learning_rate": 4.0585046419367053e-05,
+      "loss": 1.8392,
+      "step": 7985
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.45510098338127136,
+      "learning_rate": 4.057405032669582e-05,
+      "loss": 1.8877,
+      "step": 7990
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.46888962388038635,
+      "learning_rate": 4.056304930796746e-05,
+      "loss": 1.783,
+      "step": 7995
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 6.063878059387207,
+      "learning_rate": 4.055204336666158e-05,
+      "loss": 2.0542,
+      "step": 8000
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7420514822006226,
+      "learning_rate": 4.054103250625933e-05,
+      "loss": 2.0758,
+      "step": 8005
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.3912978768348694,
+      "learning_rate": 4.053001673024346e-05,
+      "loss": 1.8226,
+      "step": 8010
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.851407527923584,
+      "learning_rate": 4.051899604209823e-05,
+      "loss": 1.8753,
+      "step": 8015
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 1.0191857814788818,
+      "learning_rate": 4.0507970445309476e-05,
+      "loss": 2.0584,
+      "step": 8020
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5699238181114197,
+      "learning_rate": 4.049693994336459e-05,
+      "loss": 1.9835,
+      "step": 8025
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.5822308659553528,
+      "learning_rate": 4.0485904539752503e-05,
+      "loss": 1.8653,
+      "step": 8030
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7040959000587463,
+      "learning_rate": 4.0474864237963704e-05,
+      "loss": 1.7044,
+      "step": 8035
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6993011832237244,
+      "learning_rate": 4.046381904149024e-05,
+      "loss": 1.733,
+      "step": 8040
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.9968368411064148,
+      "learning_rate": 4.045276895382568e-05,
+      "loss": 1.8138,
+      "step": 8045
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7696248888969421,
+      "learning_rate": 4.044171397846517e-05,
+      "loss": 1.8775,
+      "step": 8050
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.6908630728721619,
+      "learning_rate": 4.0430654118905374e-05,
+      "loss": 1.8886,
+      "step": 8055
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7053937315940857,
+      "learning_rate": 4.041958937864453e-05,
+      "loss": 1.6934,
+      "step": 8060
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7439703941345215,
+      "learning_rate": 4.040851976118239e-05,
+      "loss": 1.8191,
+      "step": 8065
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.9062896370887756,
+      "learning_rate": 4.039744527002027e-05,
+      "loss": 1.6822,
+      "step": 8070
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5621775388717651,
+      "learning_rate": 4.038636590866103e-05,
+      "loss": 1.7791,
+      "step": 8075
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.6721243858337402,
+      "learning_rate": 4.0375281680609046e-05,
+      "loss": 1.9383,
+      "step": 8080
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5987261533737183,
+      "learning_rate": 4.036419258937026e-05,
+      "loss": 2.1479,
+      "step": 8085
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.3930743634700775,
+      "learning_rate": 4.0353098638452134e-05,
+      "loss": 2.0115,
+      "step": 8090
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7518051862716675,
+      "learning_rate": 4.0341999831363676e-05,
+      "loss": 2.1983,
+      "step": 8095
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.2611430883407593,
+      "learning_rate": 4.0330896171615434e-05,
+      "loss": 1.9812,
+      "step": 8100
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.8846983313560486,
+      "learning_rate": 4.031978766271948e-05,
+      "loss": 1.7216,
+      "step": 8105
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.0537869930267334,
+      "learning_rate": 4.030867430818941e-05,
+      "loss": 1.9612,
+      "step": 8110
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.5428566336631775,
+      "learning_rate": 4.0297556111540393e-05,
+      "loss": 1.9787,
+      "step": 8115
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7369449138641357,
+      "learning_rate": 4.028643307628909e-05,
+      "loss": 1.8887,
+      "step": 8120
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.3817769289016724,
+      "learning_rate": 4.027530520595371e-05,
+      "loss": 1.8445,
+      "step": 8125
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8463249802589417,
+      "learning_rate": 4.026417250405399e-05,
+      "loss": 2.0274,
+      "step": 8130
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.6427886486053467,
+      "learning_rate": 4.025303497411118e-05,
+      "loss": 2.0033,
+      "step": 8135
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6956122517585754,
+      "learning_rate": 4.024189261964808e-05,
+      "loss": 1.5285,
+      "step": 8140
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.7735291123390198,
+      "learning_rate": 4.023074544418901e-05,
+      "loss": 1.6345,
+      "step": 8145
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6430752873420715,
+      "learning_rate": 4.021959345125981e-05,
+      "loss": 1.7802,
+      "step": 8150
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8027344942092896,
+      "learning_rate": 4.0208436644387834e-05,
+      "loss": 1.6531,
+      "step": 8155
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.263455867767334,
+      "learning_rate": 4.019727502710197e-05,
+      "loss": 1.8208,
+      "step": 8160
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.4575340747833252,
+      "learning_rate": 4.018610860293264e-05,
+      "loss": 1.7946,
+      "step": 8165
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.8558323383331299,
+      "learning_rate": 4.017493737541177e-05,
+      "loss": 1.9303,
+      "step": 8170
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.5351092219352722,
+      "learning_rate": 4.0163761348072804e-05,
+      "loss": 1.8344,
+      "step": 8175
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.9965373873710632,
+      "learning_rate": 4.01525805244507e-05,
+      "loss": 1.7591,
+      "step": 8180
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5350607633590698,
+      "learning_rate": 4.014139490808196e-05,
+      "loss": 1.8909,
+      "step": 8185
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6248508095741272,
+      "learning_rate": 4.0130204502504575e-05,
+      "loss": 1.7733,
+      "step": 8190
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.5423145890235901,
+      "learning_rate": 4.0119009311258057e-05,
+      "loss": 1.9219,
+      "step": 8195
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.892905056476593,
+      "learning_rate": 4.0107809337883435e-05,
+      "loss": 1.7403,
+      "step": 8200
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.8622453808784485,
+      "learning_rate": 4.0096604585923246e-05,
+      "loss": 1.7071,
+      "step": 8205
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.459806352853775,
+      "learning_rate": 4.008539505892156e-05,
+      "loss": 1.8587,
+      "step": 8210
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.5942025184631348,
+      "learning_rate": 4.007418076042392e-05,
+      "loss": 1.7092,
+      "step": 8215
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.9066646695137024,
+      "learning_rate": 4.0062961693977406e-05,
+      "loss": 1.9558,
+      "step": 8220
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.168616533279419,
+      "learning_rate": 4.0051737863130594e-05,
+      "loss": 1.7656,
+      "step": 8225
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.33232593536376953,
+      "learning_rate": 4.004050927143358e-05,
+      "loss": 1.9195,
+      "step": 8230
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.546242892742157,
+      "learning_rate": 4.002927592243794e-05,
+      "loss": 1.9144,
+      "step": 8235
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.4449472427368164,
+      "learning_rate": 4.00180378196968e-05,
+      "loss": 1.5383,
+      "step": 8240
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6070382595062256,
+      "learning_rate": 4.000679496676473e-05,
+      "loss": 1.5511,
+      "step": 8245
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.3525921404361725,
+      "learning_rate": 3.9995547367197845e-05,
+      "loss": 1.9672,
+      "step": 8250
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.3565738201141357,
+      "learning_rate": 3.998429502455375e-05,
+      "loss": 1.61,
+      "step": 8255
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6370993256568909,
+      "learning_rate": 3.9973037942391564e-05,
+      "loss": 1.9055,
+      "step": 8260
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.565390408039093,
+      "learning_rate": 3.996177612427187e-05,
+      "loss": 1.8377,
+      "step": 8265
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6728577017784119,
+      "learning_rate": 3.995050957375678e-05,
+      "loss": 1.9646,
+      "step": 8270
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.3911441564559937,
+      "learning_rate": 3.99392382944099e-05,
+      "loss": 1.8547,
+      "step": 8275
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.7410084009170532,
+      "learning_rate": 3.992796228979631e-05,
+      "loss": 1.803,
+      "step": 8280
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.7153067588806152,
+      "learning_rate": 3.991668156348261e-05,
+      "loss": 1.7957,
+      "step": 8285
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.6121552586555481,
+      "learning_rate": 3.9905396119036876e-05,
+      "loss": 1.8041,
+      "step": 8290
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.8138649463653564,
+      "learning_rate": 3.989410596002869e-05,
+      "loss": 1.8954,
+      "step": 8295
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.1422274112701416,
+      "learning_rate": 3.9882811090029106e-05,
+      "loss": 1.8859,
+      "step": 8300
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.29510730504989624,
+      "learning_rate": 3.98715115126107e-05,
+      "loss": 1.9913,
+      "step": 8305
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.8496959209442139,
+      "learning_rate": 3.986020723134751e-05,
+      "loss": 1.8877,
+      "step": 8310
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.5750755071640015,
+      "learning_rate": 3.984889824981506e-05,
+      "loss": 1.6277,
+      "step": 8315
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.6985676288604736,
+      "learning_rate": 3.983758457159037e-05,
+      "loss": 1.823,
+      "step": 8320
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.0394619703292847,
+      "learning_rate": 3.982626620025196e-05,
+      "loss": 2.0183,
+      "step": 8325
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.700357973575592,
+      "learning_rate": 3.9814943139379815e-05,
+      "loss": 1.825,
+      "step": 8330
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.3674771785736084,
+      "learning_rate": 3.980361539255541e-05,
+      "loss": 2.0308,
+      "step": 8335
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.899864912033081,
+      "learning_rate": 3.979228296336168e-05,
+      "loss": 1.9477,
+      "step": 8340
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.9185581803321838,
+      "learning_rate": 3.978094585538308e-05,
+      "loss": 1.8853,
+      "step": 8345
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 0.47985824942588806,
+      "learning_rate": 3.9769604072205524e-05,
+      "loss": 1.742,
+      "step": 8350
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.42667725682258606,
+      "learning_rate": 3.9758257617416414e-05,
+      "loss": 1.935,
+      "step": 8355
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.3752597570419312,
+      "learning_rate": 3.974690649460461e-05,
+      "loss": 1.9357,
+      "step": 8360
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3676702380180359,
+      "learning_rate": 3.973555070736047e-05,
+      "loss": 1.7689,
+      "step": 8365
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7576048374176025,
+      "learning_rate": 3.9724190259275804e-05,
+      "loss": 1.9353,
+      "step": 8370
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8016186356544495,
+      "learning_rate": 3.9712825153943934e-05,
+      "loss": 1.8762,
+      "step": 8375
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8371986150741577,
+      "learning_rate": 3.970145539495962e-05,
+      "loss": 1.7327,
+      "step": 8380
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.9057819843292236,
+      "learning_rate": 3.9690080985919096e-05,
+      "loss": 1.5471,
+      "step": 8385
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.966324806213379,
+      "learning_rate": 3.9678701930420095e-05,
+      "loss": 1.915,
+      "step": 8390
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.0594321489334106,
+      "learning_rate": 3.966731823206179e-05,
+      "loss": 2.0922,
+      "step": 8395
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3403851389884949,
+      "learning_rate": 3.965592989444484e-05,
+      "loss": 1.8441,
+      "step": 8400
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.611869215965271,
+      "learning_rate": 3.964453692117136e-05,
+      "loss": 2.1253,
+      "step": 8405
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7241839170455933,
+      "learning_rate": 3.963313931584493e-05,
+      "loss": 1.7705,
+      "step": 8410
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.3061637878417969,
+      "learning_rate": 3.9621737082070616e-05,
+      "loss": 1.9305,
+      "step": 8415
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.108342170715332,
+      "learning_rate": 3.9610330223454926e-05,
+      "loss": 1.8427,
+      "step": 8420
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5697190761566162,
+      "learning_rate": 3.959891874360584e-05,
+      "loss": 1.7063,
+      "step": 8425
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.7562960982322693,
+      "learning_rate": 3.9587502646132804e-05,
+      "loss": 2.0017,
+      "step": 8430
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.1858628988265991,
+      "learning_rate": 3.95760819346467e-05,
+      "loss": 1.8896,
+      "step": 8435
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5219876766204834,
+      "learning_rate": 3.95646566127599e-05,
+      "loss": 1.8532,
+      "step": 8440
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.36081287264823914,
+      "learning_rate": 3.955322668408623e-05,
+      "loss": 1.7424,
+      "step": 8445
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.1630758047103882,
+      "learning_rate": 3.9541792152240965e-05,
+      "loss": 1.6692,
+      "step": 8450
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.41798633337020874,
+      "learning_rate": 3.953035302084082e-05,
+      "loss": 1.8561,
+      "step": 8455
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.932482898235321,
+      "learning_rate": 3.9518909293503994e-05,
+      "loss": 1.9695,
+      "step": 8460
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6918326616287231,
+      "learning_rate": 3.950746097385012e-05,
+      "loss": 1.8665,
+      "step": 8465
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8644784688949585,
+      "learning_rate": 3.9496008065500303e-05,
+      "loss": 1.9897,
+      "step": 8470
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.774658203125,
+      "learning_rate": 3.9484550572077075e-05,
+      "loss": 1.7163,
+      "step": 8475
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.9765104651451111,
+      "learning_rate": 3.9473088497204424e-05,
+      "loss": 1.848,
+      "step": 8480
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5454521775245667,
+      "learning_rate": 3.946162184450781e-05,
+      "loss": 1.8886,
+      "step": 8485
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.514015793800354,
+      "learning_rate": 3.945015061761412e-05,
+      "loss": 2.0871,
+      "step": 8490
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.8409627079963684,
+      "learning_rate": 3.943867482015169e-05,
+      "loss": 1.8131,
+      "step": 8495
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.5961334705352783,
+      "learning_rate": 3.94271944557503e-05,
+      "loss": 1.9685,
+      "step": 8500
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.0173949003219604,
+      "learning_rate": 3.941570952804118e-05,
+      "loss": 1.8924,
+      "step": 8505
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6633391380310059,
+      "learning_rate": 3.9404220040657e-05,
+      "loss": 1.8846,
+      "step": 8510
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 12.116934776306152,
+      "learning_rate": 3.9392725997231874e-05,
+      "loss": 1.9804,
+      "step": 8515
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.48671266436576843,
+      "learning_rate": 3.938122740140137e-05,
+      "loss": 1.921,
+      "step": 8520
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.747097373008728,
+      "learning_rate": 3.936972425680246e-05,
+      "loss": 1.962,
+      "step": 8525
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5572634935379028,
+      "learning_rate": 3.935821656707359e-05,
+      "loss": 1.6872,
+      "step": 8530
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.41499078273773193,
+      "learning_rate": 3.934670433585464e-05,
+      "loss": 1.7176,
+      "step": 8535
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.8987358212471008,
+      "learning_rate": 3.93351875667869e-05,
+      "loss": 2.0023,
+      "step": 8540
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6154505610466003,
+      "learning_rate": 3.932366626351313e-05,
+      "loss": 2.1713,
+      "step": 8545
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.38522958755493164,
+      "learning_rate": 3.931214042967749e-05,
+      "loss": 1.9133,
+      "step": 8550
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.7890143990516663,
+      "learning_rate": 3.93006100689256e-05,
+      "loss": 2.0062,
+      "step": 8555
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.5682582855224609,
+      "learning_rate": 3.92890751849045e-05,
+      "loss": 1.9966,
+      "step": 8560
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.6989255547523499,
+      "learning_rate": 3.9277535781262676e-05,
+      "loss": 1.6446,
+      "step": 8565
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.769821286201477,
+      "learning_rate": 3.926599186165001e-05,
+      "loss": 1.9349,
+      "step": 8570
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 0.46688029170036316,
+      "learning_rate": 3.925444342971783e-05,
+      "loss": 2.1585,
+      "step": 8575
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.609228789806366,
+      "learning_rate": 3.9242890489118914e-05,
+      "loss": 1.8908,
+      "step": 8580
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7954393625259399,
+      "learning_rate": 3.9231333043507436e-05,
+      "loss": 1.9862,
+      "step": 8585
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.45553261041641235,
+      "learning_rate": 3.9219771096539e-05,
+      "loss": 2.0167,
+      "step": 8590
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.3282541334629059,
+      "learning_rate": 3.9208204651870653e-05,
+      "loss": 2.0051,
+      "step": 8595
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.7702216506004333,
+      "learning_rate": 3.9196633713160834e-05,
+      "loss": 1.8989,
+      "step": 8600
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.1692415475845337,
+      "learning_rate": 3.918505828406942e-05,
+      "loss": 1.9096,
+      "step": 8605
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.8676744699478149,
+      "learning_rate": 3.917347836825773e-05,
+      "loss": 1.9068,
+      "step": 8610
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6879375576972961,
+      "learning_rate": 3.9161893969388465e-05,
+      "loss": 1.8098,
+      "step": 8615
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6627622842788696,
+      "learning_rate": 3.9150305091125746e-05,
+      "loss": 1.7859,
+      "step": 8620
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 3.1366701126098633,
+      "learning_rate": 3.913871173713515e-05,
+      "loss": 2.2024,
+      "step": 8625
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.37004053592681885,
+      "learning_rate": 3.912711391108362e-05,
+      "loss": 1.688,
+      "step": 8630
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.0141103267669678,
+      "learning_rate": 3.911551161663956e-05,
+      "loss": 2.1577,
+      "step": 8635
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5730118155479431,
+      "learning_rate": 3.9103904857472745e-05,
+      "loss": 1.8459,
+      "step": 8640
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.5071932077407837,
+      "learning_rate": 3.9092293637254395e-05,
+      "loss": 2.0975,
+      "step": 8645
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.376279354095459,
+      "learning_rate": 3.908067795965712e-05,
+      "loss": 2.0358,
+      "step": 8650
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.1163078546524048,
+      "learning_rate": 3.9069057828354947e-05,
+      "loss": 2.0103,
+      "step": 8655
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7586190700531006,
+      "learning_rate": 3.905743324702332e-05,
+      "loss": 2.0341,
+      "step": 8660
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7790800333023071,
+      "learning_rate": 3.9045804219339076e-05,
+      "loss": 1.836,
+      "step": 8665
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8625195622444153,
+      "learning_rate": 3.903417074898047e-05,
+      "loss": 2.0557,
+      "step": 8670
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.7535802721977234,
+      "learning_rate": 3.902253283962716e-05,
+      "loss": 1.9474,
+      "step": 8675
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.9609790444374084,
+      "learning_rate": 3.9010890494960195e-05,
+      "loss": 1.8005,
+      "step": 8680
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 0.8396172523498535,
+      "learning_rate": 3.8999243718662036e-05,
+      "loss": 1.8434,
+      "step": 8685
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6551474332809448,
+      "learning_rate": 3.8987592514416564e-05,
+      "loss": 1.8392,
+      "step": 8690
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.9734852313995361,
+      "learning_rate": 3.8975936885909035e-05,
+      "loss": 1.8953,
+      "step": 8695
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.37571898102760315,
+      "learning_rate": 3.896427683682611e-05,
+      "loss": 1.7285,
+      "step": 8700
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8209627866744995,
+      "learning_rate": 3.895261237085585e-05,
+      "loss": 1.833,
+      "step": 8705
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.3994194269180298,
+      "learning_rate": 3.894094349168772e-05,
+      "loss": 2.1348,
+      "step": 8710
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.766764223575592,
+      "learning_rate": 3.892927020301257e-05,
+      "loss": 1.9434,
+      "step": 8715
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7344279885292053,
+      "learning_rate": 3.891759250852266e-05,
+      "loss": 1.9841,
+      "step": 8720
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8418555855751038,
+      "learning_rate": 3.8905910411911625e-05,
+      "loss": 1.9083,
+      "step": 8725
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.784453272819519,
+      "learning_rate": 3.8894223916874494e-05,
+      "loss": 2.1818,
+      "step": 8730
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.8457757830619812,
+      "learning_rate": 3.8882533027107713e-05,
+      "loss": 2.121,
+      "step": 8735
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.5709465742111206,
+      "learning_rate": 3.887083774630908e-05,
+      "loss": 1.8507,
+      "step": 8740
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.5889132022857666,
+      "learning_rate": 3.885913807817781e-05,
+      "loss": 1.7544,
+      "step": 8745
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.483522653579712,
+      "learning_rate": 3.8847434026414516e-05,
+      "loss": 2.0034,
+      "step": 8750
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6778923869132996,
+      "learning_rate": 3.883572559472114e-05,
+      "loss": 1.7116,
+      "step": 8755
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.8211702704429626,
+      "learning_rate": 3.882401278680107e-05,
+      "loss": 1.7848,
+      "step": 8760
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.0367469787597656,
+      "learning_rate": 3.881229560635905e-05,
+      "loss": 1.9696,
+      "step": 8765
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.47565755248069763,
+      "learning_rate": 3.8800574057101227e-05,
+      "loss": 1.8395,
+      "step": 8770
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 5.1431565284729,
+      "learning_rate": 3.878884814273509e-05,
+      "loss": 1.5831,
+      "step": 8775
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6216502785682678,
+      "learning_rate": 3.877711786696956e-05,
+      "loss": 2.0567,
+      "step": 8780
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 1.0751134157180786,
+      "learning_rate": 3.8765383233514895e-05,
+      "loss": 1.9522,
+      "step": 8785
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7744889259338379,
+      "learning_rate": 3.875364424608275e-05,
+      "loss": 1.8294,
+      "step": 8790
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7963035106658936,
+      "learning_rate": 3.874190090838616e-05,
+      "loss": 2.0646,
+      "step": 8795
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6751787662506104,
+      "learning_rate": 3.873015322413954e-05,
+      "loss": 1.864,
+      "step": 8800
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.275324821472168,
+      "learning_rate": 3.871840119705866e-05,
+      "loss": 1.6613,
+      "step": 8805
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 11.015603065490723,
+      "learning_rate": 3.870664483086067e-05,
+      "loss": 1.9226,
+      "step": 8810
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6475034952163696,
+      "learning_rate": 3.869488412926411e-05,
+      "loss": 1.6963,
+      "step": 8815
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.9630057215690613,
+      "learning_rate": 3.8683119095988865e-05,
+      "loss": 2.0588,
+      "step": 8820
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.8182287812232971,
+      "learning_rate": 3.867134973475622e-05,
+      "loss": 2.4044,
+      "step": 8825
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.1581512689590454,
+      "learning_rate": 3.8659576049288806e-05,
+      "loss": 1.6287,
+      "step": 8830
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.7855355143547058,
+      "learning_rate": 3.864779804331061e-05,
+      "loss": 1.6889,
+      "step": 8835
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5677868723869324,
+      "learning_rate": 3.863601572054704e-05,
+      "loss": 1.8155,
+      "step": 8840
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.6653741002082825,
+      "learning_rate": 3.8624229084724804e-05,
+      "loss": 2.2333,
+      "step": 8845
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.5179418325424194,
+      "learning_rate": 3.861243813957201e-05,
+      "loss": 1.884,
+      "step": 8850
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.8245850801467896,
+      "learning_rate": 3.8600642888818125e-05,
+      "loss": 1.9703,
+      "step": 8855
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5349978804588318,
+      "learning_rate": 3.8588843336193994e-05,
+      "loss": 2.0805,
+      "step": 8860
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.5746064782142639,
+      "learning_rate": 3.857703948543176e-05,
+      "loss": 1.5872,
+      "step": 8865
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7160893082618713,
+      "learning_rate": 3.856523134026502e-05,
+      "loss": 1.8665,
+      "step": 8870
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.8419469594955444,
+      "learning_rate": 3.8553418904428644e-05,
+      "loss": 1.4556,
+      "step": 8875
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6989808678627014,
+      "learning_rate": 3.854160218165891e-05,
+      "loss": 1.6613,
+      "step": 8880
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.752234697341919,
+      "learning_rate": 3.8529781175693423e-05,
+      "loss": 1.8097,
+      "step": 8885
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7804884910583496,
+      "learning_rate": 3.851795589027117e-05,
+      "loss": 1.7908,
+      "step": 8890
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6645786762237549,
+      "learning_rate": 3.850612632913247e-05,
+      "loss": 2.0249,
+      "step": 8895
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.454006165266037,
+      "learning_rate": 3.849429249601901e-05,
+      "loss": 1.7901,
+      "step": 8900
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7006505131721497,
+      "learning_rate": 3.8482454394673806e-05,
+      "loss": 1.9092,
+      "step": 8905
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.7154279947280884,
+      "learning_rate": 3.8470612028841245e-05,
+      "loss": 1.8582,
+      "step": 8910
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.37950852513313293,
+      "learning_rate": 3.845876540226706e-05,
+      "loss": 1.9758,
+      "step": 8915
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5836817622184753,
+      "learning_rate": 3.844691451869832e-05,
+      "loss": 1.7204,
+      "step": 8920
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.4787684679031372,
+      "learning_rate": 3.843505938188346e-05,
+      "loss": 1.781,
+      "step": 8925
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5430890917778015,
+      "learning_rate": 3.8423199995572235e-05,
+      "loss": 2.0103,
+      "step": 8930
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.5893062353134155,
+      "learning_rate": 3.841133636351576e-05,
+      "loss": 1.9518,
+      "step": 8935
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7547013163566589,
+      "learning_rate": 3.83994684894665e-05,
+      "loss": 1.7754,
+      "step": 8940
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.5084164142608643,
+      "learning_rate": 3.8387596377178236e-05,
+      "loss": 1.6288,
+      "step": 8945
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.533129870891571,
+      "learning_rate": 3.837572003040612e-05,
+      "loss": 1.7496,
+      "step": 8950
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6257514357566833,
+      "learning_rate": 3.8363839452906627e-05,
+      "loss": 1.8528,
+      "step": 8955
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.8380822539329529,
+      "learning_rate": 3.835195464843757e-05,
+      "loss": 1.8115,
+      "step": 8960
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.9693661332130432,
+      "learning_rate": 3.834006562075809e-05,
+      "loss": 1.6759,
+      "step": 8965
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6388950943946838,
+      "learning_rate": 3.83281723736287e-05,
+      "loss": 1.7906,
+      "step": 8970
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.9138167500495911,
+      "learning_rate": 3.83162749108112e-05,
+      "loss": 1.7656,
+      "step": 8975
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5847556591033936,
+      "learning_rate": 3.830437323606876e-05,
+      "loss": 1.7698,
+      "step": 8980
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.6280438899993896,
+      "learning_rate": 3.8292467353165864e-05,
+      "loss": 2.0533,
+      "step": 8985
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 9.531767845153809,
+      "learning_rate": 3.828055726586832e-05,
+      "loss": 2.0169,
+      "step": 8990
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.7199810743331909,
+      "learning_rate": 3.826864297794329e-05,
+      "loss": 1.8667,
+      "step": 8995
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.5033280849456787,
+      "learning_rate": 3.8256724493159246e-05,
+      "loss": 1.6972,
+      "step": 9000
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.0077893733978271,
+      "learning_rate": 3.8244801815286e-05,
+      "loss": 1.8997,
+      "step": 9005
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8881186246871948,
+      "learning_rate": 3.823287494809469e-05,
+      "loss": 2.0588,
+      "step": 9010
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.8047612309455872,
+      "learning_rate": 3.822094389535775e-05,
+      "loss": 2.0178,
+      "step": 9015
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 0.4620480239391327,
+      "learning_rate": 3.8209008660848974e-05,
+      "loss": 2.0099,
+      "step": 9020
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.5921469926834106,
+      "learning_rate": 3.819706924834346e-05,
+      "loss": 1.8967,
+      "step": 9025
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6395788192749023,
+      "learning_rate": 3.818512566161765e-05,
+      "loss": 1.7851,
+      "step": 9030
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7558723092079163,
+      "learning_rate": 3.8173177904449265e-05,
+      "loss": 1.6899,
+      "step": 9035
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.2595165967941284,
+      "learning_rate": 3.816122598061739e-05,
+      "loss": 1.9932,
+      "step": 9040
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.8348121643066406,
+      "learning_rate": 3.814926989390238e-05,
+      "loss": 1.7467,
+      "step": 9045
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.8437023162841797,
+      "learning_rate": 3.8137309648085965e-05,
+      "loss": 1.5459,
+      "step": 9050
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6804198622703552,
+      "learning_rate": 3.8125345246951137e-05,
+      "loss": 1.9587,
+      "step": 9055
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.2239028215408325,
+      "learning_rate": 3.8113376694282245e-05,
+      "loss": 1.8473,
+      "step": 9060
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.9916425943374634,
+      "learning_rate": 3.8101403993864905e-05,
+      "loss": 1.7582,
+      "step": 9065
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7374111413955688,
+      "learning_rate": 3.808942714948609e-05,
+      "loss": 1.8658,
+      "step": 9070
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.6606670022010803,
+      "learning_rate": 3.807744616493405e-05,
+      "loss": 1.7126,
+      "step": 9075
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.7266793847084045,
+      "learning_rate": 3.806546104399837e-05,
+      "loss": 2.1215,
+      "step": 9080
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6009851098060608,
+      "learning_rate": 3.8053471790469935e-05,
+      "loss": 1.691,
+      "step": 9085
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6415896415710449,
+      "learning_rate": 3.8041478408140926e-05,
+      "loss": 1.4786,
+      "step": 9090
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6692395806312561,
+      "learning_rate": 3.802948090080484e-05,
+      "loss": 1.8868,
+      "step": 9095
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.1220877170562744,
+      "learning_rate": 3.8017479272256476e-05,
+      "loss": 1.8618,
+      "step": 9100
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.643251895904541,
+      "learning_rate": 3.800547352629195e-05,
+      "loss": 1.9866,
+      "step": 9105
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.6525068283081055,
+      "learning_rate": 3.799346366670864e-05,
+      "loss": 1.7077,
+      "step": 9110
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.711988091468811,
+      "learning_rate": 3.798144969730528e-05,
+      "loss": 1.7389,
+      "step": 9115
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.5814893841743469,
+      "learning_rate": 3.796943162188186e-05,
+      "loss": 2.0051,
+      "step": 9120
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.351395845413208,
+      "learning_rate": 3.79574094442397e-05,
+      "loss": 2.038,
+      "step": 9125
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.42251327633857727,
+      "learning_rate": 3.7945383168181405e-05,
+      "loss": 1.7681,
+      "step": 9130
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.0346791744232178,
+      "learning_rate": 3.793335279751085e-05,
+      "loss": 2.0539,
+      "step": 9135
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7705110311508179,
+      "learning_rate": 3.792131833603325e-05,
+      "loss": 2.0802,
+      "step": 9140
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.7287485003471375,
+      "learning_rate": 3.790927978755509e-05,
+      "loss": 1.7397,
+      "step": 9145
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5948812365531921,
+      "learning_rate": 3.789723715588416e-05,
+      "loss": 1.8581,
+      "step": 9150
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.4935462772846222,
+      "learning_rate": 3.7885190444829503e-05,
+      "loss": 1.7997,
+      "step": 9155
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.9329087138175964,
+      "learning_rate": 3.7873139658201516e-05,
+      "loss": 2.0113,
+      "step": 9160
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.4054046869277954,
+      "learning_rate": 3.786108479981183e-05,
+      "loss": 2.0666,
+      "step": 9165
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6794940829277039,
+      "learning_rate": 3.7849025873473404e-05,
+      "loss": 2.0231,
+      "step": 9170
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.5740146636962891,
+      "learning_rate": 3.7836962883000446e-05,
+      "loss": 1.82,
+      "step": 9175
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6178922057151794,
+      "learning_rate": 3.782489583220847e-05,
+      "loss": 2.0086,
+      "step": 9180
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6487544775009155,
+      "learning_rate": 3.781282472491429e-05,
+      "loss": 1.7158,
+      "step": 9185
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 0.6807827353477478,
+      "learning_rate": 3.780074956493597e-05,
+      "loss": 2.0595,
+      "step": 9190
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.687058687210083,
+      "learning_rate": 3.778867035609288e-05,
+      "loss": 1.5461,
+      "step": 9195
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5961658954620361,
+      "learning_rate": 3.777658710220564e-05,
+      "loss": 2.0018,
+      "step": 9200
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5087503790855408,
+      "learning_rate": 3.776449980709621e-05,
+      "loss": 2.0864,
+      "step": 9205
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8616182804107666,
+      "learning_rate": 3.775240847458775e-05,
+      "loss": 1.7007,
+      "step": 9210
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.5542956590652466,
+      "learning_rate": 3.774031310850477e-05,
+      "loss": 2.0111,
+      "step": 9215
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.8051539063453674,
+      "learning_rate": 3.772821371267301e-05,
+      "loss": 1.7596,
+      "step": 9220
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.0442497730255127,
+      "learning_rate": 3.771611029091948e-05,
+      "loss": 1.6548,
+      "step": 9225
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7157022953033447,
+      "learning_rate": 3.77040028470725e-05,
+      "loss": 1.5846,
+      "step": 9230
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.660575807094574,
+      "learning_rate": 3.7691891384961654e-05,
+      "loss": 2.0322,
+      "step": 9235
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.2163981199264526,
+      "learning_rate": 3.767977590841776e-05,
+      "loss": 2.1145,
+      "step": 9240
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 0.7265014052391052,
+      "learning_rate": 3.7667656421272946e-05,
+      "loss": 1.7375,
+      "step": 9245
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.1795376539230347,
+      "learning_rate": 3.765553292736059e-05,
+      "loss": 1.9286,
+      "step": 9250
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5306074023246765,
+      "learning_rate": 3.7643405430515344e-05,
+      "loss": 1.9642,
+      "step": 9255
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6961469650268555,
+      "learning_rate": 3.763127393457311e-05,
+      "loss": 1.9714,
+      "step": 9260
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.5733311772346497,
+      "learning_rate": 3.76191384433711e-05,
+      "loss": 1.6244,
+      "step": 9265
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6173118352890015,
+      "learning_rate": 3.7606998960747726e-05,
+      "loss": 1.9729,
+      "step": 9270
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7085820436477661,
+      "learning_rate": 3.7594855490542704e-05,
+      "loss": 1.8824,
+      "step": 9275
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.8288882970809937,
+      "learning_rate": 3.758270803659701e-05,
+      "loss": 1.7626,
+      "step": 9280
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.7774258255958557,
+      "learning_rate": 3.757055660275286e-05,
+      "loss": 1.7774,
+      "step": 9285
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 1.5239425897598267,
+      "learning_rate": 3.755840119285377e-05,
+      "loss": 2.165,
+      "step": 9290
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.4820414185523987,
+      "learning_rate": 3.7546241810744445e-05,
+      "loss": 2.2087,
+      "step": 9295
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6915215849876404,
+      "learning_rate": 3.75340784602709e-05,
+      "loss": 1.6879,
+      "step": 9300
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5616482496261597,
+      "learning_rate": 3.7521911145280395e-05,
+      "loss": 2.1273,
+      "step": 9305
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.9838537573814392,
+      "learning_rate": 3.750973986962145e-05,
+      "loss": 1.7339,
+      "step": 9310
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5095938444137573,
+      "learning_rate": 3.7497564637143814e-05,
+      "loss": 1.8669,
+      "step": 9315
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5213245153427124,
+      "learning_rate": 3.74853854516985e-05,
+      "loss": 1.8368,
+      "step": 9320
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6392638087272644,
+      "learning_rate": 3.747320231713778e-05,
+      "loss": 1.8477,
+      "step": 9325
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 8.093245506286621,
+      "learning_rate": 3.746101523731515e-05,
+      "loss": 1.8944,
+      "step": 9330
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5117234587669373,
+      "learning_rate": 3.744882421608541e-05,
+      "loss": 1.73,
+      "step": 9335
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5771470665931702,
+      "learning_rate": 3.743662925730453e-05,
+      "loss": 2.0571,
+      "step": 9340
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.6059357523918152,
+      "learning_rate": 3.7424430364829785e-05,
+      "loss": 1.735,
+      "step": 9345
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.48577621579170227,
+      "learning_rate": 3.7412227542519664e-05,
+      "loss": 1.8274,
+      "step": 9350
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.7163307070732117,
+      "learning_rate": 3.740002079423392e-05,
+      "loss": 2.0151,
+      "step": 9355
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6120556592941284,
+      "learning_rate": 3.738781012383351e-05,
+      "loss": 1.8655,
+      "step": 9360
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6820749640464783,
+      "learning_rate": 3.737559553518067e-05,
+      "loss": 1.8406,
+      "step": 9365
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6751840114593506,
+      "learning_rate": 3.736337703213888e-05,
+      "loss": 1.8045,
+      "step": 9370
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5991414785385132,
+      "learning_rate": 3.735115461857282e-05,
+      "loss": 1.7511,
+      "step": 9375
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6772866249084473,
+      "learning_rate": 3.733892829834845e-05,
+      "loss": 2.0622,
+      "step": 9380
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.37456727027893066,
+      "learning_rate": 3.732669807533291e-05,
+      "loss": 1.7938,
+      "step": 9385
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5582310557365417,
+      "learning_rate": 3.731446395339463e-05,
+      "loss": 1.7566,
+      "step": 9390
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.5218795537948608,
+      "learning_rate": 3.730222593640324e-05,
+      "loss": 1.6361,
+      "step": 9395
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6812488436698914,
+      "learning_rate": 3.7289984028229636e-05,
+      "loss": 1.9274,
+      "step": 9400
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.6946431398391724,
+      "learning_rate": 3.72777382327459e-05,
+      "loss": 1.8467,
+      "step": 9405
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.7319812774658203,
+      "learning_rate": 3.726548855382536e-05,
+      "loss": 1.696,
+      "step": 9410
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6396403312683105,
+      "learning_rate": 3.725323499534259e-05,
+      "loss": 1.8228,
+      "step": 9415
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.4877493977546692,
+      "learning_rate": 3.724097756117338e-05,
+      "loss": 2.1406,
+      "step": 9420
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5904247164726257,
+      "learning_rate": 3.722871625519475e-05,
+      "loss": 1.9487,
+      "step": 9425
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6823410391807556,
+      "learning_rate": 3.721645108128493e-05,
+      "loss": 1.7107,
+      "step": 9430
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.688463568687439,
+      "learning_rate": 3.720418204332338e-05,
+      "loss": 1.7925,
+      "step": 9435
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.0077917575836182,
+      "learning_rate": 3.719190914519078e-05,
+      "loss": 1.8868,
+      "step": 9440
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.1595532894134521,
+      "learning_rate": 3.7179632390769055e-05,
+      "loss": 1.6869,
+      "step": 9445
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.8011316657066345,
+      "learning_rate": 3.716735178394131e-05,
+      "loss": 1.7987,
+      "step": 9450
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5646600723266602,
+      "learning_rate": 3.7155067328591906e-05,
+      "loss": 1.8537,
+      "step": 9455
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.5149386525154114,
+      "learning_rate": 3.7142779028606405e-05,
+      "loss": 1.8974,
+      "step": 9460
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.7945959568023682,
+      "learning_rate": 3.713048688787157e-05,
+      "loss": 1.6476,
+      "step": 9465
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.4259641766548157,
+      "learning_rate": 3.7118190910275394e-05,
+      "loss": 1.8111,
+      "step": 9470
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5765038728713989,
+      "learning_rate": 3.71058910997071e-05,
+      "loss": 1.571,
+      "step": 9475
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6007339954376221,
+      "learning_rate": 3.7093587460057087e-05,
+      "loss": 1.8102,
+      "step": 9480
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.3550678491592407,
+      "learning_rate": 3.708127999521701e-05,
+      "loss": 1.8255,
+      "step": 9485
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.5561153888702393,
+      "learning_rate": 3.706896870907967e-05,
+      "loss": 1.8103,
+      "step": 9490
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.6592022180557251,
+      "learning_rate": 3.705665360553914e-05,
+      "loss": 1.7564,
+      "step": 9495
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.9379505515098572,
+      "learning_rate": 3.704433468849068e-05,
+      "loss": 1.9543,
+      "step": 9500
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.7542968392372131,
+      "learning_rate": 3.7032011961830744e-05,
+      "loss": 1.9984,
+      "step": 9505
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.1859865188598633,
+      "learning_rate": 3.7019685429456986e-05,
+      "loss": 1.8548,
+      "step": 9510
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.0546746253967285,
+      "learning_rate": 3.70073550952683e-05,
+      "loss": 1.6335,
+      "step": 9515
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.41656044125556946,
+      "learning_rate": 3.699502096316474e-05,
+      "loss": 1.817,
+      "step": 9520
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.3669623136520386,
+      "learning_rate": 3.698268303704759e-05,
+      "loss": 1.8767,
+      "step": 9525
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5422895550727844,
+      "learning_rate": 3.69728099670865e-05,
+      "loss": 2.1527,
+      "step": 9530
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6797967553138733,
+      "learning_rate": 3.6960465221579927e-05,
+      "loss": 1.7675,
+      "step": 9535
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.9210271239280701,
+      "learning_rate": 3.69481166929897e-05,
+      "loss": 1.6502,
+      "step": 9540
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8029117584228516,
+      "learning_rate": 3.693576438522167e-05,
+      "loss": 1.8089,
+      "step": 9545
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.47192418575286865,
+      "learning_rate": 3.6923408302182834e-05,
+      "loss": 2.0336,
+      "step": 9550
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.66648268699646,
+      "learning_rate": 3.691104844778142e-05,
+      "loss": 1.7003,
+      "step": 9555
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.6902146935462952,
+      "learning_rate": 3.689868482592684e-05,
+      "loss": 1.8137,
+      "step": 9560
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8713681697845459,
+      "learning_rate": 3.688631744052969e-05,
+      "loss": 2.1054,
+      "step": 9565
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8756383657455444,
+      "learning_rate": 3.687394629550178e-05,
+      "loss": 2.0403,
+      "step": 9570
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.8325067758560181,
+      "learning_rate": 3.6861571394756066e-05,
+      "loss": 1.8843,
+      "step": 9575
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.5879186987876892,
+      "learning_rate": 3.684919274220674e-05,
+      "loss": 1.9288,
+      "step": 9580
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5836779475212097,
+      "learning_rate": 3.683681034176914e-05,
+      "loss": 2.024,
+      "step": 9585
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7767724990844727,
+      "learning_rate": 3.682442419735982e-05,
+      "loss": 2.3006,
+      "step": 9590
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.923039972782135,
+      "learning_rate": 3.681203431289649e-05,
+      "loss": 1.9507,
+      "step": 9595
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.46944764256477356,
+      "learning_rate": 3.6799640692298075e-05,
+      "loss": 1.9669,
+      "step": 9600
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.29924699664115906,
+      "learning_rate": 3.678724333948466e-05,
+      "loss": 1.8226,
+      "step": 9605
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.4777134954929352,
+      "learning_rate": 3.6774842258377515e-05,
+      "loss": 2.3745,
+      "step": 9610
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8692833781242371,
+      "learning_rate": 3.676243745289909e-05,
+      "loss": 1.9515,
+      "step": 9615
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.9050225019454956,
+      "learning_rate": 3.675002892697301e-05,
+      "loss": 1.9985,
+      "step": 9620
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6240570545196533,
+      "learning_rate": 3.673761668452408e-05,
+      "loss": 1.8733,
+      "step": 9625
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.6922764778137207,
+      "learning_rate": 3.6725200729478285e-05,
+      "loss": 1.6951,
+      "step": 9630
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.4850800931453705,
+      "learning_rate": 3.671278106576278e-05,
+      "loss": 1.9839,
+      "step": 9635
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7302365303039551,
+      "learning_rate": 3.670035769730589e-05,
+      "loss": 2.0729,
+      "step": 9640
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.43896591663360596,
+      "learning_rate": 3.668793062803711e-05,
+      "loss": 1.8726,
+      "step": 9645
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8171257972717285,
+      "learning_rate": 3.667549986188712e-05,
+      "loss": 1.9601,
+      "step": 9650
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.631952166557312,
+      "learning_rate": 3.666306540278775e-05,
+      "loss": 1.745,
+      "step": 9655
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.7533524036407471,
+      "learning_rate": 3.665062725467201e-05,
+      "loss": 1.9608,
+      "step": 9660
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5362209677696228,
+      "learning_rate": 3.6638185421474084e-05,
+      "loss": 1.5627,
+      "step": 9665
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.6645787358283997,
+      "learning_rate": 3.6625739907129295e-05,
+      "loss": 1.5983,
+      "step": 9670
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.5133046507835388,
+      "learning_rate": 3.6613290715574155e-05,
+      "loss": 1.9776,
+      "step": 9675
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.0483719110488892,
+      "learning_rate": 3.660083785074634e-05,
+      "loss": 2.0728,
+      "step": 9680
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.8712103366851807,
+      "learning_rate": 3.6588381316584674e-05,
+      "loss": 1.7619,
+      "step": 9685
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.295785665512085,
+      "learning_rate": 3.6575921117029136e-05,
+      "loss": 1.8996,
+      "step": 9690
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.5898334383964539,
+      "learning_rate": 3.6563457256020884e-05,
+      "loss": 1.8068,
+      "step": 9695
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.5830318331718445,
+      "learning_rate": 3.655098973750223e-05,
+      "loss": 1.8471,
+      "step": 9700
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6544398069381714,
+      "learning_rate": 3.653851856541662e-05,
+      "loss": 1.9984,
+      "step": 9705
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.37488725781440735,
+      "learning_rate": 3.6526043743708704e-05,
+      "loss": 1.8106,
+      "step": 9710
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.239549160003662,
+      "learning_rate": 3.651356527632423e-05,
+      "loss": 2.1417,
+      "step": 9715
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.47542452812194824,
+      "learning_rate": 3.650108316721013e-05,
+      "loss": 1.6287,
+      "step": 9720
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.43621698021888733,
+      "learning_rate": 3.648859742031449e-05,
+      "loss": 1.96,
+      "step": 9725
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.5922963619232178,
+      "learning_rate": 3.6476108039586535e-05,
+      "loss": 1.8913,
+      "step": 9730
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6242755055427551,
+      "learning_rate": 3.646361502897665e-05,
+      "loss": 1.8241,
+      "step": 9735
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.1280851364135742,
+      "learning_rate": 3.645111839243636e-05,
+      "loss": 2.1662,
+      "step": 9740
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6557629704475403,
+      "learning_rate": 3.643861813391833e-05,
+      "loss": 2.0015,
+      "step": 9745
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6168567538261414,
+      "learning_rate": 3.642611425737639e-05,
+      "loss": 1.5228,
+      "step": 9750
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.4185939133167267,
+      "learning_rate": 3.6413606766765506e-05,
+      "loss": 2.1472,
+      "step": 9755
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.49710220098495483,
+      "learning_rate": 3.6401095666041765e-05,
+      "loss": 1.9794,
+      "step": 9760
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6820828914642334,
+      "learning_rate": 3.638858095916244e-05,
+      "loss": 1.7363,
+      "step": 9765
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.7763912081718445,
+      "learning_rate": 3.637606265008592e-05,
+      "loss": 1.9053,
+      "step": 9770
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.9690907001495361,
+      "learning_rate": 3.636354074277172e-05,
+      "loss": 1.6437,
+      "step": 9775
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.8117865324020386,
+      "learning_rate": 3.635101524118051e-05,
+      "loss": 1.7684,
+      "step": 9780
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.878116250038147,
+      "learning_rate": 3.63384861492741e-05,
+      "loss": 1.9089,
+      "step": 9785
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.3111972510814667,
+      "learning_rate": 3.632595347101543e-05,
+      "loss": 2.0609,
+      "step": 9790
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.38425540924072266,
+      "learning_rate": 3.6313417210368564e-05,
+      "loss": 1.7993,
+      "step": 9795
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6652374267578125,
+      "learning_rate": 3.630087737129872e-05,
+      "loss": 2.0774,
+      "step": 9800
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.9802791476249695,
+      "learning_rate": 3.628833395777224e-05,
+      "loss": 1.8374,
+      "step": 9805
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.093260407447815,
+      "learning_rate": 3.6275786973756575e-05,
+      "loss": 1.945,
+      "step": 9810
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.5013108253479004,
+      "learning_rate": 3.626323642322035e-05,
+      "loss": 1.8702,
+      "step": 9815
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7069824934005737,
+      "learning_rate": 3.6250682310133266e-05,
+      "loss": 2.0702,
+      "step": 9820
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.539284348487854,
+      "learning_rate": 3.6238124638466195e-05,
+      "loss": 1.7797,
+      "step": 9825
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.642426609992981,
+      "learning_rate": 3.622556341219111e-05,
+      "loss": 1.8979,
+      "step": 9830
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7472493052482605,
+      "learning_rate": 3.6212998635281116e-05,
+      "loss": 1.8212,
+      "step": 9835
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.1682623624801636,
+      "learning_rate": 3.620043031171043e-05,
+      "loss": 1.7595,
+      "step": 9840
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6802049279212952,
+      "learning_rate": 3.618785844545443e-05,
+      "loss": 1.9368,
+      "step": 9845
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.0584665536880493,
+      "learning_rate": 3.617528304048955e-05,
+      "loss": 1.91,
+      "step": 9850
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6849753260612488,
+      "learning_rate": 3.616270410079339e-05,
+      "loss": 1.8772,
+      "step": 9855
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.9579459428787231,
+      "learning_rate": 3.6150121630344676e-05,
+      "loss": 1.9455,
+      "step": 9860
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8555591702461243,
+      "learning_rate": 3.61375356331232e-05,
+      "loss": 1.4362,
+      "step": 9865
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.5340035557746887,
+      "learning_rate": 3.612494611310992e-05,
+      "loss": 1.9012,
+      "step": 9870
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8525397777557373,
+      "learning_rate": 3.61123530742869e-05,
+      "loss": 1.9174,
+      "step": 9875
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.6355140209197998,
+      "learning_rate": 3.6099756520637274e-05,
+      "loss": 1.7693,
+      "step": 9880
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.794104814529419,
+      "learning_rate": 3.608715645614534e-05,
+      "loss": 1.9271,
+      "step": 9885
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.5898195505142212,
+      "learning_rate": 3.6074552884796485e-05,
+      "loss": 1.969,
+      "step": 9890
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.8139387965202332,
+      "learning_rate": 3.606194581057721e-05,
+      "loss": 1.8805,
+      "step": 9895
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7168921828269958,
+      "learning_rate": 3.60493352374751e-05,
+      "loss": 1.7252,
+      "step": 9900
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.9840201735496521,
+      "learning_rate": 3.6036721169478896e-05,
+      "loss": 1.8256,
+      "step": 9905
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.7537431120872498,
+      "learning_rate": 3.60241036105784e-05,
+      "loss": 2.0456,
+      "step": 9910
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.142112374305725,
+      "learning_rate": 3.601148256476454e-05,
+      "loss": 1.987,
+      "step": 9915
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.9110394716262817,
+      "learning_rate": 3.599885803602933e-05,
+      "loss": 1.7752,
+      "step": 9920
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6123466491699219,
+      "learning_rate": 3.59862300283659e-05,
+      "loss": 2.0613,
+      "step": 9925
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.7708951234817505,
+      "learning_rate": 3.597359854576848e-05,
+      "loss": 1.8793,
+      "step": 9930
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6696035861968994,
+      "learning_rate": 3.596096359223241e-05,
+      "loss": 1.8185,
+      "step": 9935
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.329271674156189,
+      "learning_rate": 3.594832517175408e-05,
+      "loss": 2.0871,
+      "step": 9940
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.1870595216751099,
+      "learning_rate": 3.593568328833104e-05,
+      "loss": 1.6247,
+      "step": 9945
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.8780333399772644,
+      "learning_rate": 3.59230379459619e-05,
+      "loss": 1.6585,
+      "step": 9950
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 16.52887725830078,
+      "learning_rate": 3.591038914864635e-05,
+      "loss": 2.0053,
+      "step": 9955
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.8829307556152344,
+      "learning_rate": 3.5897736900385213e-05,
+      "loss": 1.7779,
+      "step": 9960
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.539667546749115,
+      "learning_rate": 3.5885081205180377e-05,
+      "loss": 1.761,
+      "step": 9965
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.6431927680969238,
+      "learning_rate": 3.5872422067034824e-05,
+      "loss": 1.908,
+      "step": 9970
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.2214795351028442,
+      "learning_rate": 3.5859759489952635e-05,
+      "loss": 1.8322,
+      "step": 9975
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.3363548517227173,
+      "learning_rate": 3.5847093477938956e-05,
+      "loss": 1.9769,
+      "step": 9980
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6570048928260803,
+      "learning_rate": 3.583442403500005e-05,
+      "loss": 1.7981,
+      "step": 9985
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.1348395347595215,
+      "learning_rate": 3.582175116514324e-05,
+      "loss": 1.7122,
+      "step": 9990
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6550275683403015,
+      "learning_rate": 3.5809074872376943e-05,
+      "loss": 1.8589,
+      "step": 9995
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.4769278764724731,
+      "learning_rate": 3.5796395160710666e-05,
+      "loss": 1.8013,
+      "step": 10000
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.5346553921699524,
+      "learning_rate": 3.5783712034154985e-05,
+      "loss": 1.988,
+      "step": 10005
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.48149240016937256,
+      "learning_rate": 3.5771025496721556e-05,
+      "loss": 1.7784,
+      "step": 10010
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.6305123567581177,
+      "learning_rate": 3.5758335552423127e-05,
+      "loss": 2.0311,
+      "step": 10015
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.48739558458328247,
+      "learning_rate": 3.574564220527351e-05,
+      "loss": 2.1409,
+      "step": 10020
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.4256173372268677,
+      "learning_rate": 3.57329454592876e-05,
+      "loss": 1.6745,
+      "step": 10025
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.9247502088546753,
+      "learning_rate": 3.572024531848136e-05,
+      "loss": 1.7922,
+      "step": 10030
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 5.438934326171875,
+      "learning_rate": 3.5707541786871845e-05,
+      "loss": 1.6375,
+      "step": 10035
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6272771954536438,
+      "learning_rate": 3.5694834868477154e-05,
+      "loss": 1.8961,
+      "step": 10040
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6681824326515198,
+      "learning_rate": 3.5682124567316485e-05,
+      "loss": 1.716,
+      "step": 10045
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6500905156135559,
+      "learning_rate": 3.5669410887410095e-05,
+      "loss": 1.8117,
+      "step": 10050
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7739772796630859,
+      "learning_rate": 3.5656693832779295e-05,
+      "loss": 1.7639,
+      "step": 10055
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.4250242710113525,
+      "learning_rate": 3.5643973407446496e-05,
+      "loss": 1.9421,
+      "step": 10060
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.9519471526145935,
+      "learning_rate": 3.5631249615435145e-05,
+      "loss": 1.9734,
+      "step": 10065
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.9607334733009338,
+      "learning_rate": 3.561852246076976e-05,
+      "loss": 1.7387,
+      "step": 10070
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.9569016695022583,
+      "learning_rate": 3.560579194747593e-05,
+      "loss": 1.6301,
+      "step": 10075
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.48008155822753906,
+      "learning_rate": 3.559305807958032e-05,
+      "loss": 1.9027,
+      "step": 10080
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.8056952953338623,
+      "learning_rate": 3.5580320861110625e-05,
+      "loss": 1.9209,
+      "step": 10085
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.4301571547985077,
+      "learning_rate": 3.556758029609561e-05,
+      "loss": 1.9884,
+      "step": 10090
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.0041254758834839,
+      "learning_rate": 3.555483638856513e-05,
+      "loss": 1.6986,
+      "step": 10095
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7964014410972595,
+      "learning_rate": 3.554208914255004e-05,
+      "loss": 1.7893,
+      "step": 10100
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7074928283691406,
+      "learning_rate": 3.552933856208229e-05,
+      "loss": 1.8304,
+      "step": 10105
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.151492953300476,
+      "learning_rate": 3.55165846511949e-05,
+      "loss": 1.6251,
+      "step": 10110
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.4866889417171478,
+      "learning_rate": 3.5503827413921885e-05,
+      "loss": 1.7865,
+      "step": 10115
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7697849273681641,
+      "learning_rate": 3.549106685429836e-05,
+      "loss": 1.669,
+      "step": 10120
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.7082263827323914,
+      "learning_rate": 3.5478302976360486e-05,
+      "loss": 1.9564,
+      "step": 10125
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.869172215461731,
+      "learning_rate": 3.5465535784145464e-05,
+      "loss": 2.0057,
+      "step": 10130
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 0.6560725569725037,
+      "learning_rate": 3.545276528169154e-05,
+      "loss": 1.9155,
+      "step": 10135
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.027157187461853,
+      "learning_rate": 3.5439991473038006e-05,
+      "loss": 1.6909,
+      "step": 10140
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7855679392814636,
+      "learning_rate": 3.542721436222521e-05,
+      "loss": 1.8473,
+      "step": 10145
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.6946498155593872,
+      "learning_rate": 3.541443395329454e-05,
+      "loss": 2.1364,
+      "step": 10150
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.001931071281433,
+      "learning_rate": 3.540165025028844e-05,
+      "loss": 1.8402,
+      "step": 10155
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.9406866431236267,
+      "learning_rate": 3.538886325725036e-05,
+      "loss": 1.9564,
+      "step": 10160
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.3592713177204132,
+      "learning_rate": 3.5376072978224824e-05,
+      "loss": 1.5525,
+      "step": 10165
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 1.2049317359924316,
+      "learning_rate": 3.536327941725739e-05,
+      "loss": 1.9862,
+      "step": 10170
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.8280571699142456,
+      "learning_rate": 3.5350482578394636e-05,
+      "loss": 1.745,
+      "step": 10175
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7504371404647827,
+      "learning_rate": 3.533768246568421e-05,
+      "loss": 1.9558,
+      "step": 10180
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.9918588995933533,
+      "learning_rate": 3.532487908317476e-05,
+      "loss": 1.979,
+      "step": 10185
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.46797382831573486,
+      "learning_rate": 3.5312072434915986e-05,
+      "loss": 1.99,
+      "step": 10190
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.5668039917945862,
+      "learning_rate": 3.5299262524958615e-05,
+      "loss": 1.7308,
+      "step": 10195
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.981918454170227,
+      "learning_rate": 3.528644935735442e-05,
+      "loss": 1.7743,
+      "step": 10200
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.9254132509231567,
+      "learning_rate": 3.527363293615619e-05,
+      "loss": 2.0246,
+      "step": 10205
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.9930950403213501,
+      "learning_rate": 3.5260813265417735e-05,
+      "loss": 1.9832,
+      "step": 10210
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7647066116333008,
+      "learning_rate": 3.524799034919392e-05,
+      "loss": 2.1351,
+      "step": 10215
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.070019245147705,
+      "learning_rate": 3.523516419154062e-05,
+      "loss": 1.942,
+      "step": 10220
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.79027259349823,
+      "learning_rate": 3.522233479651472e-05,
+      "loss": 1.9146,
+      "step": 10225
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.4503646194934845,
+      "learning_rate": 3.520950216817417e-05,
+      "loss": 2.0948,
+      "step": 10230
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7014816999435425,
+      "learning_rate": 3.519666631057789e-05,
+      "loss": 1.7254,
+      "step": 10235
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.0421819686889648,
+      "learning_rate": 3.518382722778587e-05,
+      "loss": 1.8633,
+      "step": 10240
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.7871792316436768,
+      "learning_rate": 3.5170984923859095e-05,
+      "loss": 1.7731,
+      "step": 10245
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.5907774567604065,
+      "learning_rate": 3.515813940285957e-05,
+      "loss": 1.9179,
+      "step": 10250
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.4600268304347992,
+      "learning_rate": 3.514529066885032e-05,
+      "loss": 1.7114,
+      "step": 10255
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6806623339653015,
+      "learning_rate": 3.51324387258954e-05,
+      "loss": 1.6823,
+      "step": 10260
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7156978249549866,
+      "learning_rate": 3.5119583578059846e-05,
+      "loss": 1.709,
+      "step": 10265
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.7115353345870972,
+      "learning_rate": 3.510672522940975e-05,
+      "loss": 1.7488,
+      "step": 10270
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.8930706977844238,
+      "learning_rate": 3.509386368401218e-05,
+      "loss": 1.6939,
+      "step": 10275
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.555728018283844,
+      "learning_rate": 3.508099894593523e-05,
+      "loss": 1.7732,
+      "step": 10280
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6877692341804504,
+      "learning_rate": 3.506813101924802e-05,
+      "loss": 1.799,
+      "step": 10285
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.9091320633888245,
+      "learning_rate": 3.5055259908020645e-05,
+      "loss": 1.6652,
+      "step": 10290
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.6716402173042297,
+      "learning_rate": 3.504238561632424e-05,
+      "loss": 1.6759,
+      "step": 10295
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.615459144115448,
+      "learning_rate": 3.502950814823092e-05,
+      "loss": 1.7817,
+      "step": 10300
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.8127800822257996,
+      "learning_rate": 3.501662750781382e-05,
+      "loss": 2.2239,
+      "step": 10305
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.9452939033508301,
+      "learning_rate": 3.5003743699147076e-05,
+      "loss": 2.1167,
+      "step": 10310
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7668461799621582,
+      "learning_rate": 3.499085672630582e-05,
+      "loss": 1.6316,
+      "step": 10315
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7267149090766907,
+      "learning_rate": 3.497796659336619e-05,
+      "loss": 2.0752,
+      "step": 10320
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7957209944725037,
+      "learning_rate": 3.4965073304405326e-05,
+      "loss": 1.8228,
+      "step": 10325
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.5732484459877014,
+      "learning_rate": 3.495217686350136e-05,
+      "loss": 1.8593,
+      "step": 10330
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.3122698068618774,
+      "learning_rate": 3.493927727473343e-05,
+      "loss": 1.6822,
+      "step": 10335
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.5948165655136108,
+      "learning_rate": 3.492637454218166e-05,
+      "loss": 1.9183,
+      "step": 10340
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.938048243522644,
+      "learning_rate": 3.491346866992716e-05,
+      "loss": 2.0546,
+      "step": 10345
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.6968057751655579,
+      "learning_rate": 3.490055966205207e-05,
+      "loss": 1.8256,
+      "step": 10350
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.5820090174674988,
+      "learning_rate": 3.4887647522639464e-05,
+      "loss": 1.9695,
+      "step": 10355
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.7669956088066101,
+      "learning_rate": 3.4874732255773465e-05,
+      "loss": 1.7784,
+      "step": 10360
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.5515414476394653,
+      "learning_rate": 3.486181386553916e-05,
+      "loss": 1.9826,
+      "step": 10365
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 7.995153427124023,
+      "learning_rate": 3.484889235602261e-05,
+      "loss": 1.8934,
+      "step": 10370
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.8782067894935608,
+      "learning_rate": 3.483596773131088e-05,
+      "loss": 1.7218,
+      "step": 10375
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.6286051273345947,
+      "learning_rate": 3.482303999549201e-05,
+      "loss": 1.9686,
+      "step": 10380
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.4949547350406647,
+      "learning_rate": 3.481010915265504e-05,
+      "loss": 2.0401,
+      "step": 10385
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.9229368567466736,
+      "learning_rate": 3.4797175206889984e-05,
+      "loss": 1.6785,
+      "step": 10390
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.5106030702590942,
+      "learning_rate": 3.478423816228784e-05,
+      "loss": 1.9079,
+      "step": 10395
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.9068930745124817,
+      "learning_rate": 3.477129802294057e-05,
+      "loss": 1.9197,
+      "step": 10400
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.7757366299629211,
+      "learning_rate": 3.475835479294114e-05,
+      "loss": 1.9806,
+      "step": 10405
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.814594566822052,
+      "learning_rate": 3.4745408476383476e-05,
+      "loss": 2.0118,
+      "step": 10410
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.643321692943573,
+      "learning_rate": 3.473245907736248e-05,
+      "loss": 1.9629,
+      "step": 10415
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7608657479286194,
+      "learning_rate": 3.471950659997404e-05,
+      "loss": 1.8221,
+      "step": 10420
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.1012566089630127,
+      "learning_rate": 3.470655104831501e-05,
+      "loss": 1.87,
+      "step": 10425
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7623602151870728,
+      "learning_rate": 3.469359242648321e-05,
+      "loss": 1.9752,
+      "step": 10430
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.9718965888023376,
+      "learning_rate": 3.468063073857747e-05,
+      "loss": 1.784,
+      "step": 10435
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.79907488822937,
+      "learning_rate": 3.466766598869753e-05,
+      "loss": 2.1134,
+      "step": 10440
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 15.143911361694336,
+      "learning_rate": 3.4654698180944134e-05,
+      "loss": 2.1167,
+      "step": 10445
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.5659872889518738,
+      "learning_rate": 3.4641727319418995e-05,
+      "loss": 2.2435,
+      "step": 10450
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6032043695449829,
+      "learning_rate": 3.4628753408224765e-05,
+      "loss": 2.0959,
+      "step": 10455
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.6183554530143738,
+      "learning_rate": 3.461577645146511e-05,
+      "loss": 1.8351,
+      "step": 10460
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.806084156036377,
+      "learning_rate": 3.4602796453244615e-05,
+      "loss": 1.6052,
+      "step": 10465
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7526884078979492,
+      "learning_rate": 3.4589813417668823e-05,
+      "loss": 1.9665,
+      "step": 10470
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.9556746482849121,
+      "learning_rate": 3.457682734884428e-05,
+      "loss": 1.8647,
+      "step": 10475
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.5155915021896362,
+      "learning_rate": 3.456383825087846e-05,
+      "loss": 1.9263,
+      "step": 10480
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.661363422870636,
+      "learning_rate": 3.45508461278798e-05,
+      "loss": 1.7769,
+      "step": 10485
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.5568251609802246,
+      "learning_rate": 3.453785098395769e-05,
+      "loss": 1.9676,
+      "step": 10490
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6141484975814819,
+      "learning_rate": 3.45248528232225e-05,
+      "loss": 2.0722,
+      "step": 10495
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6515774726867676,
+      "learning_rate": 3.4511851649785506e-05,
+      "loss": 1.9458,
+      "step": 10500
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6424517631530762,
+      "learning_rate": 3.4498847467759e-05,
+      "loss": 1.9665,
+      "step": 10505
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.39145955443382263,
+      "learning_rate": 3.4485840281256176e-05,
+      "loss": 1.9062,
+      "step": 10510
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.9824689030647278,
+      "learning_rate": 3.447283009439119e-05,
+      "loss": 1.6084,
+      "step": 10515
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.44701799750328064,
+      "learning_rate": 3.445981691127915e-05,
+      "loss": 1.9367,
+      "step": 10520
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.8132763504981995,
+      "learning_rate": 3.4446800736036134e-05,
+      "loss": 1.8618,
+      "step": 10525
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.5754613876342773,
+      "learning_rate": 3.4433781572779126e-05,
+      "loss": 1.72,
+      "step": 10530
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.46498289704322815,
+      "learning_rate": 3.442075942562608e-05,
+      "loss": 1.9345,
+      "step": 10535
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.9286674857139587,
+      "learning_rate": 3.4407734298695895e-05,
+      "loss": 1.7018,
+      "step": 10540
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.5865563750267029,
+      "learning_rate": 3.439470619610839e-05,
+      "loss": 1.9787,
+      "step": 10545
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.0381340980529785,
+      "learning_rate": 3.438167512198436e-05,
+      "loss": 1.7573,
+      "step": 10550
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.0353657007217407,
+      "learning_rate": 3.4368641080445504e-05,
+      "loss": 1.7011,
+      "step": 10555
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.5357837677001953,
+      "learning_rate": 3.4355604075614493e-05,
+      "loss": 2.1718,
+      "step": 10560
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7515667080879211,
+      "learning_rate": 3.434256411161491e-05,
+      "loss": 1.9217,
+      "step": 10565
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.7363743185997009,
+      "learning_rate": 3.432952119257128e-05,
+      "loss": 1.9062,
+      "step": 10570
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6471055746078491,
+      "learning_rate": 3.431647532260908e-05,
+      "loss": 2.0589,
+      "step": 10575
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 0.6963765621185303,
+      "learning_rate": 3.4303426505854695e-05,
+      "loss": 1.9334,
+      "step": 10580
+    },
+    {
+      "epoch": 1.89,
+      "grad_norm": 1.006638765335083,
+      "learning_rate": 3.4290374746435454e-05,
+      "loss": 1.8345,
+      "step": 10585
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.5938763618469238,
+      "learning_rate": 3.427732004847961e-05,
+      "loss": 1.7264,
+      "step": 10590
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.7721897959709167,
+      "learning_rate": 3.4264262416116366e-05,
+      "loss": 1.7195,
+      "step": 10595
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6883012652397156,
+      "learning_rate": 3.425120185347584e-05,
+      "loss": 1.5873,
+      "step": 10600
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.9483616352081299,
+      "learning_rate": 3.423813836468904e-05,
+      "loss": 1.8211,
+      "step": 10605
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.6309101581573486,
+      "learning_rate": 3.4225071953887976e-05,
+      "loss": 1.9336,
+      "step": 10610
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8639585971832275,
+      "learning_rate": 3.421200262520553e-05,
+      "loss": 2.0485,
+      "step": 10615
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8056395649909973,
+      "learning_rate": 3.419893038277552e-05,
+      "loss": 1.8187,
+      "step": 10620
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 1.7254058122634888,
+      "learning_rate": 3.418585523073266e-05,
+      "loss": 1.8431,
+      "step": 10625
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.5998656153678894,
+      "learning_rate": 3.417277717321264e-05,
+      "loss": 1.8515,
+      "step": 10630
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.8533723950386047,
+      "learning_rate": 3.4159696214352014e-05,
+      "loss": 1.9317,
+      "step": 10635
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.44307661056518555,
+      "learning_rate": 3.414661235828829e-05,
+      "loss": 1.7573,
+      "step": 10640
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.5623877644538879,
+      "learning_rate": 3.413352560915988e-05,
+      "loss": 1.9952,
+      "step": 10645
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.5950538516044617,
+      "learning_rate": 3.4120435971106105e-05,
+      "loss": 2.0118,
+      "step": 10650
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7914562225341797,
+      "learning_rate": 3.410734344826719e-05,
+      "loss": 1.6905,
+      "step": 10655
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.2145016193389893,
+      "learning_rate": 3.409424804478431e-05,
+      "loss": 1.9919,
+      "step": 10660
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 1.1050101518630981,
+      "learning_rate": 3.408114976479953e-05,
+      "loss": 1.8972,
+      "step": 10665
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7993802428245544,
+      "learning_rate": 3.406804861245581e-05,
+      "loss": 1.7432,
+      "step": 10670
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.7174950838088989,
+      "learning_rate": 3.4054944591897034e-05,
+      "loss": 1.7941,
+      "step": 10675
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 2.201014518737793,
+      "learning_rate": 3.4041837707267996e-05,
+      "loss": 1.8024,
+      "step": 10680
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.560925304889679,
+      "learning_rate": 3.402872796271438e-05,
+      "loss": 1.6103,
+      "step": 10685
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6595713496208191,
+      "learning_rate": 3.4015615362382814e-05,
+      "loss": 1.983,
+      "step": 10690
+    },
+    {
+      "epoch": 1.91,
+      "grad_norm": 0.6843532919883728,
+      "learning_rate": 3.400249991042076e-05,
+      "loss": 2.0138,
+      "step": 10695
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.5372428297996521,
+      "learning_rate": 3.398938161097665e-05,
+      "loss": 2.0141,
+      "step": 10700
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6535273790359497,
+      "learning_rate": 3.3976260468199785e-05,
+      "loss": 2.0042,
+      "step": 10705
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6673722863197327,
+      "learning_rate": 3.3963136486240366e-05,
+      "loss": 1.8941,
+      "step": 10710
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6364368200302124,
+      "learning_rate": 3.3950009669249497e-05,
+      "loss": 1.7138,
+      "step": 10715
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6928640007972717,
+      "learning_rate": 3.393688002137919e-05,
+      "loss": 1.6929,
+      "step": 10720
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.8325533270835876,
+      "learning_rate": 3.392374754678231e-05,
+      "loss": 1.9077,
+      "step": 10725
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 1.1905181407928467,
+      "learning_rate": 3.3910612249612674e-05,
+      "loss": 1.9971,
+      "step": 10730
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6571048498153687,
+      "learning_rate": 3.3897474134024953e-05,
+      "loss": 2.2925,
+      "step": 10735
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.5732169151306152,
+      "learning_rate": 3.3884333204174724e-05,
+      "loss": 1.9575,
+      "step": 10740
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.4101000130176544,
+      "learning_rate": 3.3871189464218445e-05,
+      "loss": 2.1045,
+      "step": 10745
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.73232102394104,
+      "learning_rate": 3.385804291831347e-05,
+      "loss": 1.9257,
+      "step": 10750
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6931615471839905,
+      "learning_rate": 3.384489357061804e-05,
+      "loss": 1.878,
+      "step": 10755
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.5579245686531067,
+      "learning_rate": 3.3831741425291294e-05,
+      "loss": 1.9672,
+      "step": 10760
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.762345016002655,
+      "learning_rate": 3.381858648649322e-05,
+      "loss": 1.8285,
+      "step": 10765
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.3375375270843506,
+      "learning_rate": 3.380542875838472e-05,
+      "loss": 2.3338,
+      "step": 10770
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.3226951360702515,
+      "learning_rate": 3.379226824512758e-05,
+      "loss": 1.9114,
+      "step": 10775
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7157413363456726,
+      "learning_rate": 3.3779104950884446e-05,
+      "loss": 1.8751,
+      "step": 10780
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.9126617312431335,
+      "learning_rate": 3.376593887981887e-05,
+      "loss": 2.0082,
+      "step": 10785
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.915809690952301,
+      "learning_rate": 3.3752770036095236e-05,
+      "loss": 2.2042,
+      "step": 10790
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 1.350906252861023,
+      "learning_rate": 3.373959842387888e-05,
+      "loss": 2.0439,
+      "step": 10795
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.6320156455039978,
+      "learning_rate": 3.3726424047335943e-05,
+      "loss": 2.0582,
+      "step": 10800
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.42162302136421204,
+      "learning_rate": 3.3713246910633473e-05,
+      "loss": 2.2003,
+      "step": 10805
+    },
+    {
+      "epoch": 1.93,
+      "grad_norm": 0.7747112512588501,
+      "learning_rate": 3.370006701793939e-05,
+      "loss": 1.7979,
+      "step": 10810
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.5144717693328857,
+      "learning_rate": 3.3686884373422465e-05,
+      "loss": 1.8804,
+      "step": 10815
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.6446430683135986,
+      "learning_rate": 3.367369898125238e-05,
+      "loss": 2.1296,
+      "step": 10820
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.71882164478302,
+      "learning_rate": 3.366051084559965e-05,
+      "loss": 2.0616,
+      "step": 10825
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.0291924476623535,
+      "learning_rate": 3.3647319970635665e-05,
+      "loss": 1.9716,
+      "step": 10830
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.1064375638961792,
+      "learning_rate": 3.363412636053269e-05,
+      "loss": 1.7938,
+      "step": 10835
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.4092391729354858,
+      "learning_rate": 3.362093001946386e-05,
+      "loss": 2.3134,
+      "step": 10840
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.5512030124664307,
+      "learning_rate": 3.360773095160315e-05,
+      "loss": 1.8288,
+      "step": 10845
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 2.6880881786346436,
+      "learning_rate": 3.3594529161125424e-05,
+      "loss": 1.606,
+      "step": 10850
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.32284656167030334,
+      "learning_rate": 3.358132465220639e-05,
+      "loss": 1.8226,
+      "step": 10855
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.3082622289657593,
+      "learning_rate": 3.356811742902262e-05,
+      "loss": 1.9702,
+      "step": 10860
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.579648494720459,
+      "learning_rate": 3.355490749575155e-05,
+      "loss": 1.8468,
+      "step": 10865
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.6377948522567749,
+      "learning_rate": 3.354169485657147e-05,
+      "loss": 2.1329,
+      "step": 10870
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.47102659940719604,
+      "learning_rate": 3.3528479515661514e-05,
+      "loss": 1.7664,
+      "step": 10875
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.959218442440033,
+      "learning_rate": 3.35152614772017e-05,
+      "loss": 1.4881,
+      "step": 10880
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.5902130007743835,
+      "learning_rate": 3.3502040745372874e-05,
+      "loss": 1.9162,
+      "step": 10885
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.0470603704452515,
+      "learning_rate": 3.348881732435673e-05,
+      "loss": 1.4373,
+      "step": 10890
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.8185494542121887,
+      "learning_rate": 3.347559121833584e-05,
+      "loss": 1.8553,
+      "step": 10895
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.7984978556632996,
+      "learning_rate": 3.34623624314936e-05,
+      "loss": 1.7071,
+      "step": 10900
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.5953229665756226,
+      "learning_rate": 3.344913096801425e-05,
+      "loss": 1.9628,
+      "step": 10905
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 0.4407386779785156,
+      "learning_rate": 3.3435896832082916e-05,
+      "loss": 1.8533,
+      "step": 10910
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 9.959710121154785,
+      "learning_rate": 3.3422660027885515e-05,
+      "loss": 1.8898,
+      "step": 10915
+    },
+    {
+      "epoch": 1.95,
+      "grad_norm": 1.026064157485962,
+      "learning_rate": 3.340942055960886e-05,
+      "loss": 1.9131,
+      "step": 10920
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8007922172546387,
+      "learning_rate": 3.339617843144057e-05,
+      "loss": 1.7598,
+      "step": 10925
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.5460824966430664,
+      "learning_rate": 3.3382933647569115e-05,
+      "loss": 2.0906,
+      "step": 10930
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6225911378860474,
+      "learning_rate": 3.33696862121838e-05,
+      "loss": 1.8498,
+      "step": 10935
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6068547368049622,
+      "learning_rate": 3.33564361294748e-05,
+      "loss": 1.8413,
+      "step": 10940
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.0755717754364014,
+      "learning_rate": 3.334318340363309e-05,
+      "loss": 2.0584,
+      "step": 10945
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.5796096920967102,
+      "learning_rate": 3.332992803885049e-05,
+      "loss": 2.1057,
+      "step": 10950
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.20377254486084,
+      "learning_rate": 3.331667003931967e-05,
+      "loss": 1.8257,
+      "step": 10955
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.8372711539268494,
+      "learning_rate": 3.330340940923411e-05,
+      "loss": 1.9423,
+      "step": 10960
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.5656445622444153,
+      "learning_rate": 3.3290146152788136e-05,
+      "loss": 1.7365,
+      "step": 10965
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.3391028344631195,
+      "learning_rate": 3.3276880274176927e-05,
+      "loss": 1.7596,
+      "step": 10970
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 1.2373957633972168,
+      "learning_rate": 3.326361177759643e-05,
+      "loss": 1.7008,
+      "step": 10975
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.6096294522285461,
+      "learning_rate": 3.325034066724348e-05,
+      "loss": 1.8067,
+      "step": 10980
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7483140230178833,
+      "learning_rate": 3.323706694731572e-05,
+      "loss": 1.6209,
+      "step": 10985
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.4939500391483307,
+      "learning_rate": 3.3223790622011595e-05,
+      "loss": 1.7462,
+      "step": 10990
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.016813039779663,
+      "learning_rate": 3.321051169553042e-05,
+      "loss": 1.8239,
+      "step": 10995
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.5926358103752136,
+      "learning_rate": 3.3197230172072294e-05,
+      "loss": 2.0223,
+      "step": 11000
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7104119062423706,
+      "learning_rate": 3.318394605583813e-05,
+      "loss": 1.7247,
+      "step": 11005
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 1.1022257804870605,
+      "learning_rate": 3.317065935102972e-05,
+      "loss": 1.6968,
+      "step": 11010
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.7626497745513916,
+      "learning_rate": 3.315737006184961e-05,
+      "loss": 1.9668,
+      "step": 11015
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.606796383857727,
+      "learning_rate": 3.314407819250119e-05,
+      "loss": 1.8814,
+      "step": 11020
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.34029826521873474,
+      "learning_rate": 3.313078374718868e-05,
+      "loss": 1.6823,
+      "step": 11025
+    },
+    {
+      "epoch": 1.97,
+      "grad_norm": 0.5755228400230408,
+      "learning_rate": 3.311748673011709e-05,
+      "loss": 1.6421,
+      "step": 11030
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7753984332084656,
+      "learning_rate": 3.310418714549225e-05,
+      "loss": 2.0978,
+      "step": 11035
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.5059102177619934,
+      "learning_rate": 3.309088499752082e-05,
+      "loss": 2.0625,
+      "step": 11040
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6016696691513062,
+      "learning_rate": 3.307758029041024e-05,
+      "loss": 1.8624,
+      "step": 11045
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7924507856369019,
+      "learning_rate": 3.306427302836879e-05,
+      "loss": 1.8362,
+      "step": 11050
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7843905687332153,
+      "learning_rate": 3.3050963215605526e-05,
+      "loss": 2.0905,
+      "step": 11055
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.0598030090332031,
+      "learning_rate": 3.3037650856330354e-05,
+      "loss": 1.7577,
+      "step": 11060
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7119802832603455,
+      "learning_rate": 3.3024335954753946e-05,
+      "loss": 1.896,
+      "step": 11065
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.5891833901405334,
+      "learning_rate": 3.301101851508779e-05,
+      "loss": 2.03,
+      "step": 11070
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.3929714262485504,
+      "learning_rate": 3.2997698541544184e-05,
+      "loss": 1.6645,
+      "step": 11075
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.6735256910324097,
+      "learning_rate": 3.298437603833622e-05,
+      "loss": 2.2231,
+      "step": 11080
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.9890113472938538,
+      "learning_rate": 3.29710510096778e-05,
+      "loss": 1.8691,
+      "step": 11085
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.49276089668273926,
+      "learning_rate": 3.295772345978361e-05,
+      "loss": 1.8855,
+      "step": 11090
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7664940357208252,
+      "learning_rate": 3.2944393392869145e-05,
+      "loss": 2.1009,
+      "step": 11095
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7249503135681152,
+      "learning_rate": 3.2931060813150685e-05,
+      "loss": 1.8206,
+      "step": 11100
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 1.5629411935806274,
+      "learning_rate": 3.291772572484533e-05,
+      "loss": 1.7675,
+      "step": 11105
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.38072723150253296,
+      "learning_rate": 3.2904388132170936e-05,
+      "loss": 1.8236,
+      "step": 11110
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7046849727630615,
+      "learning_rate": 3.2891048039346177e-05,
+      "loss": 1.7859,
+      "step": 11115
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.5897841453552246,
+      "learning_rate": 3.2877705450590526e-05,
+      "loss": 1.9604,
+      "step": 11120
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6484891772270203,
+      "learning_rate": 3.2864360370124206e-05,
+      "loss": 1.8623,
+      "step": 11125
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.7585556507110596,
+      "learning_rate": 3.285101280216827e-05,
+      "loss": 1.746,
+      "step": 11130
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.6930608153343201,
+      "learning_rate": 3.2837662750944535e-05,
+      "loss": 1.5794,
+      "step": 11135
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.684201717376709,
+      "learning_rate": 3.282431022067561e-05,
+      "loss": 1.8985,
+      "step": 11140
+    },
+    {
+      "epoch": 1.99,
+      "grad_norm": 0.8467980027198792,
+      "learning_rate": 3.281095521558488e-05,
+      "loss": 1.8468,
+      "step": 11145
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.533331871032715,
+      "learning_rate": 3.2797597739896545e-05,
+      "loss": 1.6749,
+      "step": 11150
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.5449894666671753,
+      "learning_rate": 3.278423779783554e-05,
+      "loss": 1.7722,
+      "step": 11155
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.5651440024375916,
+      "learning_rate": 3.27708753936276e-05,
+      "loss": 1.8541,
+      "step": 11160
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.9599629640579224,
+      "learning_rate": 3.2757510531499256e-05,
+      "loss": 1.9124,
+      "step": 11165
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.530594527721405,
+      "learning_rate": 3.274414321567779e-05,
+      "loss": 2.0135,
+      "step": 11170
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.8075482249259949,
+      "learning_rate": 3.273077345039127e-05,
+      "loss": 1.8081,
+      "step": 11175
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.5179402232170105,
+      "learning_rate": 3.271740123986856e-05,
+      "loss": 1.8615,
+      "step": 11180
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6782111525535583,
+      "learning_rate": 3.270402658833924e-05,
+      "loss": 1.9938,
+      "step": 11185
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.5111522674560547,
+      "learning_rate": 3.2690649500033726e-05,
+      "loss": 1.5514,
+      "step": 11190
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.3096226453781128,
+      "learning_rate": 3.267726997918318e-05,
+      "loss": 1.7645,
+      "step": 11195
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.4817447364330292,
+      "learning_rate": 3.266388803001951e-05,
+      "loss": 1.7603,
+      "step": 11200
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.8163765668869019,
+      "learning_rate": 3.2650503656775446e-05,
+      "loss": 1.8734,
+      "step": 11205
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.504024863243103,
+      "learning_rate": 3.2637116863684426e-05,
+      "loss": 1.5327,
+      "step": 11210
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.4556671380996704,
+      "learning_rate": 3.2623727654980686e-05,
+      "loss": 1.9113,
+      "step": 11215
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.9641478657722473,
+      "learning_rate": 3.261033603489923e-05,
+      "loss": 1.8351,
+      "step": 11220
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.5921196937561035,
+      "learning_rate": 3.259694200767579e-05,
+      "loss": 1.9184,
+      "step": 11225
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.8275460004806519,
+      "learning_rate": 3.258354557754691e-05,
+      "loss": 1.7349,
+      "step": 11230
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.7848659753799438,
+      "learning_rate": 3.257014674874986e-05,
+      "loss": 1.7097,
+      "step": 11235
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.5022772550582886,
+      "learning_rate": 3.255674552552267e-05,
+      "loss": 1.7641,
+      "step": 11240
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 1.0205901861190796,
+      "learning_rate": 3.254334191210414e-05,
+      "loss": 1.9206,
+      "step": 11245
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.4597795009613037,
+      "learning_rate": 3.252993591273382e-05,
+      "loss": 1.9408,
+      "step": 11250
+    },
+    {
+      "epoch": 2.01,
+      "grad_norm": 0.4975505769252777,
+      "learning_rate": 3.251652753165202e-05,
+      "loss": 1.9817,
+      "step": 11255
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.8713247776031494,
+      "learning_rate": 3.2503116773099786e-05,
+      "loss": 1.69,
+      "step": 11260
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6077904105186462,
+      "learning_rate": 3.248970364131894e-05,
+      "loss": 1.8712,
+      "step": 11265
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.489361971616745,
+      "learning_rate": 3.2476288140552026e-05,
+      "loss": 1.7067,
+      "step": 11270
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7086228728294373,
+      "learning_rate": 3.246287027504237e-05,
+      "loss": 1.7504,
+      "step": 11275
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7792586088180542,
+      "learning_rate": 3.2449450049034024e-05,
+      "loss": 1.874,
+      "step": 11280
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.6796438097953796,
+      "learning_rate": 3.243602746677179e-05,
+      "loss": 1.8736,
+      "step": 11285
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.0525963306427002,
+      "learning_rate": 3.242260253250122e-05,
+      "loss": 1.7056,
+      "step": 11290
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.3810064196586609,
+      "learning_rate": 3.240917525046862e-05,
+      "loss": 1.7772,
+      "step": 11295
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.3621659278869629,
+      "learning_rate": 3.2395745624921e-05,
+      "loss": 1.8997,
+      "step": 11300
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.7644492387771606,
+      "learning_rate": 3.238231366010616e-05,
+      "loss": 1.7473,
+      "step": 11305
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 0.8346506953239441,
+      "learning_rate": 3.2368879360272606e-05,
+      "loss": 1.7338,
+      "step": 11310
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.7360906004905701,
+      "learning_rate": 3.235544272966961e-05,
+      "loss": 1.6659,
+      "step": 11315
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.8705644607543945,
+      "learning_rate": 3.2342003772547145e-05,
+      "loss": 1.9399,
+      "step": 11320
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.1617865562438965,
+      "learning_rate": 3.232856249315595e-05,
+      "loss": 1.7525,
+      "step": 11325
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.6664028167724609,
+      "learning_rate": 3.231511889574748e-05,
+      "loss": 1.7085,
+      "step": 11330
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.378801167011261,
+      "learning_rate": 3.230167298457395e-05,
+      "loss": 2.1498,
+      "step": 11335
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 2.0433850288391113,
+      "learning_rate": 3.2288224763888274e-05,
+      "loss": 1.9798,
+      "step": 11340
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.8449577689170837,
+      "learning_rate": 3.227477423794412e-05,
+      "loss": 1.9539,
+      "step": 11345
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.6414170265197754,
+      "learning_rate": 3.226132141099586e-05,
+      "loss": 1.8589,
+      "step": 11350
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.9575155377388,
+      "learning_rate": 3.2247866287298634e-05,
+      "loss": 1.749,
+      "step": 11355
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 1.6495996713638306,
+      "learning_rate": 3.2234408871108266e-05,
+      "loss": 1.7329,
+      "step": 11360
+    },
+    {
+      "epoch": 2.03,
+      "grad_norm": 0.5812102556228638,
+      "learning_rate": 3.2220949166681335e-05,
+      "loss": 1.9318,
+      "step": 11365
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.8048359155654907,
+      "learning_rate": 3.220748717827513e-05,
+      "loss": 1.853,
+      "step": 11370
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7342902421951294,
+      "learning_rate": 3.219402291014766e-05,
+      "loss": 1.9342,
+      "step": 11375
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7486286163330078,
+      "learning_rate": 3.218055636655766e-05,
+      "loss": 2.1832,
+      "step": 11380
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.1646840572357178,
+      "learning_rate": 3.21670875517646e-05,
+      "loss": 1.7443,
+      "step": 11385
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.269112467765808,
+      "learning_rate": 3.215361647002863e-05,
+      "loss": 1.9097,
+      "step": 11390
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7798238396644592,
+      "learning_rate": 3.2140143125610654e-05,
+      "loss": 1.9486,
+      "step": 11395
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.9379304647445679,
+      "learning_rate": 3.212666752277228e-05,
+      "loss": 1.7346,
+      "step": 11400
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 1.0831714868545532,
+      "learning_rate": 3.211318966577581e-05,
+      "loss": 2.0097,
+      "step": 11405
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.3674708902835846,
+      "learning_rate": 3.20997095588843e-05,
+      "loss": 2.05,
+      "step": 11410
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.9015005826950073,
+      "learning_rate": 3.20862272063615e-05,
+      "loss": 1.7768,
+      "step": 11415
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.8480490446090698,
+      "learning_rate": 3.207274261247183e-05,
+      "loss": 1.768,
+      "step": 11420
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.48236319422721863,
+      "learning_rate": 3.2059255781480475e-05,
+      "loss": 1.9133,
+      "step": 11425
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 1.2355780601501465,
+      "learning_rate": 3.204576671765331e-05,
+      "loss": 1.9076,
+      "step": 11430
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6403537392616272,
+      "learning_rate": 3.2032275425256916e-05,
+      "loss": 1.7401,
+      "step": 11435
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7580591440200806,
+      "learning_rate": 3.201878190855857e-05,
+      "loss": 1.8118,
+      "step": 11440
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.9600961208343506,
+      "learning_rate": 3.2005286171826256e-05,
+      "loss": 2.1253,
+      "step": 11445
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.496980756521225,
+      "learning_rate": 3.199178821932865e-05,
+      "loss": 1.9104,
+      "step": 11450
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6483522653579712,
+      "learning_rate": 3.1978288055335164e-05,
+      "loss": 1.7537,
+      "step": 11455
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.5102001428604126,
+      "learning_rate": 3.196478568411589e-05,
+      "loss": 1.8392,
+      "step": 11460
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.5300964713096619,
+      "learning_rate": 3.195128110994159e-05,
+      "loss": 2.1472,
+      "step": 11465
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7955127358436584,
+      "learning_rate": 3.193777433708376e-05,
+      "loss": 1.9474,
+      "step": 11470
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.7563099265098572,
+      "learning_rate": 3.192426536981459e-05,
+      "loss": 1.9155,
+      "step": 11475
+    },
+    {
+      "epoch": 2.05,
+      "grad_norm": 0.6697816252708435,
+      "learning_rate": 3.191075421240694e-05,
+      "loss": 1.9573,
+      "step": 11480
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7206788063049316,
+      "learning_rate": 3.189724086913438e-05,
+      "loss": 1.8042,
+      "step": 11485
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.4096161127090454,
+      "learning_rate": 3.188372534427117e-05,
+      "loss": 2.0095,
+      "step": 11490
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6973406672477722,
+      "learning_rate": 3.1870207642092246e-05,
+      "loss": 1.4596,
+      "step": 11495
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.4881574213504791,
+      "learning_rate": 3.185668776687326e-05,
+      "loss": 1.9685,
+      "step": 11500
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6438494920730591,
+      "learning_rate": 3.184316572289053e-05,
+      "loss": 1.9556,
+      "step": 11505
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7263823747634888,
+      "learning_rate": 3.182964151442107e-05,
+      "loss": 2.1189,
+      "step": 11510
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.909968912601471,
+      "learning_rate": 3.181611514574255e-05,
+      "loss": 1.7608,
+      "step": 11515
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6028372645378113,
+      "learning_rate": 3.180258662113338e-05,
+      "loss": 1.7227,
+      "step": 11520
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.7164033651351929,
+      "learning_rate": 3.17890559448726e-05,
+      "loss": 1.978,
+      "step": 11525
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.8094786405563354,
+      "learning_rate": 3.177552312123995e-05,
+      "loss": 1.957,
+      "step": 11530
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 0.6405372619628906,
+      "learning_rate": 3.1761988154515864e-05,
+      "loss": 1.9312,
+      "step": 11535
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.37368497252464294,
+      "learning_rate": 3.1748451048981424e-05,
+      "loss": 2.1636,
+      "step": 11540
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.5130252242088318,
+      "learning_rate": 3.1734911808918406e-05,
+      "loss": 1.5754,
+      "step": 11545
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.6798210740089417,
+      "learning_rate": 3.172137043860927e-05,
+      "loss": 1.8592,
+      "step": 11550
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.8292738199234009,
+      "learning_rate": 3.170782694233712e-05,
+      "loss": 1.6516,
+      "step": 11555
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.5600850582122803,
+      "learning_rate": 3.169428132438576e-05,
+      "loss": 1.7959,
+      "step": 11560
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.9339403510093689,
+      "learning_rate": 3.168073358903966e-05,
+      "loss": 1.8377,
+      "step": 11565
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.5029419660568237,
+      "learning_rate": 3.166718374058395e-05,
+      "loss": 2.0719,
+      "step": 11570
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 1.2558355331420898,
+      "learning_rate": 3.165363178330444e-05,
+      "loss": 1.8537,
+      "step": 11575
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.9240514039993286,
+      "learning_rate": 3.16400777214876e-05,
+      "loss": 1.5651,
+      "step": 11580
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.5042521357536316,
+      "learning_rate": 3.1626521559420556e-05,
+      "loss": 1.6923,
+      "step": 11585
+    },
+    {
+      "epoch": 2.07,
+      "grad_norm": 0.8794254064559937,
+      "learning_rate": 3.161296330139111e-05,
+      "loss": 1.6979,
+      "step": 11590
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6789154410362244,
+      "learning_rate": 3.1599402951687744e-05,
+      "loss": 1.9248,
+      "step": 11595
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.4555959403514862,
+      "learning_rate": 3.1585840514599574e-05,
+      "loss": 1.7823,
+      "step": 11600
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.4320109784603119,
+      "learning_rate": 3.1572275994416376e-05,
+      "loss": 1.9124,
+      "step": 11605
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.5282646417617798,
+      "learning_rate": 3.155870939542861e-05,
+      "loss": 1.8452,
+      "step": 11610
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 1.2354097366333008,
+      "learning_rate": 3.154514072192736e-05,
+      "loss": 2.0078,
+      "step": 11615
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.7362090945243835,
+      "learning_rate": 3.153156997820441e-05,
+      "loss": 1.7717,
+      "step": 11620
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.5220533609390259,
+      "learning_rate": 3.151799716855215e-05,
+      "loss": 1.6834,
+      "step": 11625
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6556394696235657,
+      "learning_rate": 3.150442229726366e-05,
+      "loss": 1.7327,
+      "step": 11630
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.4095345139503479,
+      "learning_rate": 3.1490845368632645e-05,
+      "loss": 1.852,
+      "step": 11635
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6544451117515564,
+      "learning_rate": 3.147726638695349e-05,
+      "loss": 1.8067,
+      "step": 11640
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.5166847705841064,
+      "learning_rate": 3.14636853565212e-05,
+      "loss": 1.7826,
+      "step": 11645
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.8674813508987427,
+      "learning_rate": 3.145010228163145e-05,
+      "loss": 1.9693,
+      "step": 11650
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7544355988502502,
+      "learning_rate": 3.1436517166580565e-05,
+      "loss": 1.6409,
+      "step": 11655
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.8396444916725159,
+      "learning_rate": 3.1422930015665484e-05,
+      "loss": 1.9227,
+      "step": 11660
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7315942049026489,
+      "learning_rate": 3.140934083318382e-05,
+      "loss": 1.6208,
+      "step": 11665
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 3.2649600505828857,
+      "learning_rate": 3.139574962343381e-05,
+      "loss": 1.7472,
+      "step": 11670
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.7523483037948608,
+      "learning_rate": 3.138215639071435e-05,
+      "loss": 1.7216,
+      "step": 11675
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.9226972460746765,
+      "learning_rate": 3.1368561139324956e-05,
+      "loss": 1.6117,
+      "step": 11680
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 1.0279871225357056,
+      "learning_rate": 3.13549638735658e-05,
+      "loss": 2.0684,
+      "step": 11685
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.8119601607322693,
+      "learning_rate": 3.1341364597737686e-05,
+      "loss": 1.7426,
+      "step": 11690
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.5219517350196838,
+      "learning_rate": 3.132776331614205e-05,
+      "loss": 2.1323,
+      "step": 11695
+    },
+    {
+      "epoch": 2.09,
+      "grad_norm": 0.8303288221359253,
+      "learning_rate": 3.131416003308097e-05,
+      "loss": 1.7259,
+      "step": 11700
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.494863897562027,
+      "learning_rate": 3.130055475285714e-05,
+      "loss": 1.9558,
+      "step": 11705
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.1602771282196045,
+      "learning_rate": 3.12869474797739e-05,
+      "loss": 1.8207,
+      "step": 11710
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6153246164321899,
+      "learning_rate": 3.127333821813522e-05,
+      "loss": 1.8699,
+      "step": 11715
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.7908428907394409,
+      "learning_rate": 3.1259726972245694e-05,
+      "loss": 1.8281,
+      "step": 11720
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.8119539618492126,
+      "learning_rate": 3.124611374641056e-05,
+      "loss": 1.8657,
+      "step": 11725
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.5964948534965515,
+      "learning_rate": 3.1232498544935635e-05,
+      "loss": 1.8478,
+      "step": 11730
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.046675682067871,
+      "learning_rate": 3.121888137212742e-05,
+      "loss": 1.7418,
+      "step": 11735
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6229978799819946,
+      "learning_rate": 3.120526223229302e-05,
+      "loss": 1.8828,
+      "step": 11740
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.7325795888900757,
+      "learning_rate": 3.119164112974014e-05,
+      "loss": 1.6943,
+      "step": 11745
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.5110823512077332,
+      "learning_rate": 3.1178018068777125e-05,
+      "loss": 2.0465,
+      "step": 11750
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.5617272257804871,
+      "learning_rate": 3.1164393053712944e-05,
+      "loss": 2.0785,
+      "step": 11755
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6913292407989502,
+      "learning_rate": 3.115076608885716e-05,
+      "loss": 1.8045,
+      "step": 11760
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.4254094362258911,
+      "learning_rate": 3.1137137178519985e-05,
+      "loss": 1.7095,
+      "step": 11765
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7044830322265625,
+      "learning_rate": 3.112350632701222e-05,
+      "loss": 1.9113,
+      "step": 11770
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.994310200214386,
+      "learning_rate": 3.110987353864529e-05,
+      "loss": 1.6473,
+      "step": 11775
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.0846924781799316,
+      "learning_rate": 3.109623881773124e-05,
+      "loss": 1.7661,
+      "step": 11780
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.5573928356170654,
+      "learning_rate": 3.108260216858272e-05,
+      "loss": 1.8612,
+      "step": 11785
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 1.210252046585083,
+      "learning_rate": 3.106896359551299e-05,
+      "loss": 1.9547,
+      "step": 11790
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.9023343920707703,
+      "learning_rate": 3.1055323102835895e-05,
+      "loss": 1.7782,
+      "step": 11795
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 21.55355453491211,
+      "learning_rate": 3.1041680694865935e-05,
+      "loss": 1.7996,
+      "step": 11800
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.7996496558189392,
+      "learning_rate": 3.102803637591818e-05,
+      "loss": 2.1707,
+      "step": 11805
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 0.8086062073707581,
+      "learning_rate": 3.1014390150308326e-05,
+      "loss": 2.1546,
+      "step": 11810
+    },
+    {
+      "epoch": 2.11,
+      "grad_norm": 3.054373025894165,
+      "learning_rate": 3.1003471799924244e-05,
+      "loss": 2.0085,
+      "step": 11815
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.8012299537658691,
+      "learning_rate": 3.098982215320005e-05,
+      "loss": 1.6379,
+      "step": 11820
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.008513331413269,
+      "learning_rate": 3.097617061190086e-05,
+      "loss": 1.9324,
+      "step": 11825
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.827659547328949,
+      "learning_rate": 3.096251718034466e-05,
+      "loss": 2.1212,
+      "step": 11830
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.8780943155288696,
+      "learning_rate": 3.0948861862850005e-05,
+      "loss": 1.4271,
+      "step": 11835
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.5806859731674194,
+      "learning_rate": 3.093520466373607e-05,
+      "loss": 1.9781,
+      "step": 11840
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.8726419806480408,
+      "learning_rate": 3.0921545587322605e-05,
+      "loss": 1.7361,
+      "step": 11845
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.8282438516616821,
+      "learning_rate": 3.090788463792996e-05,
+      "loss": 1.4147,
+      "step": 11850
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.0596107244491577,
+      "learning_rate": 3.0894221819879094e-05,
+      "loss": 1.8455,
+      "step": 11855
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.9032155275344849,
+      "learning_rate": 3.0880557137491546e-05,
+      "loss": 1.897,
+      "step": 11860
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 2.334949016571045,
+      "learning_rate": 3.0866890595089414e-05,
+      "loss": 2.0026,
+      "step": 11865
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.5920724868774414,
+      "learning_rate": 3.085322219699544e-05,
+      "loss": 1.7887,
+      "step": 11870
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.8088483214378357,
+      "learning_rate": 3.083955194753291e-05,
+      "loss": 1.6292,
+      "step": 11875
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.0029832124710083,
+      "learning_rate": 3.082587985102572e-05,
+      "loss": 1.8685,
+      "step": 11880
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.9885581135749817,
+      "learning_rate": 3.0812205911798336e-05,
+      "loss": 1.812,
+      "step": 11885
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.0413260459899902,
+      "learning_rate": 3.079853013417582e-05,
+      "loss": 2.1284,
+      "step": 11890
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.0664329528808594,
+      "learning_rate": 3.078485252248379e-05,
+      "loss": 2.0554,
+      "step": 11895
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.3859197795391083,
+      "learning_rate": 3.077117308104846e-05,
+      "loss": 2.0608,
+      "step": 11900
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.6689034104347229,
+      "learning_rate": 3.0757491814196665e-05,
+      "loss": 1.7033,
+      "step": 11905
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.5232422351837158,
+      "learning_rate": 3.074380872625573e-05,
+      "loss": 1.7849,
+      "step": 11910
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.7212871313095093,
+      "learning_rate": 3.073012382155362e-05,
+      "loss": 1.9207,
+      "step": 11915
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 0.5604449510574341,
+      "learning_rate": 3.071643710441886e-05,
+      "loss": 1.8573,
+      "step": 11920
+    },
+    {
+      "epoch": 2.13,
+      "grad_norm": 1.3720588684082031,
+      "learning_rate": 3.070274857918054e-05,
+      "loss": 1.696,
+      "step": 11925
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.821916937828064,
+      "learning_rate": 3.068905825016834e-05,
+      "loss": 1.5985,
+      "step": 11930
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.8182123899459839,
+      "learning_rate": 3.0675366121712476e-05,
+      "loss": 2.0291,
+      "step": 11935
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.4312397241592407,
+      "learning_rate": 3.066167219814376e-05,
+      "loss": 1.5368,
+      "step": 11940
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.5408018231391907,
+      "learning_rate": 3.0647976483793584e-05,
+      "loss": 1.828,
+      "step": 11945
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6024120450019836,
+      "learning_rate": 3.063427898299388e-05,
+      "loss": 1.8548,
+      "step": 11950
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.9336116313934326,
+      "learning_rate": 3.062057970007715e-05,
+      "loss": 1.6518,
+      "step": 11955
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 2.2404088973999023,
+      "learning_rate": 3.060687863937647e-05,
+      "loss": 2.039,
+      "step": 11960
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6012473106384277,
+      "learning_rate": 3.059317580522546e-05,
+      "loss": 2.2862,
+      "step": 11965
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.5676471590995789,
+      "learning_rate": 3.057947120195833e-05,
+      "loss": 2.1326,
+      "step": 11970
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.6940319538116455,
+      "learning_rate": 3.056576483390983e-05,
+      "loss": 2.0709,
+      "step": 11975
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 0.5876992344856262,
+      "learning_rate": 3.055205670541527e-05,
+      "loss": 1.8725,
+      "step": 11980
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.3545783758163452,
+      "learning_rate": 3.05383468208105e-05,
+      "loss": 1.7014,
+      "step": 11985
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.659228503704071,
+      "learning_rate": 3.052463518443196e-05,
+      "loss": 1.6168,
+      "step": 11990
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.3598320782184601,
+      "learning_rate": 3.0510921800616633e-05,
+      "loss": 1.812,
+      "step": 11995
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 1.2681336402893066,
+      "learning_rate": 3.049720667370204e-05,
+      "loss": 1.6559,
+      "step": 12000
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.691615879535675,
+      "learning_rate": 3.048348980802626e-05,
+      "loss": 1.7639,
+      "step": 12005
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6840038299560547,
+      "learning_rate": 3.0469771207927932e-05,
+      "loss": 1.7349,
+      "step": 12010
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.38991519808769226,
+      "learning_rate": 3.0456050877746228e-05,
+      "loss": 2.1607,
+      "step": 12015
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6570853590965271,
+      "learning_rate": 3.044232882182088e-05,
+      "loss": 2.0544,
+      "step": 12020
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6140663027763367,
+      "learning_rate": 3.042860504449217e-05,
+      "loss": 1.7201,
+      "step": 12025
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.6192611455917358,
+      "learning_rate": 3.0414879550100894e-05,
+      "loss": 1.895,
+      "step": 12030
+    },
+    {
+      "epoch": 2.15,
+      "grad_norm": 0.8460562825202942,
+      "learning_rate": 3.0401152342988426e-05,
+      "loss": 1.7679,
+      "step": 12035
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6160430312156677,
+      "learning_rate": 3.0387423427496674e-05,
+      "loss": 1.7706,
+      "step": 12040
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.532014787197113,
+      "learning_rate": 3.0373692807968074e-05,
+      "loss": 2.0207,
+      "step": 12045
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6476467847824097,
+      "learning_rate": 3.0359960488745598e-05,
+      "loss": 1.6105,
+      "step": 12050
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.520656406879425,
+      "learning_rate": 3.0346226474172783e-05,
+      "loss": 1.8744,
+      "step": 12055
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.8512612581253052,
+      "learning_rate": 3.0332490768593675e-05,
+      "loss": 1.8777,
+      "step": 12060
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6043909788131714,
+      "learning_rate": 3.0318753376352866e-05,
+      "loss": 1.6042,
+      "step": 12065
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.555930256843567,
+      "learning_rate": 3.030501430179548e-05,
+      "loss": 1.7851,
+      "step": 12070
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.8706346750259399,
+      "learning_rate": 3.0291273549267173e-05,
+      "loss": 1.9402,
+      "step": 12075
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7135600447654724,
+      "learning_rate": 3.027753112311413e-05,
+      "loss": 2.1418,
+      "step": 12080
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.1376060247421265,
+      "learning_rate": 3.026378702768307e-05,
+      "loss": 1.8637,
+      "step": 12085
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6475717425346375,
+      "learning_rate": 3.0250041267321232e-05,
+      "loss": 2.0024,
+      "step": 12090
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.3832349181175232,
+      "learning_rate": 3.02362938463764e-05,
+      "loss": 1.837,
+      "step": 12095
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.567743182182312,
+      "learning_rate": 3.0222544769196858e-05,
+      "loss": 1.7733,
+      "step": 12100
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7041072249412537,
+      "learning_rate": 3.0208794040131426e-05,
+      "loss": 1.5469,
+      "step": 12105
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7668848633766174,
+      "learning_rate": 3.0195041663529456e-05,
+      "loss": 1.9095,
+      "step": 12110
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.9507167935371399,
+      "learning_rate": 3.01812876437408e-05,
+      "loss": 1.8337,
+      "step": 12115
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.7399880886077881,
+      "learning_rate": 3.0167531985115842e-05,
+      "loss": 1.6083,
+      "step": 12120
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.739559531211853,
+      "learning_rate": 3.0153774692005492e-05,
+      "loss": 1.6946,
+      "step": 12125
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.6560977697372437,
+      "learning_rate": 3.0140015768761164e-05,
+      "loss": 1.874,
+      "step": 12130
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.5614013075828552,
+      "learning_rate": 3.012625521973479e-05,
+      "loss": 1.8449,
+      "step": 12135
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.746187686920166,
+      "learning_rate": 3.011249304927883e-05,
+      "loss": 1.7156,
+      "step": 12140
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 0.5409942865371704,
+      "learning_rate": 3.0098729261746228e-05,
+      "loss": 2.0243,
+      "step": 12145
+    },
+    {
+      "epoch": 2.17,
+      "grad_norm": 1.222755789756775,
+      "learning_rate": 3.0084963861490468e-05,
+      "loss": 1.7754,
+      "step": 12150
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.2106776237487793,
+      "learning_rate": 3.0071196852865528e-05,
+      "loss": 1.9023,
+      "step": 12155
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.8225529193878174,
+      "learning_rate": 3.0057428240225897e-05,
+      "loss": 2.0112,
+      "step": 12160
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.5646623969078064,
+      "learning_rate": 3.0043658027926585e-05,
+      "loss": 1.5788,
+      "step": 12165
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7176980972290039,
+      "learning_rate": 3.0029886220323082e-05,
+      "loss": 1.9195,
+      "step": 12170
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.6058060526847839,
+      "learning_rate": 3.0016112821771418e-05,
+      "loss": 1.9483,
+      "step": 12175
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.7406783699989319,
+      "learning_rate": 3.000233783662808e-05,
+      "loss": 1.8359,
+      "step": 12180
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.0844707489013672,
+      "learning_rate": 2.998856126925011e-05,
+      "loss": 1.972,
+      "step": 12185
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 0.5166330933570862,
+      "learning_rate": 2.9974783123995005e-05,
+      "loss": 1.7504,
+      "step": 12190
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.1674580574035645,
+      "learning_rate": 2.9961003405220774e-05,
+      "loss": 1.6967,
+      "step": 12195
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.0376147031784058,
+      "learning_rate": 2.9947222117285945e-05,
+      "loss": 1.8488,
+      "step": 12200
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.062048077583313,
+      "learning_rate": 2.9933439264549518e-05,
+      "loss": 1.6505,
+      "step": 12205
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7804186344146729,
+      "learning_rate": 2.9919654851370998e-05,
+      "loss": 1.6969,
+      "step": 12210
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.533622145652771,
+      "learning_rate": 2.9905868882110378e-05,
+      "loss": 1.9222,
+      "step": 12215
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6791996955871582,
+      "learning_rate": 2.989208136112815e-05,
+      "loss": 1.8727,
+      "step": 12220
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.5420125722885132,
+      "learning_rate": 2.9878292292785293e-05,
+      "loss": 1.4943,
+      "step": 12225
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7872235774993896,
+      "learning_rate": 2.9864501681443276e-05,
+      "loss": 1.5611,
+      "step": 12230
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7012057900428772,
+      "learning_rate": 2.985070953146406e-05,
+      "loss": 1.6986,
+      "step": 12235
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.6529300212860107,
+      "learning_rate": 2.983691584721008e-05,
+      "loss": 1.6813,
+      "step": 12240
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.9842402935028076,
+      "learning_rate": 2.982312063304427e-05,
+      "loss": 1.9067,
+      "step": 12245
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.594903826713562,
+      "learning_rate": 2.9809323893330043e-05,
+      "loss": 1.6995,
+      "step": 12250
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 1.0693496465682983,
+      "learning_rate": 2.9795525632431297e-05,
+      "loss": 1.4401,
+      "step": 12255
+    },
+    {
+      "epoch": 2.19,
+      "grad_norm": 0.7465589046478271,
+      "learning_rate": 2.978172585471241e-05,
+      "loss": 1.9998,
+      "step": 12260
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.5380420684814453,
+      "learning_rate": 2.976792456453823e-05,
+      "loss": 1.8497,
+      "step": 12265
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.618114709854126,
+      "learning_rate": 2.975412176627409e-05,
+      "loss": 1.7707,
+      "step": 12270
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.5173262357711792,
+      "learning_rate": 2.9740317464285816e-05,
+      "loss": 1.8605,
+      "step": 12275
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.9060313701629639,
+      "learning_rate": 2.9726511662939695e-05,
+      "loss": 1.8223,
+      "step": 12280
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.1166000366210938,
+      "learning_rate": 2.971270436660247e-05,
+      "loss": 1.7321,
+      "step": 12285
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.9780591726303101,
+      "learning_rate": 2.969889557964139e-05,
+      "loss": 1.6634,
+      "step": 12290
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.0793691873550415,
+      "learning_rate": 2.9685085306424154e-05,
+      "loss": 2.0318,
+      "step": 12295
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.39177799224853516,
+      "learning_rate": 2.967127355131894e-05,
+      "loss": 1.8835,
+      "step": 12300
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.9409979581832886,
+      "learning_rate": 2.9657460318694407e-05,
+      "loss": 1.7953,
+      "step": 12305
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.2627776861190796,
+      "learning_rate": 2.9643645612919646e-05,
+      "loss": 1.9271,
+      "step": 12310
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.5884200930595398,
+      "learning_rate": 2.9629829438364248e-05,
+      "loss": 1.6694,
+      "step": 12315
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.5460672378540039,
+      "learning_rate": 2.9616011799398253e-05,
+      "loss": 1.9784,
+      "step": 12320
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.2458994388580322,
+      "learning_rate": 2.9602192700392168e-05,
+      "loss": 2.0202,
+      "step": 12325
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 1.5217969417572021,
+      "learning_rate": 2.9588372145716948e-05,
+      "loss": 1.7552,
+      "step": 12330
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.5435410737991333,
+      "learning_rate": 2.957455013974404e-05,
+      "loss": 1.8492,
+      "step": 12335
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.8544949293136597,
+      "learning_rate": 2.956072668684532e-05,
+      "loss": 1.9221,
+      "step": 12340
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.841983437538147,
+      "learning_rate": 2.9546901791393134e-05,
+      "loss": 1.8283,
+      "step": 12345
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6002446413040161,
+      "learning_rate": 2.9533075457760296e-05,
+      "loss": 1.7687,
+      "step": 12350
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6456916332244873,
+      "learning_rate": 2.951924769032004e-05,
+      "loss": 2.0222,
+      "step": 12355
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.909529447555542,
+      "learning_rate": 2.9505418493446087e-05,
+      "loss": 1.9912,
+      "step": 12360
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.6185993552207947,
+      "learning_rate": 2.9491587871512598e-05,
+      "loss": 1.5441,
+      "step": 12365
+    },
+    {
+      "epoch": 2.21,
+      "grad_norm": 0.7631831169128418,
+      "learning_rate": 2.947775582889419e-05,
+      "loss": 2.0657,
+      "step": 12370
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7581762075424194,
+      "learning_rate": 2.9463922369965917e-05,
+      "loss": 1.6854,
+      "step": 12375
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6361576318740845,
+      "learning_rate": 2.9450087499103297e-05,
+      "loss": 2.0618,
+      "step": 12380
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.9206913113594055,
+      "learning_rate": 2.9436251220682275e-05,
+      "loss": 1.8886,
+      "step": 12385
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7134048342704773,
+      "learning_rate": 2.9422413539079263e-05,
+      "loss": 1.6918,
+      "step": 12390
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.76563960313797,
+      "learning_rate": 2.9408574458671106e-05,
+      "loss": 1.9828,
+      "step": 12395
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.9646211862564087,
+      "learning_rate": 2.9394733983835082e-05,
+      "loss": 1.8371,
+      "step": 12400
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6462224125862122,
+      "learning_rate": 2.938089211894892e-05,
+      "loss": 1.7227,
+      "step": 12405
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.7270526885986328,
+      "learning_rate": 2.9367048868390812e-05,
+      "loss": 1.8023,
+      "step": 12410
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6550453305244446,
+      "learning_rate": 2.9353204236539334e-05,
+      "loss": 1.9456,
+      "step": 12415
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.5939522385597229,
+      "learning_rate": 2.933935822777355e-05,
+      "loss": 2.0319,
+      "step": 12420
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.8000824451446533,
+      "learning_rate": 2.932551084647293e-05,
+      "loss": 1.7618,
+      "step": 12425
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.1937888860702515,
+      "learning_rate": 2.931166209701739e-05,
+      "loss": 1.9905,
+      "step": 12430
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.8587405681610107,
+      "learning_rate": 2.929781198378727e-05,
+      "loss": 1.8422,
+      "step": 12435
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.4814565181732178,
+      "learning_rate": 2.928396051116336e-05,
+      "loss": 1.8622,
+      "step": 12440
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6059118509292603,
+      "learning_rate": 2.9270107683526866e-05,
+      "loss": 1.7572,
+      "step": 12445
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7698599696159363,
+      "learning_rate": 2.9256253505259406e-05,
+      "loss": 1.8953,
+      "step": 12450
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.1599043607711792,
+      "learning_rate": 2.924239798074307e-05,
+      "loss": 1.6808,
+      "step": 12455
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7120723128318787,
+      "learning_rate": 2.922854111436032e-05,
+      "loss": 1.7554,
+      "step": 12460
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.5135558247566223,
+      "learning_rate": 2.9214682910494096e-05,
+      "loss": 2.0082,
+      "step": 12465
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.6316121816635132,
+      "learning_rate": 2.9200823373527713e-05,
+      "loss": 2.05,
+      "step": 12470
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.9234532713890076,
+      "learning_rate": 2.918696250784494e-05,
+      "loss": 1.6301,
+      "step": 12475
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 0.7441158294677734,
+      "learning_rate": 2.917310031782995e-05,
+      "loss": 2.0701,
+      "step": 12480
+    },
+    {
+      "epoch": 2.23,
+      "grad_norm": 1.686782717704773,
+      "learning_rate": 2.9159236807867345e-05,
+      "loss": 1.6456,
+      "step": 12485
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.481157511472702,
+      "learning_rate": 2.9145371982342145e-05,
+      "loss": 2.0247,
+      "step": 12490
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.3480195999145508,
+      "learning_rate": 2.913150584563977e-05,
+      "loss": 1.7576,
+      "step": 12495
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.9952110052108765,
+      "learning_rate": 2.911763840214607e-05,
+      "loss": 1.8107,
+      "step": 12500
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.2428638935089111,
+      "learning_rate": 2.91037696562473e-05,
+      "loss": 1.8556,
+      "step": 12505
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.8979426026344299,
+      "learning_rate": 2.908989961233014e-05,
+      "loss": 1.8072,
+      "step": 12510
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.5153095722198486,
+      "learning_rate": 2.9076028274781663e-05,
+      "loss": 2.073,
+      "step": 12515
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7544877529144287,
+      "learning_rate": 2.9062155647989364e-05,
+      "loss": 1.8057,
+      "step": 12520
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.3223302364349365,
+      "learning_rate": 2.904828173634114e-05,
+      "loss": 1.9224,
+      "step": 12525
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.5328083634376526,
+      "learning_rate": 2.90344065442253e-05,
+      "loss": 1.9428,
+      "step": 12530
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.7167631387710571,
+      "learning_rate": 2.902053007603055e-05,
+      "loss": 2.0441,
+      "step": 12535
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 1.436819314956665,
+      "learning_rate": 2.9006652336146e-05,
+      "loss": 2.0272,
+      "step": 12540
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.5790032148361206,
+      "learning_rate": 2.8992773328961177e-05,
+      "loss": 1.9511,
+      "step": 12545
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.8261239528656006,
+      "learning_rate": 2.8978893058865987e-05,
+      "loss": 1.8447,
+      "step": 12550
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.655191957950592,
+      "learning_rate": 2.896501153025075e-05,
+      "loss": 1.9432,
+      "step": 12555
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.519561231136322,
+      "learning_rate": 2.8951128747506184e-05,
+      "loss": 2.0886,
+      "step": 12560
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7673883438110352,
+      "learning_rate": 2.8937244715023386e-05,
+      "loss": 1.8421,
+      "step": 12565
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6510341167449951,
+      "learning_rate": 2.8923359437193877e-05,
+      "loss": 1.6667,
+      "step": 12570
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6979449987411499,
+      "learning_rate": 2.8909472918409552e-05,
+      "loss": 1.8703,
+      "step": 12575
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.6348164677619934,
+      "learning_rate": 2.88955851630627e-05,
+      "loss": 1.782,
+      "step": 12580
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.9662460088729858,
+      "learning_rate": 2.8881696175546014e-05,
+      "loss": 1.9121,
+      "step": 12585
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.7864895462989807,
+      "learning_rate": 2.886780596025256e-05,
+      "loss": 1.7273,
+      "step": 12590
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 1.4080077409744263,
+      "learning_rate": 2.8853914521575797e-05,
+      "loss": 1.6424,
+      "step": 12595
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.8357173800468445,
+      "learning_rate": 2.8840021863909583e-05,
+      "loss": 1.9837,
+      "step": 12600
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.4318985641002655,
+      "learning_rate": 2.882612799164815e-05,
+      "loss": 1.7542,
+      "step": 12605
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.6467979550361633,
+      "learning_rate": 2.8812232909186103e-05,
+      "loss": 1.8621,
+      "step": 12610
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7634948492050171,
+      "learning_rate": 2.8798336620918464e-05,
+      "loss": 1.719,
+      "step": 12615
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7211525440216064,
+      "learning_rate": 2.878443913124059e-05,
+      "loss": 1.7789,
+      "step": 12620
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.5822710394859314,
+      "learning_rate": 2.8770540444548272e-05,
+      "loss": 1.9444,
+      "step": 12625
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.549856424331665,
+      "learning_rate": 2.8756640565237637e-05,
+      "loss": 2.1138,
+      "step": 12630
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.3974264860153198,
+      "learning_rate": 2.8742739497705208e-05,
+      "loss": 1.9182,
+      "step": 12635
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.5173325538635254,
+      "learning_rate": 2.8728837246347867e-05,
+      "loss": 1.8516,
+      "step": 12640
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.45196789503097534,
+      "learning_rate": 2.8714933815562894e-05,
+      "loss": 1.7245,
+      "step": 12645
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 0.7211712598800659,
+      "learning_rate": 2.870102920974793e-05,
+      "loss": 1.924,
+      "step": 12650
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7739723920822144,
+      "learning_rate": 2.868712343330099e-05,
+      "loss": 1.7423,
+      "step": 12655
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.738148033618927,
+      "learning_rate": 2.8673216490620452e-05,
+      "loss": 1.9875,
+      "step": 12660
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.5198020935058594,
+      "learning_rate": 2.8659308386105065e-05,
+      "loss": 1.8309,
+      "step": 12665
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.786413848400116,
+      "learning_rate": 2.864539912415396e-05,
+      "loss": 2.0195,
+      "step": 12670
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.3828994035720825,
+      "learning_rate": 2.8631488709166625e-05,
+      "loss": 1.9273,
+      "step": 12675
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7168930172920227,
+      "learning_rate": 2.8617577145542902e-05,
+      "loss": 1.8926,
+      "step": 12680
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.955230176448822,
+      "learning_rate": 2.8603664437682998e-05,
+      "loss": 1.6857,
+      "step": 12685
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.8382362127304077,
+      "learning_rate": 2.8589750589987506e-05,
+      "loss": 2.0365,
+      "step": 12690
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.8835332989692688,
+      "learning_rate": 2.8575835606857356e-05,
+      "loss": 2.0321,
+      "step": 12695
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.4743799865245819,
+      "learning_rate": 2.8561919492693845e-05,
+      "loss": 2.0926,
+      "step": 12700
+    },
+    {
+      "epoch": 2.27,
+      "grad_norm": 0.7596274614334106,
+      "learning_rate": 2.854800225189862e-05,
+      "loss": 1.8994,
+      "step": 12705
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7106197476387024,
+      "learning_rate": 2.8534083888873697e-05,
+      "loss": 1.7113,
+      "step": 12710
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.7384430766105652,
+      "learning_rate": 2.8520164408021432e-05,
+      "loss": 1.742,
+      "step": 12715
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.9046937227249146,
+      "learning_rate": 2.850624381374456e-05,
+      "loss": 1.9561,
+      "step": 12720
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.8590306639671326,
+      "learning_rate": 2.8492322110446146e-05,
+      "loss": 1.8945,
+      "step": 12725
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.8791319727897644,
+      "learning_rate": 2.8478399302529597e-05,
+      "loss": 1.8657,
+      "step": 12730
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6100798845291138,
+      "learning_rate": 2.84644753943987e-05,
+      "loss": 1.8644,
+      "step": 12735
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6638292670249939,
+      "learning_rate": 2.8450550390457565e-05,
+      "loss": 1.8004,
+      "step": 12740
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.814220666885376,
+      "learning_rate": 2.8436624295110663e-05,
+      "loss": 1.885,
+      "step": 12745
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 1.095098614692688,
+      "learning_rate": 2.8422697112762803e-05,
+      "loss": 2.1555,
+      "step": 12750
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.6496490836143494,
+      "learning_rate": 2.8408768847819133e-05,
+      "loss": 1.7966,
+      "step": 12755
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.9682343006134033,
+      "learning_rate": 2.8394839504685162e-05,
+      "loss": 1.8505,
+      "step": 12760
+    },
+    {
+      "epoch": 2.28,
+      "grad_norm": 0.9972975850105286,
+      "learning_rate": 2.8380909087766717e-05,
+      "loss": 1.7333,
+      "step": 12765
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.9131410121917725,
+      "learning_rate": 2.8366977601469986e-05,
+      "loss": 1.9954,
+      "step": 12770
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.8180422782897949,
+      "learning_rate": 2.8353045050201465e-05,
+      "loss": 1.7182,
+      "step": 12775
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.5822163820266724,
+      "learning_rate": 2.8339111438368032e-05,
+      "loss": 1.6172,
+      "step": 12780
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.9533834457397461,
+      "learning_rate": 2.832517677037686e-05,
+      "loss": 1.971,
+      "step": 12785
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.9467795491218567,
+      "learning_rate": 2.8311241050635467e-05,
+      "loss": 1.8465,
+      "step": 12790
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.5219370722770691,
+      "learning_rate": 2.8297304283551728e-05,
+      "loss": 1.932,
+      "step": 12795
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.45871034264564514,
+      "learning_rate": 2.8283366473533808e-05,
+      "loss": 2.0782,
+      "step": 12800
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.48181986808776855,
+      "learning_rate": 2.8269427624990224e-05,
+      "loss": 2.0225,
+      "step": 12805
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6003161668777466,
+      "learning_rate": 2.8255487742329838e-05,
+      "loss": 1.7363,
+      "step": 12810
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.6212576031684875,
+      "learning_rate": 2.8241546829961808e-05,
+      "loss": 1.887,
+      "step": 12815
+    },
+    {
+      "epoch": 2.29,
+      "grad_norm": 0.5601562261581421,
+      "learning_rate": 2.8227604892295628e-05,
+      "loss": 1.8317,
+      "step": 12820
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.9833465814590454,
+      "learning_rate": 2.8213661933741132e-05,
+      "loss": 1.8634,
+      "step": 12825
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.311710357666016,
+      "learning_rate": 2.8199717958708455e-05,
+      "loss": 1.6882,
+      "step": 12830
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.8071384429931641,
+      "learning_rate": 2.818577297160806e-05,
+      "loss": 2.0808,
+      "step": 12835
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.8173998594284058,
+      "learning_rate": 2.8171826976850753e-05,
+      "loss": 1.7736,
+      "step": 12840
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.7156624794006348,
+      "learning_rate": 2.815787997884761e-05,
+      "loss": 1.89,
+      "step": 12845
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.1727192401885986,
+      "learning_rate": 2.8143931982010068e-05,
+      "loss": 1.8794,
+      "step": 12850
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.5154191255569458,
+      "learning_rate": 2.8129982990749872e-05,
+      "loss": 1.9237,
+      "step": 12855
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.5282955169677734,
+      "learning_rate": 2.8116033009479058e-05,
+      "loss": 1.9519,
+      "step": 12860
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 0.8198376297950745,
+      "learning_rate": 2.8102082042610007e-05,
+      "loss": 1.8245,
+      "step": 12865
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.2161682844161987,
+      "learning_rate": 2.8088130094555394e-05,
+      "loss": 1.7885,
+      "step": 12870
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 4.205202579498291,
+      "learning_rate": 2.8074177169728188e-05,
+      "loss": 1.9236,
+      "step": 12875
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.6405089497566223,
+      "learning_rate": 2.8060223272541706e-05,
+      "loss": 1.863,
+      "step": 12880
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.1483161449432373,
+      "learning_rate": 2.8046268407409555e-05,
+      "loss": 1.8218,
+      "step": 12885
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 1.4012528657913208,
+      "learning_rate": 2.8032312578745634e-05,
+      "loss": 1.7995,
+      "step": 12890
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.5227799415588379,
+      "learning_rate": 2.8018355790964153e-05,
+      "loss": 2.0666,
+      "step": 12895
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7263579368591309,
+      "learning_rate": 2.8004398048479646e-05,
+      "loss": 1.5815,
+      "step": 12900
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.8712220191955566,
+      "learning_rate": 2.799043935570692e-05,
+      "loss": 1.6493,
+      "step": 12905
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.4584944248199463,
+      "learning_rate": 2.797647971706111e-05,
+      "loss": 2.0632,
+      "step": 12910
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.9673914909362793,
+      "learning_rate": 2.7962519136957625e-05,
+      "loss": 1.7794,
+      "step": 12915
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.5151163339614868,
+      "learning_rate": 2.7948557619812183e-05,
+      "loss": 1.8515,
+      "step": 12920
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.7004354596138,
+      "learning_rate": 2.7934595170040802e-05,
+      "loss": 1.7434,
+      "step": 12925
+    },
+    {
+      "epoch": 2.31,
+      "grad_norm": 0.5903719663619995,
+      "learning_rate": 2.7920631792059797e-05,
+      "loss": 1.9251,
+      "step": 12930
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.2344070672988892,
+      "learning_rate": 2.7906667490285765e-05,
+      "loss": 1.7713,
+      "step": 12935
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.2571988105773926,
+      "learning_rate": 2.7892702269135595e-05,
+      "loss": 1.5384,
+      "step": 12940
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.5835108757019043,
+      "learning_rate": 2.787873613302649e-05,
+      "loss": 1.7925,
+      "step": 12945
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.5990691184997559,
+      "learning_rate": 2.7864769086375903e-05,
+      "loss": 1.7678,
+      "step": 12950
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.5972462892532349,
+      "learning_rate": 2.7850801133601616e-05,
+      "loss": 1.8765,
+      "step": 12955
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 2.4750313758850098,
+      "learning_rate": 2.7836832279121678e-05,
+      "loss": 2.0151,
+      "step": 12960
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.888633131980896,
+      "learning_rate": 2.782286252735441e-05,
+      "loss": 1.9451,
+      "step": 12965
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.58510422706604,
+      "learning_rate": 2.7808891882718445e-05,
+      "loss": 2.0672,
+      "step": 12970
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7262510657310486,
+      "learning_rate": 2.7794920349632675e-05,
+      "loss": 1.9223,
+      "step": 12975
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.7543483972549438,
+      "learning_rate": 2.7780947932516294e-05,
+      "loss": 1.9088,
+      "step": 12980
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 5.534985542297363,
+      "learning_rate": 2.7766974635788743e-05,
+      "loss": 2.0211,
+      "step": 12985
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6918667554855347,
+      "learning_rate": 2.7753000463869794e-05,
+      "loss": 1.7378,
+      "step": 12990
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7650964260101318,
+      "learning_rate": 2.7739025421179433e-05,
+      "loss": 1.5172,
+      "step": 12995
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.662441611289978,
+      "learning_rate": 2.7725049512137967e-05,
+      "loss": 1.9162,
+      "step": 13000
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6917517781257629,
+      "learning_rate": 2.771107274116597e-05,
+      "loss": 1.7873,
+      "step": 13005
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.6874393820762634,
+      "learning_rate": 2.769709511268427e-05,
+      "loss": 1.9959,
+      "step": 13010
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.8905332684516907,
+      "learning_rate": 2.7683116631113976e-05,
+      "loss": 1.9593,
+      "step": 13015
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.5240230560302734,
+      "learning_rate": 2.7669137300876467e-05,
+      "loss": 1.6583,
+      "step": 13020
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 1.5556925535202026,
+      "learning_rate": 2.7655157126393404e-05,
+      "loss": 1.6085,
+      "step": 13025
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.9076879620552063,
+      "learning_rate": 2.7641176112086698e-05,
+      "loss": 1.6375,
+      "step": 13030
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.4439728260040283,
+      "learning_rate": 2.762719426237853e-05,
+      "loss": 1.9813,
+      "step": 13035
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.7149770259857178,
+      "learning_rate": 2.761321158169134e-05,
+      "loss": 1.9144,
+      "step": 13040
+    },
+    {
+      "epoch": 2.33,
+      "grad_norm": 0.49863100051879883,
+      "learning_rate": 2.759922807444784e-05,
+      "loss": 2.1206,
+      "step": 13045
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.2692310810089111,
+      "learning_rate": 2.7585243745071003e-05,
+      "loss": 1.9092,
+      "step": 13050
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.9751031398773193,
+      "learning_rate": 2.757125859798405e-05,
+      "loss": 2.052,
+      "step": 13055
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.8686983585357666,
+      "learning_rate": 2.7557272637610483e-05,
+      "loss": 1.8621,
+      "step": 13060
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.48079320788383484,
+      "learning_rate": 2.7543285868374037e-05,
+      "loss": 1.6149,
+      "step": 13065
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.8371298909187317,
+      "learning_rate": 2.7529298294698712e-05,
+      "loss": 1.9314,
+      "step": 13070
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.3321702182292938,
+      "learning_rate": 2.7515309921008775e-05,
+      "loss": 1.8415,
+      "step": 13075
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.8658364415168762,
+      "learning_rate": 2.7501320751728732e-05,
+      "loss": 1.7368,
+      "step": 13080
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.8022560477256775,
+      "learning_rate": 2.7487330791283335e-05,
+      "loss": 1.6356,
+      "step": 13085
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.0247143507003784,
+      "learning_rate": 2.7473340044097605e-05,
+      "loss": 1.8981,
+      "step": 13090
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.5062198042869568,
+      "learning_rate": 2.7459348514596796e-05,
+      "loss": 2.1258,
+      "step": 13095
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 6.957434177398682,
+      "learning_rate": 2.7445356207206424e-05,
+      "loss": 1.7353,
+      "step": 13100
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.8973393440246582,
+      "learning_rate": 2.7431363126352232e-05,
+      "loss": 1.5304,
+      "step": 13105
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7698760628700256,
+      "learning_rate": 2.7417369276460215e-05,
+      "loss": 1.8007,
+      "step": 13110
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6482462882995605,
+      "learning_rate": 2.7403374661956626e-05,
+      "loss": 1.9363,
+      "step": 13115
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.528232216835022,
+      "learning_rate": 2.7389379287267953e-05,
+      "loss": 1.5902,
+      "step": 13120
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.5823779702186584,
+      "learning_rate": 2.7375383156820895e-05,
+      "loss": 1.8967,
+      "step": 13125
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.153917908668518,
+      "learning_rate": 2.7361386275042434e-05,
+      "loss": 1.8673,
+      "step": 13130
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.1044551134109497,
+      "learning_rate": 2.734738864635977e-05,
+      "loss": 1.9955,
+      "step": 13135
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.7324090600013733,
+      "learning_rate": 2.733339027520032e-05,
+      "loss": 1.8696,
+      "step": 13140
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6724053621292114,
+      "learning_rate": 2.7319391165991786e-05,
+      "loss": 2.0311,
+      "step": 13145
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 1.5398845672607422,
+      "learning_rate": 2.7305391323162056e-05,
+      "loss": 1.8971,
+      "step": 13150
+    },
+    {
+      "epoch": 2.35,
+      "grad_norm": 0.6307987570762634,
+      "learning_rate": 2.7291390751139263e-05,
+      "loss": 1.998,
+      "step": 13155
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.5545775890350342,
+      "learning_rate": 2.7277389454351777e-05,
+      "loss": 1.9609,
+      "step": 13160
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.9437121748924255,
+      "learning_rate": 2.7263387437228204e-05,
+      "loss": 1.9158,
+      "step": 13165
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.5333415865898132,
+      "learning_rate": 2.724938470419736e-05,
+      "loss": 2.0586,
+      "step": 13170
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.7932983636856079,
+      "learning_rate": 2.72353812596883e-05,
+      "loss": 1.9107,
+      "step": 13175
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.802893877029419,
+      "learning_rate": 2.7221377108130292e-05,
+      "loss": 2.0469,
+      "step": 13180
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.4990454614162445,
+      "learning_rate": 2.720737225395284e-05,
+      "loss": 1.9607,
+      "step": 13185
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.1566194295883179,
+      "learning_rate": 2.7193366701585677e-05,
+      "loss": 1.781,
+      "step": 13190
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.773506224155426,
+      "learning_rate": 2.717936045545873e-05,
+      "loss": 1.8001,
+      "step": 13195
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.6509556174278259,
+      "learning_rate": 2.7165353520002167e-05,
+      "loss": 1.9832,
+      "step": 13200
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.9481406807899475,
+      "learning_rate": 2.7151345899646363e-05,
+      "loss": 1.8591,
+      "step": 13205
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.6072911620140076,
+      "learning_rate": 2.7137337598821932e-05,
+      "loss": 1.8942,
+      "step": 13210
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.9392735362052917,
+      "learning_rate": 2.712332862195967e-05,
+      "loss": 1.8107,
+      "step": 13215
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.8217921257019043,
+      "learning_rate": 2.7109318973490606e-05,
+      "loss": 1.8219,
+      "step": 13220
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.2502552270889282,
+      "learning_rate": 2.709530865784598e-05,
+      "loss": 2.1015,
+      "step": 13225
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.7115715742111206,
+      "learning_rate": 2.7081297679457236e-05,
+      "loss": 1.8984,
+      "step": 13230
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6438332796096802,
+      "learning_rate": 2.706728604275604e-05,
+      "loss": 1.9235,
+      "step": 13235
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.5191622376441956,
+      "learning_rate": 2.705327375217427e-05,
+      "loss": 1.9651,
+      "step": 13240
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 1.5185312032699585,
+      "learning_rate": 2.703926081214397e-05,
+      "loss": 1.9163,
+      "step": 13245
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.4574000835418701,
+      "learning_rate": 2.702524722709744e-05,
+      "loss": 1.8515,
+      "step": 13250
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.6382564902305603,
+      "learning_rate": 2.7011233001467166e-05,
+      "loss": 2.0203,
+      "step": 13255
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.9020372629165649,
+      "learning_rate": 2.6997218139685825e-05,
+      "loss": 1.7434,
+      "step": 13260
+    },
+    {
+      "epoch": 2.37,
+      "grad_norm": 0.9500882625579834,
+      "learning_rate": 2.6983202646186306e-05,
+      "loss": 1.6759,
+      "step": 13265
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.5408741235733032,
+      "learning_rate": 2.69691865254017e-05,
+      "loss": 1.9249,
+      "step": 13270
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.5403386950492859,
+      "learning_rate": 2.695516978176529e-05,
+      "loss": 1.9793,
+      "step": 13275
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.1900765895843506,
+      "learning_rate": 2.694115241971056e-05,
+      "loss": 1.5909,
+      "step": 13280
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.8298088908195496,
+      "learning_rate": 2.6927134443671187e-05,
+      "loss": 1.6979,
+      "step": 13285
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.30263030529022217,
+      "learning_rate": 2.6913115858081038e-05,
+      "loss": 2.0684,
+      "step": 13290
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.6629304885864258,
+      "learning_rate": 2.6899096667374187e-05,
+      "loss": 1.7992,
+      "step": 13295
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.8511559367179871,
+      "learning_rate": 2.6885076875984888e-05,
+      "loss": 1.8706,
+      "step": 13300
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.7309024333953857,
+      "learning_rate": 2.6871056488347583e-05,
+      "loss": 1.6392,
+      "step": 13305
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.796414315700531,
+      "learning_rate": 2.6857035508896917e-05,
+      "loss": 1.8747,
+      "step": 13310
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 0.9297599792480469,
+      "learning_rate": 2.6843013942067702e-05,
+      "loss": 1.527,
+      "step": 13315
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.2013111114501953,
+      "learning_rate": 2.6828991792294943e-05,
+      "loss": 1.8731,
+      "step": 13320
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.5262635946273804,
+      "learning_rate": 2.6814969064013846e-05,
+      "loss": 1.898,
+      "step": 13325
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.6759445071220398,
+      "learning_rate": 2.6800945761659784e-05,
+      "loss": 1.9081,
+      "step": 13330
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.5417052507400513,
+      "learning_rate": 2.6786921889668303e-05,
+      "loss": 1.689,
+      "step": 13335
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7009138464927673,
+      "learning_rate": 2.677289745247515e-05,
+      "loss": 1.9229,
+      "step": 13340
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.689923882484436,
+      "learning_rate": 2.675887245451624e-05,
+      "loss": 1.8571,
+      "step": 13345
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.4503270387649536,
+      "learning_rate": 2.6744846900227654e-05,
+      "loss": 1.8241,
+      "step": 13350
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 1.711902141571045,
+      "learning_rate": 2.673082079404568e-05,
+      "loss": 1.7829,
+      "step": 13355
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.37441059947013855,
+      "learning_rate": 2.6716794140406758e-05,
+      "loss": 1.8407,
+      "step": 13360
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.5013666749000549,
+      "learning_rate": 2.6702766943747493e-05,
+      "loss": 1.9828,
+      "step": 13365
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.7650493383407593,
+      "learning_rate": 2.6688739208504686e-05,
+      "loss": 1.7357,
+      "step": 13370
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.6591855883598328,
+      "learning_rate": 2.667471093911529e-05,
+      "loss": 2.0887,
+      "step": 13375
+    },
+    {
+      "epoch": 2.39,
+      "grad_norm": 0.5767456889152527,
+      "learning_rate": 2.6660682140016435e-05,
+      "loss": 1.9841,
+      "step": 13380
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.6293467283248901,
+      "learning_rate": 2.6646652815645412e-05,
+      "loss": 2.0894,
+      "step": 13385
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.4856579601764679,
+      "learning_rate": 2.6632622970439698e-05,
+      "loss": 1.8616,
+      "step": 13390
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.9088009595870972,
+      "learning_rate": 2.6618592608836896e-05,
+      "loss": 1.8978,
+      "step": 13395
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.0228464603424072,
+      "learning_rate": 2.660456173527482e-05,
+      "loss": 1.7805,
+      "step": 13400
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.8853181004524231,
+      "learning_rate": 2.6590530354191405e-05,
+      "loss": 1.683,
+      "step": 13405
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7312606573104858,
+      "learning_rate": 2.657649847002477e-05,
+      "loss": 1.9513,
+      "step": 13410
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7548524737358093,
+      "learning_rate": 2.6562466087213188e-05,
+      "loss": 1.7518,
+      "step": 13415
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.1628096103668213,
+      "learning_rate": 2.654843321019508e-05,
+      "loss": 1.9882,
+      "step": 13420
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.8537325263023376,
+      "learning_rate": 2.653439984340904e-05,
+      "loss": 2.1221,
+      "step": 13425
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.9286057949066162,
+      "learning_rate": 2.65203659912938e-05,
+      "loss": 1.7583,
+      "step": 13430
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.7813141942024231,
+      "learning_rate": 2.650633165828827e-05,
+      "loss": 1.7862,
+      "step": 13435
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.770072877407074,
+      "learning_rate": 2.6492296848831476e-05,
+      "loss": 1.7671,
+      "step": 13440
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7098547220230103,
+      "learning_rate": 2.6478261567362628e-05,
+      "loss": 2.02,
+      "step": 13445
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.0164227485656738,
+      "learning_rate": 2.646422581832107e-05,
+      "loss": 1.6729,
+      "step": 13450
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.548316478729248,
+      "learning_rate": 2.645018960614629e-05,
+      "loss": 1.694,
+      "step": 13455
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7778927683830261,
+      "learning_rate": 2.6436152935277936e-05,
+      "loss": 1.775,
+      "step": 13460
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.7858615517616272,
+      "learning_rate": 2.642211581015579e-05,
+      "loss": 1.6574,
+      "step": 13465
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.9521759748458862,
+      "learning_rate": 2.640807823521978e-05,
+      "loss": 1.7402,
+      "step": 13470
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 1.4299689531326294,
+      "learning_rate": 2.6394040214909976e-05,
+      "loss": 1.7257,
+      "step": 13475
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.9395617246627808,
+      "learning_rate": 2.6380001753666593e-05,
+      "loss": 1.8034,
+      "step": 13480
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.4695361256599426,
+      "learning_rate": 2.636596285592999e-05,
+      "loss": 1.8073,
+      "step": 13485
+    },
+    {
+      "epoch": 2.41,
+      "grad_norm": 0.31577566266059875,
+      "learning_rate": 2.635192352614064e-05,
+      "loss": 1.9157,
+      "step": 13490
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.4319940507411957,
+      "learning_rate": 2.6337883768739192e-05,
+      "loss": 1.969,
+      "step": 13495
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.9783098101615906,
+      "learning_rate": 2.632384358816638e-05,
+      "loss": 2.0406,
+      "step": 13500
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.060631513595581,
+      "learning_rate": 2.6309802988863124e-05,
+      "loss": 1.7076,
+      "step": 13505
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.9120562672615051,
+      "learning_rate": 2.6295761975270432e-05,
+      "loss": 1.958,
+      "step": 13510
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.6509984135627747,
+      "learning_rate": 2.628172055182948e-05,
+      "loss": 1.9328,
+      "step": 13515
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.6889690160751343,
+      "learning_rate": 2.626767872298156e-05,
+      "loss": 2.1166,
+      "step": 13520
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 1.167043924331665,
+      "learning_rate": 2.6253636493168065e-05,
+      "loss": 1.9546,
+      "step": 13525
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.5428257584571838,
+      "learning_rate": 2.623959386683056e-05,
+      "loss": 1.8359,
+      "step": 13530
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.911302387714386,
+      "learning_rate": 2.6225550848410706e-05,
+      "loss": 1.7099,
+      "step": 13535
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.8027520775794983,
+      "learning_rate": 2.6211507442350303e-05,
+      "loss": 1.9308,
+      "step": 13540
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 0.5085805058479309,
+      "learning_rate": 2.619746365309125e-05,
+      "loss": 1.7125,
+      "step": 13545
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.828308641910553,
+      "learning_rate": 2.6183419485075605e-05,
+      "loss": 2.0749,
+      "step": 13550
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.7583789825439453,
+      "learning_rate": 2.6169374942745505e-05,
+      "loss": 1.8622,
+      "step": 13555
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.5300992131233215,
+      "learning_rate": 2.615533003054324e-05,
+      "loss": 1.4851,
+      "step": 13560
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7071987986564636,
+      "learning_rate": 2.6141284752911205e-05,
+      "loss": 1.7153,
+      "step": 13565
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.1922173500061035,
+      "learning_rate": 2.612723911429189e-05,
+      "loss": 1.8535,
+      "step": 13570
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.631430447101593,
+      "learning_rate": 2.6113193119127934e-05,
+      "loss": 2.0645,
+      "step": 13575
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.7159907817840576,
+      "learning_rate": 2.6099146771862064e-05,
+      "loss": 1.748,
+      "step": 13580
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.622367262840271,
+      "learning_rate": 2.6085100076937126e-05,
+      "loss": 1.5421,
+      "step": 13585
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.9370682239532471,
+      "learning_rate": 2.6071053038796077e-05,
+      "loss": 1.8222,
+      "step": 13590
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 0.9740180969238281,
+      "learning_rate": 2.6057005661881982e-05,
+      "loss": 1.6439,
+      "step": 13595
+    },
+    {
+      "epoch": 2.43,
+      "grad_norm": 1.1773573160171509,
+      "learning_rate": 2.6042957950638013e-05,
+      "loss": 1.9287,
+      "step": 13600
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.8247451186180115,
+      "learning_rate": 2.602890990950745e-05,
+      "loss": 1.8199,
+      "step": 13605
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7261055111885071,
+      "learning_rate": 2.6014861542933676e-05,
+      "loss": 1.7212,
+      "step": 13610
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.8083186149597168,
+      "learning_rate": 2.600081285536016e-05,
+      "loss": 1.9574,
+      "step": 13615
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.9959443211555481,
+      "learning_rate": 2.5986763851230506e-05,
+      "loss": 1.982,
+      "step": 13620
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7056101560592651,
+      "learning_rate": 2.5972714534988403e-05,
+      "loss": 1.7576,
+      "step": 13625
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.7754709720611572,
+      "learning_rate": 2.5958664911077614e-05,
+      "loss": 1.8267,
+      "step": 13630
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.6835538744926453,
+      "learning_rate": 2.5944614983942044e-05,
+      "loss": 1.7875,
+      "step": 13635
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.8215372562408447,
+      "learning_rate": 2.5930564758025665e-05,
+      "loss": 1.462,
+      "step": 13640
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.608349621295929,
+      "learning_rate": 2.5916514237772543e-05,
+      "loss": 1.9472,
+      "step": 13645
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.9655773639678955,
+      "learning_rate": 2.590246342762685e-05,
+      "loss": 1.7016,
+      "step": 13650
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.6545687317848206,
+      "learning_rate": 2.5888412332032847e-05,
+      "loss": 1.7792,
+      "step": 13655
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.8544228672981262,
+      "learning_rate": 2.587436095543488e-05,
+      "loss": 1.7444,
+      "step": 13660
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.4463987946510315,
+      "learning_rate": 2.5860309302277373e-05,
+      "loss": 1.8348,
+      "step": 13665
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.6795549392700195,
+      "learning_rate": 2.5846257377004874e-05,
+      "loss": 1.95,
+      "step": 13670
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.8079219460487366,
+      "learning_rate": 2.583220518406197e-05,
+      "loss": 1.7795,
+      "step": 13675
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.4398018419742584,
+      "learning_rate": 2.581815272789337e-05,
+      "loss": 2.0165,
+      "step": 13680
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.9223191738128662,
+      "learning_rate": 2.5804100012943855e-05,
+      "loss": 1.7688,
+      "step": 13685
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.9821711778640747,
+      "learning_rate": 2.579004704365827e-05,
+      "loss": 1.8864,
+      "step": 13690
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.7216715812683105,
+      "learning_rate": 2.5775993824481565e-05,
+      "loss": 1.8994,
+      "step": 13695
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.4654800295829773,
+      "learning_rate": 2.5761940359858767e-05,
+      "loss": 2.0082,
+      "step": 13700
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.3378876447677612,
+      "learning_rate": 2.5747886654234967e-05,
+      "loss": 1.799,
+      "step": 13705
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 1.4554535150527954,
+      "learning_rate": 2.5733832712055328e-05,
+      "loss": 1.8686,
+      "step": 13710
+    },
+    {
+      "epoch": 2.45,
+      "grad_norm": 0.4863360822200775,
+      "learning_rate": 2.5719778537765115e-05,
+      "loss": 2.0027,
+      "step": 13715
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 6.204029560089111,
+      "learning_rate": 2.570572413580964e-05,
+      "loss": 2.0457,
+      "step": 13720
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.010169267654419,
+      "learning_rate": 2.5691669510634302e-05,
+      "loss": 1.829,
+      "step": 13725
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.463751882314682,
+      "learning_rate": 2.5677614666684567e-05,
+      "loss": 1.9938,
+      "step": 13730
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.6773427724838257,
+      "learning_rate": 2.566355960840596e-05,
+      "loss": 2.0168,
+      "step": 13735
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.6346594095230103,
+      "learning_rate": 2.5649504340244085e-05,
+      "loss": 1.7716,
+      "step": 13740
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.40613624453544617,
+      "learning_rate": 2.5635448866644613e-05,
+      "loss": 1.7708,
+      "step": 13745
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7240647077560425,
+      "learning_rate": 2.5621393192053278e-05,
+      "loss": 2.1106,
+      "step": 13750
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.6415520906448364,
+      "learning_rate": 2.560733732091587e-05,
+      "loss": 1.5509,
+      "step": 13755
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.5846617221832275,
+      "learning_rate": 2.559328125767826e-05,
+      "loss": 1.7258,
+      "step": 13760
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.5477493405342102,
+      "learning_rate": 2.557922500678635e-05,
+      "loss": 1.7324,
+      "step": 13765
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.5833098292350769,
+      "learning_rate": 2.556516857268613e-05,
+      "loss": 1.9903,
+      "step": 13770
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.5727924108505249,
+      "learning_rate": 2.555111195982364e-05,
+      "loss": 1.9487,
+      "step": 13775
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.5207077264785767,
+      "learning_rate": 2.553705517264496e-05,
+      "loss": 1.9171,
+      "step": 13780
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.6740149259567261,
+      "learning_rate": 2.5522998215596245e-05,
+      "loss": 1.8721,
+      "step": 13785
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 2.3125998973846436,
+      "learning_rate": 2.55089410931237e-05,
+      "loss": 1.8894,
+      "step": 13790
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.8759080767631531,
+      "learning_rate": 2.5494883809673576e-05,
+      "loss": 1.7691,
+      "step": 13795
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.6559151411056519,
+      "learning_rate": 2.5480826369692178e-05,
+      "loss": 1.8659,
+      "step": 13800
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.1714446544647217,
+      "learning_rate": 2.5466768777625872e-05,
+      "loss": 1.8289,
+      "step": 13805
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 1.3722327947616577,
+      "learning_rate": 2.5452711037921033e-05,
+      "loss": 1.5297,
+      "step": 13810
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.6491648554801941,
+      "learning_rate": 2.5438653155024145e-05,
+      "loss": 1.8004,
+      "step": 13815
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.7832010984420776,
+      "learning_rate": 2.542459513338168e-05,
+      "loss": 1.7827,
+      "step": 13820
+    },
+    {
+      "epoch": 2.47,
+      "grad_norm": 0.854123055934906,
+      "learning_rate": 2.541053697744018e-05,
+      "loss": 2.0299,
+      "step": 13825
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.6451117992401123,
+      "learning_rate": 2.5396478691646227e-05,
+      "loss": 1.848,
+      "step": 13830
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.6874570846557617,
+      "learning_rate": 2.538242028044645e-05,
+      "loss": 1.7095,
+      "step": 13835
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.2778282165527344,
+      "learning_rate": 2.53683617482875e-05,
+      "loss": 1.7892,
+      "step": 13840
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7223300337791443,
+      "learning_rate": 2.535430309961609e-05,
+      "loss": 1.8236,
+      "step": 13845
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7660056948661804,
+      "learning_rate": 2.5340244338878948e-05,
+      "loss": 2.1062,
+      "step": 13850
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.7210371494293213,
+      "learning_rate": 2.5326185470522852e-05,
+      "loss": 1.7827,
+      "step": 13855
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.532279372215271,
+      "learning_rate": 2.5312126498994603e-05,
+      "loss": 1.7299,
+      "step": 13860
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.1334110498428345,
+      "learning_rate": 2.5298067428741034e-05,
+      "loss": 1.8157,
+      "step": 13865
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 2.2065303325653076,
+      "learning_rate": 2.5284008264209037e-05,
+      "loss": 1.842,
+      "step": 13870
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.4892088174819946,
+      "learning_rate": 2.5269949009845495e-05,
+      "loss": 2.1346,
+      "step": 13875
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.8326558470726013,
+      "learning_rate": 2.525588967009734e-05,
+      "loss": 1.7902,
+      "step": 13880
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.9677508473396301,
+      "learning_rate": 2.5241830249411523e-05,
+      "loss": 1.4476,
+      "step": 13885
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.45370617508888245,
+      "learning_rate": 2.5227770752235043e-05,
+      "loss": 2.0553,
+      "step": 13890
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.6809149980545044,
+      "learning_rate": 2.5213711183014888e-05,
+      "loss": 1.8837,
+      "step": 13895
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.7268977165222168,
+      "learning_rate": 2.5199651546198084e-05,
+      "loss": 1.9655,
+      "step": 13900
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.5356599688529968,
+      "learning_rate": 2.51855918462317e-05,
+      "loss": 1.7184,
+      "step": 13905
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.551061749458313,
+      "learning_rate": 2.5171532087562782e-05,
+      "loss": 1.8176,
+      "step": 13910
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.8018007278442383,
+      "learning_rate": 2.5157472274638437e-05,
+      "loss": 2.0117,
+      "step": 13915
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.2345116138458252,
+      "learning_rate": 2.5143412411905766e-05,
+      "loss": 1.6238,
+      "step": 13920
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.5574769377708435,
+      "learning_rate": 2.5129352503811876e-05,
+      "loss": 1.8545,
+      "step": 13925
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 0.5945510268211365,
+      "learning_rate": 2.5115292554803915e-05,
+      "loss": 1.8348,
+      "step": 13930
+    },
+    {
+      "epoch": 2.49,
+      "grad_norm": 1.0902730226516724,
+      "learning_rate": 2.5101232569329042e-05,
+      "loss": 1.7905,
+      "step": 13935
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.5143458843231201,
+      "learning_rate": 2.50871725518344e-05,
+      "loss": 1.7065,
+      "step": 13940
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.1873408555984497,
+      "learning_rate": 2.5073112506767156e-05,
+      "loss": 1.9102,
+      "step": 13945
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.836678147315979,
+      "learning_rate": 2.5059052438574504e-05,
+      "loss": 1.9174,
+      "step": 13950
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.9396674036979675,
+      "learning_rate": 2.5044992351703617e-05,
+      "loss": 1.8286,
+      "step": 13955
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.3221425712108612,
+      "learning_rate": 2.5030932250601696e-05,
+      "loss": 2.0022,
+      "step": 13960
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.404722809791565,
+      "learning_rate": 2.5016872139715934e-05,
+      "loss": 1.8733,
+      "step": 13965
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.7955238819122314,
+      "learning_rate": 2.5002812023493526e-05,
+      "loss": 1.7287,
+      "step": 13970
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.7894271016120911,
+      "learning_rate": 2.498875190638168e-05,
+      "loss": 1.9474,
+      "step": 13975
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.6397268176078796,
+      "learning_rate": 2.4974691792827598e-05,
+      "loss": 1.7299,
+      "step": 13980
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.6210111975669861,
+      "learning_rate": 2.4960631687278475e-05,
+      "loss": 1.9172,
+      "step": 13985
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.5007354617118835,
+      "learning_rate": 2.494657159418151e-05,
+      "loss": 2.2217,
+      "step": 13990
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.6257606744766235,
+      "learning_rate": 2.4932511517983906e-05,
+      "loss": 1.8129,
+      "step": 13995
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.7970540523529053,
+      "learning_rate": 2.4918451463132843e-05,
+      "loss": 1.7572,
+      "step": 14000
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.6435466408729553,
+      "learning_rate": 2.4904391434075495e-05,
+      "loss": 1.827,
+      "step": 14005
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.1177036762237549,
+      "learning_rate": 2.489033143525905e-05,
+      "loss": 1.8556,
+      "step": 14010
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.462189793586731,
+      "learning_rate": 2.4876271471130664e-05,
+      "loss": 1.8425,
+      "step": 14015
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.4156283438205719,
+      "learning_rate": 2.486221154613749e-05,
+      "loss": 1.9366,
+      "step": 14020
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.2464938163757324,
+      "learning_rate": 2.484815166472668e-05,
+      "loss": 1.6903,
+      "step": 14025
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.604915201663971,
+      "learning_rate": 2.4834091831345343e-05,
+      "loss": 1.6903,
+      "step": 14030
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.5225396156311035,
+      "learning_rate": 2.4820032050440602e-05,
+      "loss": 1.7955,
+      "step": 14035
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 0.6467166543006897,
+      "learning_rate": 2.4805972326459555e-05,
+      "loss": 1.9819,
+      "step": 14040
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.0003973245620728,
+      "learning_rate": 2.479191266384928e-05,
+      "loss": 1.5891,
+      "step": 14045
+    },
+    {
+      "epoch": 2.51,
+      "grad_norm": 1.0571430921554565,
+      "learning_rate": 2.4777853067056824e-05,
+      "loss": 1.7885,
+      "step": 14050
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.7798799872398376,
+      "learning_rate": 2.4763793540529226e-05,
+      "loss": 1.5714,
+      "step": 14055
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.5746772289276123,
+      "learning_rate": 2.4749734088713502e-05,
+      "loss": 1.8001,
+      "step": 14060
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.2530102729797363,
+      "learning_rate": 2.473567471605665e-05,
+      "loss": 1.8787,
+      "step": 14065
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.9434322714805603,
+      "learning_rate": 2.4721615427005628e-05,
+      "loss": 2.0167,
+      "step": 14070
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.5446343421936035,
+      "learning_rate": 2.4707556226007377e-05,
+      "loss": 1.5517,
+      "step": 14075
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.794992983341217,
+      "learning_rate": 2.469349711750881e-05,
+      "loss": 2.0354,
+      "step": 14080
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.5005180239677429,
+      "learning_rate": 2.467943810595682e-05,
+      "loss": 1.7495,
+      "step": 14085
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.5188912153244019,
+      "learning_rate": 2.4665379195798247e-05,
+      "loss": 1.722,
+      "step": 14090
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 3.6539127826690674,
+      "learning_rate": 2.4651320391479902e-05,
+      "loss": 1.621,
+      "step": 14095
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.638843297958374,
+      "learning_rate": 2.463726169744859e-05,
+      "loss": 1.8223,
+      "step": 14100
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.6514829397201538,
+      "learning_rate": 2.462320311815105e-05,
+      "loss": 1.6494,
+      "step": 14105
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.295366883277893,
+      "learning_rate": 2.4609144658034e-05,
+      "loss": 1.8201,
+      "step": 14110
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.8184346556663513,
+      "learning_rate": 2.4595086321544125e-05,
+      "loss": 2.0595,
+      "step": 14115
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.1989121437072754,
+      "learning_rate": 2.458102811312805e-05,
+      "loss": 1.9181,
+      "step": 14120
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.7824451923370361,
+      "learning_rate": 2.4566970037232385e-05,
+      "loss": 1.7497,
+      "step": 14125
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.6604727506637573,
+      "learning_rate": 2.4552912098303675e-05,
+      "loss": 2.0074,
+      "step": 14130
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.795128583908081,
+      "learning_rate": 2.453885430078845e-05,
+      "loss": 2.2564,
+      "step": 14135
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": Infinity,
+      "learning_rate": 2.4527608167581993e-05,
+      "loss": 2.0231,
+      "step": 14140
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 0.7836667895317078,
+      "learning_rate": 2.451355063581608e-05,
+      "loss": 1.9791,
+      "step": 14145
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.0901525020599365,
+      "learning_rate": 2.4499493257913634e-05,
+      "loss": 1.497,
+      "step": 14150
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 2.3279173374176025,
+      "learning_rate": 2.4485436038320983e-05,
+      "loss": 2.0759,
+      "step": 14155
+    },
+    {
+      "epoch": 2.53,
+      "grad_norm": 1.8435465097427368,
+      "learning_rate": 2.447137898148443e-05,
+      "loss": 2.0998,
+      "step": 14160
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.8504188060760498,
+      "learning_rate": 2.4457322091850183e-05,
+      "loss": 1.6399,
+      "step": 14165
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.5952614545822144,
+      "learning_rate": 2.444326537386444e-05,
+      "loss": 2.058,
+      "step": 14170
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.9884322285652161,
+      "learning_rate": 2.4429208831973333e-05,
+      "loss": 1.7262,
+      "step": 14175
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.2859538793563843,
+      "learning_rate": 2.441515247062291e-05,
+      "loss": 1.7738,
+      "step": 14180
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.47512462735176086,
+      "learning_rate": 2.4401096294259217e-05,
+      "loss": 1.8085,
+      "step": 14185
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.804999828338623,
+      "learning_rate": 2.438704030732818e-05,
+      "loss": 1.7905,
+      "step": 14190
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.0102812051773071,
+      "learning_rate": 2.43729845142757e-05,
+      "loss": 2.0173,
+      "step": 14195
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.7093918919563293,
+      "learning_rate": 2.4358928919547616e-05,
+      "loss": 1.8977,
+      "step": 14200
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.1343001127243042,
+      "learning_rate": 2.43448735275897e-05,
+      "loss": 1.5511,
+      "step": 14205
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 0.6553163528442383,
+      "learning_rate": 2.4330818342847655e-05,
+      "loss": 1.5871,
+      "step": 14210
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.924633502960205,
+      "learning_rate": 2.4316763369767127e-05,
+      "loss": 2.0279,
+      "step": 14215
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.7396385669708252,
+      "learning_rate": 2.4302708612793686e-05,
+      "loss": 1.8374,
+      "step": 14220
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.5724390149116516,
+      "learning_rate": 2.4288654076372842e-05,
+      "loss": 2.1751,
+      "step": 14225
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.5967565774917603,
+      "learning_rate": 2.4274599764950034e-05,
+      "loss": 1.8294,
+      "step": 14230
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.9949921369552612,
+      "learning_rate": 2.426054568297062e-05,
+      "loss": 1.8312,
+      "step": 14235
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.9570918083190918,
+      "learning_rate": 2.4246491834879892e-05,
+      "loss": 1.8949,
+      "step": 14240
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.5771768093109131,
+      "learning_rate": 2.423243822512307e-05,
+      "loss": 1.8138,
+      "step": 14245
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.2878669500350952,
+      "learning_rate": 2.4218384858145306e-05,
+      "loss": 1.7788,
+      "step": 14250
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.2565678358078003,
+      "learning_rate": 2.420433173839165e-05,
+      "loss": 1.6618,
+      "step": 14255
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 2.0233705043792725,
+      "learning_rate": 2.41902788703071e-05,
+      "loss": 1.8718,
+      "step": 14260
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 1.1918349266052246,
+      "learning_rate": 2.4176226258336574e-05,
+      "loss": 1.7568,
+      "step": 14265
+    },
+    {
+      "epoch": 2.55,
+      "grad_norm": 0.5747388601303101,
+      "learning_rate": 2.4162173906924882e-05,
+      "loss": 1.975,
+      "step": 14270
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.6768869161605835,
+      "learning_rate": 2.4148121820516785e-05,
+      "loss": 2.0272,
+      "step": 14275
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.4785836338996887,
+      "learning_rate": 2.413407000355694e-05,
+      "loss": 1.7301,
+      "step": 14280
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7693573236465454,
+      "learning_rate": 2.412001846048991e-05,
+      "loss": 1.7174,
+      "step": 14285
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.5978009700775146,
+      "learning_rate": 2.4105967195760192e-05,
+      "loss": 1.8307,
+      "step": 14290
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.6870273947715759,
+      "learning_rate": 2.40919162138122e-05,
+      "loss": 1.6285,
+      "step": 14295
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.6152418255805969,
+      "learning_rate": 2.407786551909023e-05,
+      "loss": 2.058,
+      "step": 14300
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.458072543144226,
+      "learning_rate": 2.4063815116038512e-05,
+      "loss": 1.4863,
+      "step": 14305
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.782149612903595,
+      "learning_rate": 2.4049765009101165e-05,
+      "loss": 1.7991,
+      "step": 14310
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.7315062880516052,
+      "learning_rate": 2.4035715202722237e-05,
+      "loss": 1.6951,
+      "step": 14315
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 2.745988130569458,
+      "learning_rate": 2.4021665701345667e-05,
+      "loss": 1.8848,
+      "step": 14320
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.6406875848770142,
+      "learning_rate": 2.40076165094153e-05,
+      "loss": 2.0148,
+      "step": 14325
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.698235809803009,
+      "learning_rate": 2.3993567631374862e-05,
+      "loss": 2.1415,
+      "step": 14330
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.499514639377594,
+      "learning_rate": 2.397951907166802e-05,
+      "loss": 2.03,
+      "step": 14335
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.9097588062286377,
+      "learning_rate": 2.3965470834738312e-05,
+      "loss": 1.8086,
+      "step": 14340
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.8855141401290894,
+      "learning_rate": 2.3951422925029184e-05,
+      "loss": 1.465,
+      "step": 14345
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.5352545976638794,
+      "learning_rate": 2.393737534698398e-05,
+      "loss": 1.7741,
+      "step": 14350
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 2.3942019939422607,
+      "learning_rate": 2.3923328105045928e-05,
+      "loss": 2.0296,
+      "step": 14355
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 1.186553716659546,
+      "learning_rate": 2.390928120365816e-05,
+      "loss": 2.0242,
+      "step": 14360
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.4111693203449249,
+      "learning_rate": 2.3895234647263704e-05,
+      "loss": 1.9497,
+      "step": 14365
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.9013734459877014,
+      "learning_rate": 2.3881188440305468e-05,
+      "loss": 1.9767,
+      "step": 14370
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.960102379322052,
+      "learning_rate": 2.3867142587226247e-05,
+      "loss": 2.0041,
+      "step": 14375
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.4831627905368805,
+      "learning_rate": 2.3853097092468732e-05,
+      "loss": 1.5793,
+      "step": 14380
+    },
+    {
+      "epoch": 2.57,
+      "grad_norm": 0.7343013882637024,
+      "learning_rate": 2.383905196047551e-05,
+      "loss": 1.647,
+      "step": 14385
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.5119633674621582,
+      "learning_rate": 2.382500719568903e-05,
+      "loss": 1.7457,
+      "step": 14390
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.567718505859375,
+      "learning_rate": 2.381096280255165e-05,
+      "loss": 1.9885,
+      "step": 14395
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.3946149349212646,
+      "learning_rate": 2.3796918785505587e-05,
+      "loss": 1.9862,
+      "step": 14400
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.8330457210540771,
+      "learning_rate": 2.3782875148992954e-05,
+      "loss": 1.8574,
+      "step": 14405
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.6044062376022339,
+      "learning_rate": 2.3768831897455745e-05,
+      "loss": 1.796,
+      "step": 14410
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.53191077709198,
+      "learning_rate": 2.375478903533582e-05,
+      "loss": 1.9155,
+      "step": 14415
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.6012637615203857,
+      "learning_rate": 2.3740746567074935e-05,
+      "loss": 2.1244,
+      "step": 14420
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.6626573204994202,
+      "learning_rate": 2.3726704497114687e-05,
+      "loss": 1.6726,
+      "step": 14425
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.1490142345428467,
+      "learning_rate": 2.371266282989659e-05,
+      "loss": 1.7387,
+      "step": 14430
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.9652857780456543,
+      "learning_rate": 2.3698621569861996e-05,
+      "loss": 1.6093,
+      "step": 14435
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.8638285398483276,
+      "learning_rate": 2.3684580721452153e-05,
+      "loss": 1.9487,
+      "step": 14440
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.6422768235206604,
+      "learning_rate": 2.3670540289108153e-05,
+      "loss": 1.8297,
+      "step": 14445
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.0574924945831299,
+      "learning_rate": 2.3656500277270983e-05,
+      "loss": 1.9849,
+      "step": 14450
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.574671745300293,
+      "learning_rate": 2.3642460690381495e-05,
+      "loss": 2.0423,
+      "step": 14455
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7926428914070129,
+      "learning_rate": 2.362842153288037e-05,
+      "loss": 1.5625,
+      "step": 14460
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.752262532711029,
+      "learning_rate": 2.361438280920821e-05,
+      "loss": 2.0275,
+      "step": 14465
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.7062698602676392,
+      "learning_rate": 2.3600344523805424e-05,
+      "loss": 2.1019,
+      "step": 14470
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.3775525689125061,
+      "learning_rate": 2.3586306681112323e-05,
+      "loss": 1.9513,
+      "step": 14475
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 1.0212444067001343,
+      "learning_rate": 2.3572269285569054e-05,
+      "loss": 1.9915,
+      "step": 14480
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.917746901512146,
+      "learning_rate": 2.3558232341615643e-05,
+      "loss": 1.8467,
+      "step": 14485
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.5298284888267517,
+      "learning_rate": 2.354419585369195e-05,
+      "loss": 1.8151,
+      "step": 14490
+    },
+    {
+      "epoch": 2.59,
+      "grad_norm": 0.6283917427062988,
+      "learning_rate": 2.3530159826237707e-05,
+      "loss": 2.2388,
+      "step": 14495
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.9578955769538879,
+      "learning_rate": 2.3516124263692502e-05,
+      "loss": 1.5422,
+      "step": 14500
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.6109092235565186,
+      "learning_rate": 2.3502089170495762e-05,
+      "loss": 1.9209,
+      "step": 14505
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.5404300093650818,
+      "learning_rate": 2.3488054551086785e-05,
+      "loss": 1.9089,
+      "step": 14510
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.5609441995620728,
+      "learning_rate": 2.347402040990469e-05,
+      "loss": 1.8755,
+      "step": 14515
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 4.9784345626831055,
+      "learning_rate": 2.3459986751388478e-05,
+      "loss": 1.9347,
+      "step": 14520
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 2.010941505432129,
+      "learning_rate": 2.3445953579976968e-05,
+      "loss": 1.7676,
+      "step": 14525
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.6856008172035217,
+      "learning_rate": 2.343192090010885e-05,
+      "loss": 1.9795,
+      "step": 14530
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.712049126625061,
+      "learning_rate": 2.3417888716222643e-05,
+      "loss": 1.8283,
+      "step": 14535
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.41500189900398254,
+      "learning_rate": 2.3403857032756706e-05,
+      "loss": 1.8836,
+      "step": 14540
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.3549801707267761,
+      "learning_rate": 2.3389825854149263e-05,
+      "loss": 1.8883,
+      "step": 14545
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.8253669142723083,
+      "learning_rate": 2.337579518483835e-05,
+      "loss": 1.522,
+      "step": 14550
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.8156489729881287,
+      "learning_rate": 2.3361765029261863e-05,
+      "loss": 1.7102,
+      "step": 14555
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.8357172608375549,
+      "learning_rate": 2.334773539185752e-05,
+      "loss": 1.7754,
+      "step": 14560
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7100852727890015,
+      "learning_rate": 2.333370627706288e-05,
+      "loss": 1.4992,
+      "step": 14565
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7589089870452881,
+      "learning_rate": 2.3319677689315338e-05,
+      "loss": 2.224,
+      "step": 14570
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.49599790573120117,
+      "learning_rate": 2.3305649633052133e-05,
+      "loss": 1.5944,
+      "step": 14575
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.9817653894424438,
+      "learning_rate": 2.3291622112710316e-05,
+      "loss": 1.6229,
+      "step": 14580
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.6684935688972473,
+      "learning_rate": 2.327759513272678e-05,
+      "loss": 1.863,
+      "step": 14585
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7037563323974609,
+      "learning_rate": 2.326356869753825e-05,
+      "loss": 2.113,
+      "step": 14590
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.6015907526016235,
+      "learning_rate": 2.324954281158127e-05,
+      "loss": 1.7673,
+      "step": 14595
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 0.7683700323104858,
+      "learning_rate": 2.323551747929222e-05,
+      "loss": 1.9316,
+      "step": 14600
+    },
+    {
+      "epoch": 2.61,
+      "grad_norm": 3.4007463455200195,
+      "learning_rate": 2.3221492705107294e-05,
+      "loss": 1.6905,
+      "step": 14605
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.5388293266296387,
+      "learning_rate": 2.320746849346251e-05,
+      "loss": 2.086,
+      "step": 14610
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.5219169855117798,
+      "learning_rate": 2.3193444848793713e-05,
+      "loss": 1.9448,
+      "step": 14615
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.7064671516418457,
+      "learning_rate": 2.317942177553658e-05,
+      "loss": 1.5938,
+      "step": 14620
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.014631986618042,
+      "learning_rate": 2.3165399278126583e-05,
+      "loss": 1.9672,
+      "step": 14625
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.6399771571159363,
+      "learning_rate": 2.3151377360999025e-05,
+      "loss": 1.986,
+      "step": 14630
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.9371715784072876,
+      "learning_rate": 2.3137356028589042e-05,
+      "loss": 1.8568,
+      "step": 14635
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.1944173574447632,
+      "learning_rate": 2.3123335285331547e-05,
+      "loss": 2.001,
+      "step": 14640
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.6138972640037537,
+      "learning_rate": 2.3109315135661302e-05,
+      "loss": 1.8779,
+      "step": 14645
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 2.2491681575775146,
+      "learning_rate": 2.3095295584012865e-05,
+      "loss": 1.6365,
+      "step": 14650
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.8489646911621094,
+      "learning_rate": 2.3081276634820595e-05,
+      "loss": 1.7617,
+      "step": 14655
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.6606999635696411,
+      "learning_rate": 2.3067258292518675e-05,
+      "loss": 1.9637,
+      "step": 14660
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 0.7351112961769104,
+      "learning_rate": 2.3053240561541107e-05,
+      "loss": 1.6616,
+      "step": 14665
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.5655643343925476,
+      "learning_rate": 2.303922344632167e-05,
+      "loss": 1.6532,
+      "step": 14670
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.6417161226272583,
+      "learning_rate": 2.3025206951293973e-05,
+      "loss": 1.8476,
+      "step": 14675
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.9483529925346375,
+      "learning_rate": 2.301119108089142e-05,
+      "loss": 1.5566,
+      "step": 14680
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7348762154579163,
+      "learning_rate": 2.2997175839547215e-05,
+      "loss": 1.8403,
+      "step": 14685
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.085129976272583,
+      "learning_rate": 2.2983161231694364e-05,
+      "loss": 1.5614,
+      "step": 14690
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.830359697341919,
+      "learning_rate": 2.2969147261765675e-05,
+      "loss": 1.5355,
+      "step": 14695
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.0169621706008911,
+      "learning_rate": 2.2955133934193762e-05,
+      "loss": 2.0021,
+      "step": 14700
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.5929272174835205,
+      "learning_rate": 2.2941121253411008e-05,
+      "loss": 1.65,
+      "step": 14705
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 1.135832667350769,
+      "learning_rate": 2.2927109223849626e-05,
+      "loss": 1.8439,
+      "step": 14710
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.7020158767700195,
+      "learning_rate": 2.2913097849941602e-05,
+      "loss": 1.857,
+      "step": 14715
+    },
+    {
+      "epoch": 2.63,
+      "grad_norm": 0.45904645323753357,
+      "learning_rate": 2.2899087136118717e-05,
+      "loss": 1.6038,
+      "step": 14720
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.3881944417953491,
+      "learning_rate": 2.2885077086812555e-05,
+      "loss": 2.0703,
+      "step": 14725
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.6580604910850525,
+      "learning_rate": 2.2871067706454472e-05,
+      "loss": 1.718,
+      "step": 14730
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.6806758642196655,
+      "learning_rate": 2.285705899947563e-05,
+      "loss": 1.5215,
+      "step": 14735
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.8426775932312012,
+      "learning_rate": 2.284305097030696e-05,
+      "loss": 1.8828,
+      "step": 14740
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.8580935597419739,
+      "learning_rate": 2.2829043623379202e-05,
+      "loss": 1.9616,
+      "step": 14745
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.5852141380310059,
+      "learning_rate": 2.281503696312285e-05,
+      "loss": 1.8733,
+      "step": 14750
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.9512518644332886,
+      "learning_rate": 2.2801030993968213e-05,
+      "loss": 2.0427,
+      "step": 14755
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.8054193258285522,
+      "learning_rate": 2.278702572034535e-05,
+      "loss": 1.6561,
+      "step": 14760
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.729395866394043,
+      "learning_rate": 2.277302114668412e-05,
+      "loss": 1.7398,
+      "step": 14765
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.750243067741394,
+      "learning_rate": 2.2759017277414166e-05,
+      "loss": 1.7189,
+      "step": 14770
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.8767192959785461,
+      "learning_rate": 2.2745014116964884e-05,
+      "loss": 1.7503,
+      "step": 14775
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.0850597620010376,
+      "learning_rate": 2.273101166976548e-05,
+      "loss": 1.8113,
+      "step": 14780
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.1997833251953125,
+      "learning_rate": 2.2717009940244892e-05,
+      "loss": 1.9981,
+      "step": 14785
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.7039926648139954,
+      "learning_rate": 2.2703008932831874e-05,
+      "loss": 1.8546,
+      "step": 14790
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.5928205847740173,
+      "learning_rate": 2.268900865195491e-05,
+      "loss": 1.786,
+      "step": 14795
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 2.150042772293091,
+      "learning_rate": 2.2675009102042297e-05,
+      "loss": 2.0728,
+      "step": 14800
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.6576092839241028,
+      "learning_rate": 2.2661010287522057e-05,
+      "loss": 1.9185,
+      "step": 14805
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.43387851119041443,
+      "learning_rate": 2.2647012212822023e-05,
+      "loss": 2.0831,
+      "step": 14810
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 2.2747631072998047,
+      "learning_rate": 2.263301488236976e-05,
+      "loss": 1.9771,
+      "step": 14815
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.780261218547821,
+      "learning_rate": 2.2619018300592603e-05,
+      "loss": 1.7917,
+      "step": 14820
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 0.6798030734062195,
+      "learning_rate": 2.260502247191768e-05,
+      "loss": 1.5835,
+      "step": 14825
+    },
+    {
+      "epoch": 2.65,
+      "grad_norm": 1.7987754344940186,
+      "learning_rate": 2.2591027400771838e-05,
+      "loss": 1.5079,
+      "step": 14830
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.6671625971794128,
+      "learning_rate": 2.2577033091581715e-05,
+      "loss": 1.9078,
+      "step": 14835
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.9358892440795898,
+      "learning_rate": 2.25630395487737e-05,
+      "loss": 1.7289,
+      "step": 14840
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.6218108534812927,
+      "learning_rate": 2.2549046776773932e-05,
+      "loss": 1.6662,
+      "step": 14845
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.6321266889572144,
+      "learning_rate": 2.2535054780008302e-05,
+      "loss": 1.9075,
+      "step": 14850
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.0128587484359741,
+      "learning_rate": 2.252106356290248e-05,
+      "loss": 1.9468,
+      "step": 14855
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7265017628669739,
+      "learning_rate": 2.2507073129881863e-05,
+      "loss": 1.8416,
+      "step": 14860
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.9810836911201477,
+      "learning_rate": 2.249308348537162e-05,
+      "loss": 1.5815,
+      "step": 14865
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7224981784820557,
+      "learning_rate": 2.247909463379666e-05,
+      "loss": 1.9182,
+      "step": 14870
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.7442790269851685,
+      "learning_rate": 2.246510657958164e-05,
+      "loss": 1.9225,
+      "step": 14875
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 0.815650224685669,
+      "learning_rate": 2.245111932715097e-05,
+      "loss": 1.8601,
+      "step": 14880
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.262229323387146,
+      "learning_rate": 2.2437132880928817e-05,
+      "loss": 2.0791,
+      "step": 14885
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.8286881446838379,
+      "learning_rate": 2.2423147245339062e-05,
+      "loss": 1.9286,
+      "step": 14890
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.7588956356048584,
+      "learning_rate": 2.240916242480535e-05,
+      "loss": 1.697,
+      "step": 14895
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.852202296257019,
+      "learning_rate": 2.2395178423751076e-05,
+      "loss": 1.7105,
+      "step": 14900
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.785529375076294,
+      "learning_rate": 2.2381195246599356e-05,
+      "loss": 1.7383,
+      "step": 14905
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.9461120963096619,
+      "learning_rate": 2.236721289777306e-05,
+      "loss": 1.8059,
+      "step": 14910
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.5357199311256409,
+      "learning_rate": 2.2353231381694797e-05,
+      "loss": 2.066,
+      "step": 14915
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 2.159280300140381,
+      "learning_rate": 2.23392507027869e-05,
+      "loss": 1.8313,
+      "step": 14920
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.9953837394714355,
+      "learning_rate": 2.2325270865471444e-05,
+      "loss": 1.8698,
+      "step": 14925
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.6990963220596313,
+      "learning_rate": 2.231129187417025e-05,
+      "loss": 1.7352,
+      "step": 14930
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.6935223937034607,
+      "learning_rate": 2.2297313733304845e-05,
+      "loss": 1.514,
+      "step": 14935
+    },
+    {
+      "epoch": 2.67,
+      "grad_norm": 0.5469807386398315,
+      "learning_rate": 2.22833364472965e-05,
+      "loss": 1.8815,
+      "step": 14940
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.5818589925765991,
+      "learning_rate": 2.2269360020566232e-05,
+      "loss": 1.9459,
+      "step": 14945
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.0151652097702026,
+      "learning_rate": 2.2255384457534757e-05,
+      "loss": 1.8557,
+      "step": 14950
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.0191221237182617,
+      "learning_rate": 2.2241409762622534e-05,
+      "loss": 1.5992,
+      "step": 14955
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.0580757856369019,
+      "learning_rate": 2.222743594024976e-05,
+      "loss": 1.8454,
+      "step": 14960
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.6717436909675598,
+      "learning_rate": 2.2213462994836323e-05,
+      "loss": 2.0803,
+      "step": 14965
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7515537738800049,
+      "learning_rate": 2.2199490930801862e-05,
+      "loss": 1.6758,
+      "step": 14970
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.05194890499115,
+      "learning_rate": 2.2185519752565733e-05,
+      "loss": 1.7331,
+      "step": 14975
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.2085919380187988,
+      "learning_rate": 2.2171549464547e-05,
+      "loss": 1.8599,
+      "step": 14980
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7703227400779724,
+      "learning_rate": 2.215758007116444e-05,
+      "loss": 1.8927,
+      "step": 14985
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7714345455169678,
+      "learning_rate": 2.214361157683658e-05,
+      "loss": 1.4521,
+      "step": 14990
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.7530484795570374,
+      "learning_rate": 2.212964398598162e-05,
+      "loss": 1.9209,
+      "step": 14995
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.6522998213768005,
+      "learning_rate": 2.211567730301751e-05,
+      "loss": 1.9337,
+      "step": 15000
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.6403537392616272,
+      "learning_rate": 2.21017115323619e-05,
+      "loss": 1.8753,
+      "step": 15005
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.9187439680099487,
+      "learning_rate": 2.2087746678432135e-05,
+      "loss": 1.7973,
+      "step": 15010
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.840508222579956,
+      "learning_rate": 2.20737827456453e-05,
+      "loss": 2.1179,
+      "step": 15015
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.1166287660598755,
+      "learning_rate": 2.205981973841817e-05,
+      "loss": 1.5053,
+      "step": 15020
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.1791390180587769,
+      "learning_rate": 2.2045857661167235e-05,
+      "loss": 1.7282,
+      "step": 15025
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 1.0932645797729492,
+      "learning_rate": 2.203189651830867e-05,
+      "loss": 1.9909,
+      "step": 15030
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 2.5372047424316406,
+      "learning_rate": 2.2017936314258385e-05,
+      "loss": 2.1166,
+      "step": 15035
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.7102367877960205,
+      "learning_rate": 2.2003977053431972e-05,
+      "loss": 2.0205,
+      "step": 15040
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 3.9497880935668945,
+      "learning_rate": 2.1990018740244733e-05,
+      "loss": 1.8872,
+      "step": 15045
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.714215874671936,
+      "learning_rate": 2.1976061379111677e-05,
+      "loss": 1.9495,
+      "step": 15050
+    },
+    {
+      "epoch": 2.69,
+      "grad_norm": 0.8167940974235535,
+      "learning_rate": 2.196210497444749e-05,
+      "loss": 1.8913,
+      "step": 15055
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.2327383756637573,
+      "learning_rate": 2.194814953066659e-05,
+      "loss": 1.9469,
+      "step": 15060
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.683508574962616,
+      "learning_rate": 2.1934195052183047e-05,
+      "loss": 1.9397,
+      "step": 15065
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.7274824380874634,
+      "learning_rate": 2.1920241543410673e-05,
+      "loss": 1.9172,
+      "step": 15070
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.8672675490379333,
+      "learning_rate": 2.1906289008762926e-05,
+      "loss": 1.7328,
+      "step": 15075
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.210166335105896,
+      "learning_rate": 2.1892337452652996e-05,
+      "loss": 1.6836,
+      "step": 15080
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.42006659507751465,
+      "learning_rate": 2.1878386879493732e-05,
+      "loss": 1.7958,
+      "step": 15085
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.6762058734893799,
+      "learning_rate": 2.18644372936977e-05,
+      "loss": 1.9698,
+      "step": 15090
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 2.0761775970458984,
+      "learning_rate": 2.1850488699677137e-05,
+      "loss": 1.711,
+      "step": 15095
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.4220300614833832,
+      "learning_rate": 2.1836541101843964e-05,
+      "loss": 1.9976,
+      "step": 15100
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.0192917585372925,
+      "learning_rate": 2.1822594504609802e-05,
+      "loss": 1.6178,
+      "step": 15105
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.5573521256446838,
+      "learning_rate": 2.180864891238594e-05,
+      "loss": 1.8285,
+      "step": 15110
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7691249847412109,
+      "learning_rate": 2.1794704329583353e-05,
+      "loss": 1.8416,
+      "step": 15115
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.6303778290748596,
+      "learning_rate": 2.178076076061272e-05,
+      "loss": 1.9445,
+      "step": 15120
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.794248342514038,
+      "learning_rate": 2.1766818209884355e-05,
+      "loss": 1.772,
+      "step": 15125
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.5555631518363953,
+      "learning_rate": 2.1752876681808272e-05,
+      "loss": 1.9228,
+      "step": 15130
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.8108105063438416,
+      "learning_rate": 2.1738936180794173e-05,
+      "loss": 1.7449,
+      "step": 15135
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.45947834849357605,
+      "learning_rate": 2.1724996711251426e-05,
+      "loss": 2.1185,
+      "step": 15140
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.4830736815929413,
+      "learning_rate": 2.1711058277589066e-05,
+      "loss": 1.8752,
+      "step": 15145
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.6274514198303223,
+      "learning_rate": 2.1697120884215817e-05,
+      "loss": 2.2124,
+      "step": 15150
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.7346020340919495,
+      "learning_rate": 2.1683184535540046e-05,
+      "loss": 1.7558,
+      "step": 15155
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 1.0452593564987183,
+      "learning_rate": 2.166924923596982e-05,
+      "loss": 1.692,
+      "step": 15160
+    },
+    {
+      "epoch": 2.71,
+      "grad_norm": 0.8475756645202637,
+      "learning_rate": 2.165531498991286e-05,
+      "loss": 1.6763,
+      "step": 15165
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.5865801572799683,
+      "learning_rate": 2.1641381801776557e-05,
+      "loss": 1.998,
+      "step": 15170
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.8659747242927551,
+      "learning_rate": 2.1627449675967945e-05,
+      "loss": 1.9164,
+      "step": 15175
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.834214985370636,
+      "learning_rate": 2.161351861689376e-05,
+      "loss": 1.8393,
+      "step": 15180
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.6650885939598083,
+      "learning_rate": 2.1599588628960384e-05,
+      "loss": 1.8873,
+      "step": 15185
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.6627525687217712,
+      "learning_rate": 2.158565971657385e-05,
+      "loss": 1.8317,
+      "step": 15190
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 2.031599998474121,
+      "learning_rate": 2.157173188413987e-05,
+      "loss": 2.1292,
+      "step": 15195
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.6217500567436218,
+      "learning_rate": 2.1557805136063787e-05,
+      "loss": 1.9539,
+      "step": 15200
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.5643689036369324,
+      "learning_rate": 2.1543879476750632e-05,
+      "loss": 2.012,
+      "step": 15205
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.619275689125061,
+      "learning_rate": 2.1529954910605084e-05,
+      "loss": 1.8118,
+      "step": 15210
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.6966983675956726,
+      "learning_rate": 2.1516031442031452e-05,
+      "loss": 1.9702,
+      "step": 15215
+    },
+    {
+      "epoch": 2.72,
+      "grad_norm": 0.824952244758606,
+      "learning_rate": 2.1502109075433725e-05,
+      "loss": 1.8539,
+      "step": 15220
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.7268326878547668,
+      "learning_rate": 2.1488187815215527e-05,
+      "loss": 1.8079,
+      "step": 15225
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.47795695066452026,
+      "learning_rate": 2.1474267665780153e-05,
+      "loss": 1.8452,
+      "step": 15230
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.0499993562698364,
+      "learning_rate": 2.146034863153052e-05,
+      "loss": 1.9181,
+      "step": 15235
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.5682938694953918,
+      "learning_rate": 2.1446430716869215e-05,
+      "loss": 1.7971,
+      "step": 15240
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.4888734519481659,
+      "learning_rate": 2.143251392619845e-05,
+      "loss": 1.979,
+      "step": 15245
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.2355575561523438,
+      "learning_rate": 2.14185982639201e-05,
+      "loss": 1.6015,
+      "step": 15250
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 2.380397319793701,
+      "learning_rate": 2.140468373443568e-05,
+      "loss": 1.8096,
+      "step": 15255
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 5.283520698547363,
+      "learning_rate": 2.1390770342146338e-05,
+      "loss": 1.8967,
+      "step": 15260
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 1.0056599378585815,
+      "learning_rate": 2.137685809145286e-05,
+      "loss": 1.6557,
+      "step": 15265
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.4523605704307556,
+      "learning_rate": 2.1362946986755678e-05,
+      "loss": 1.7145,
+      "step": 15270
+    },
+    {
+      "epoch": 2.73,
+      "grad_norm": 0.9183207154273987,
+      "learning_rate": 2.134903703245487e-05,
+      "loss": 2.0983,
+      "step": 15275
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.6543048620223999,
+      "learning_rate": 2.1335128232950135e-05,
+      "loss": 2.0348,
+      "step": 15280
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.8366292119026184,
+      "learning_rate": 2.132122059264082e-05,
+      "loss": 1.6401,
+      "step": 15285
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.6985805034637451,
+      "learning_rate": 2.1307314115925888e-05,
+      "loss": 1.749,
+      "step": 15290
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.4857269823551178,
+      "learning_rate": 2.1293408807203947e-05,
+      "loss": 1.916,
+      "step": 15295
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.9709117412567139,
+      "learning_rate": 2.1279504670873245e-05,
+      "loss": 2.0241,
+      "step": 15300
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.052292823791504,
+      "learning_rate": 2.1265601711331636e-05,
+      "loss": 1.5705,
+      "step": 15305
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7669774889945984,
+      "learning_rate": 2.1251699932976606e-05,
+      "loss": 1.7281,
+      "step": 15310
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.21905517578125,
+      "learning_rate": 2.1237799340205283e-05,
+      "loss": 2.013,
+      "step": 15315
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.9224725365638733,
+      "learning_rate": 2.1223899937414415e-05,
+      "loss": 1.8188,
+      "step": 15320
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.9945480823516846,
+      "learning_rate": 2.1210001729000357e-05,
+      "loss": 1.7489,
+      "step": 15325
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.0888783931732178,
+      "learning_rate": 2.1196104719359115e-05,
+      "loss": 1.8444,
+      "step": 15330
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 0.7765944004058838,
+      "learning_rate": 2.1182208912886283e-05,
+      "loss": 1.9981,
+      "step": 15335
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.3819294571876526,
+      "learning_rate": 2.11683143139771e-05,
+      "loss": 1.8742,
+      "step": 15340
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.1613242626190186,
+      "learning_rate": 2.1154420927026424e-05,
+      "loss": 1.526,
+      "step": 15345
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.8339876532554626,
+      "learning_rate": 2.114052875642871e-05,
+      "loss": 2.2053,
+      "step": 15350
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.7060007452964783,
+      "learning_rate": 2.1126637806578027e-05,
+      "loss": 1.9372,
+      "step": 15355
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.6718441247940063,
+      "learning_rate": 2.111274808186809e-05,
+      "loss": 1.7533,
+      "step": 15360
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.6338340044021606,
+      "learning_rate": 2.1098859586692184e-05,
+      "loss": 1.8297,
+      "step": 15365
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.91750568151474,
+      "learning_rate": 2.108497232544324e-05,
+      "loss": 1.8179,
+      "step": 15370
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.8784101009368896,
+      "learning_rate": 2.1071086302513793e-05,
+      "loss": 1.7564,
+      "step": 15375
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.9052116870880127,
+      "learning_rate": 2.105720152229596e-05,
+      "loss": 1.8512,
+      "step": 15380
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 1.1224867105484009,
+      "learning_rate": 2.1043317989181497e-05,
+      "loss": 1.7728,
+      "step": 15385
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.5035684704780579,
+      "learning_rate": 2.102943570756175e-05,
+      "loss": 1.9403,
+      "step": 15390
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.207096815109253,
+      "learning_rate": 2.1015554681827665e-05,
+      "loss": 1.7796,
+      "step": 15395
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.99371737241745,
+      "learning_rate": 2.1001674916369812e-05,
+      "loss": 2.0566,
+      "step": 15400
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.190355658531189,
+      "learning_rate": 2.0987796415578333e-05,
+      "loss": 1.8686,
+      "step": 15405
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.4896886944770813,
+      "learning_rate": 2.0973919183842983e-05,
+      "loss": 1.9051,
+      "step": 15410
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.8628058433532715,
+      "learning_rate": 2.096004322555312e-05,
+      "loss": 1.5376,
+      "step": 15415
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 2.157165765762329,
+      "learning_rate": 2.0946168545097708e-05,
+      "loss": 1.5625,
+      "step": 15420
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.9407357573509216,
+      "learning_rate": 2.0932295146865277e-05,
+      "loss": 1.674,
+      "step": 15425
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.7408738136291504,
+      "learning_rate": 2.0918423035243983e-05,
+      "loss": 1.8977,
+      "step": 15430
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.6515112519264221,
+      "learning_rate": 2.090455221462156e-05,
+      "loss": 1.8421,
+      "step": 15435
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.649272620677948,
+      "learning_rate": 2.0890682689385324e-05,
+      "loss": 1.9893,
+      "step": 15440
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.8793091177940369,
+      "learning_rate": 2.0876814463922213e-05,
+      "loss": 1.5602,
+      "step": 15445
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.6581525206565857,
+      "learning_rate": 2.086294754261872e-05,
+      "loss": 1.7371,
+      "step": 15450
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.5119839906692505,
+      "learning_rate": 2.084908192986093e-05,
+      "loss": 2.0622,
+      "step": 15455
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.5535023808479309,
+      "learning_rate": 2.083521763003453e-05,
+      "loss": 2.2007,
+      "step": 15460
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.1636162996292114,
+      "learning_rate": 2.08213546475248e-05,
+      "loss": 1.9082,
+      "step": 15465
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.48526015877723694,
+      "learning_rate": 2.0807492986716566e-05,
+      "loss": 2.0381,
+      "step": 15470
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.322176456451416,
+      "learning_rate": 2.0793632651994268e-05,
+      "loss": 1.7674,
+      "step": 15475
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.6021888256072998,
+      "learning_rate": 2.0779773647741924e-05,
+      "loss": 1.6768,
+      "step": 15480
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.8677191138267517,
+      "learning_rate": 2.076591597834311e-05,
+      "loss": 1.6359,
+      "step": 15485
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.6849371790885925,
+      "learning_rate": 2.0752059648181008e-05,
+      "loss": 1.7706,
+      "step": 15490
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 1.1195003986358643,
+      "learning_rate": 2.073820466163835e-05,
+      "loss": 1.7344,
+      "step": 15495
+    },
+    {
+      "epoch": 2.77,
+      "grad_norm": 0.6078161001205444,
+      "learning_rate": 2.072435102309745e-05,
+      "loss": 1.687,
+      "step": 15500
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.8624263405799866,
+      "learning_rate": 2.0710498736940208e-05,
+      "loss": 1.8988,
+      "step": 15505
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.518776535987854,
+      "learning_rate": 2.06966478075481e-05,
+      "loss": 1.9482,
+      "step": 15510
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.0695035457611084,
+      "learning_rate": 2.068279823930214e-05,
+      "loss": 1.7764,
+      "step": 15515
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.537037193775177,
+      "learning_rate": 2.066895003658295e-05,
+      "loss": 1.8001,
+      "step": 15520
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7043410539627075,
+      "learning_rate": 2.0655103203770702e-05,
+      "loss": 1.8046,
+      "step": 15525
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.6016290783882141,
+      "learning_rate": 2.0641257745245124e-05,
+      "loss": 1.8151,
+      "step": 15530
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 1.0047852993011475,
+      "learning_rate": 2.0627413665385533e-05,
+      "loss": 1.9005,
+      "step": 15535
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.5631789565086365,
+      "learning_rate": 2.06135709685708e-05,
+      "loss": 1.7507,
+      "step": 15540
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.7759091258049011,
+      "learning_rate": 2.059972965917934e-05,
+      "loss": 1.7911,
+      "step": 15545
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.9690524339675903,
+      "learning_rate": 2.0585889741589155e-05,
+      "loss": 1.6226,
+      "step": 15550
+    },
+    {
+      "epoch": 2.78,
+      "grad_norm": 0.9261754751205444,
+      "learning_rate": 2.0572051220177807e-05,
+      "loss": 1.722,
+      "step": 15555
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7827115058898926,
+      "learning_rate": 2.0558214099322393e-05,
+      "loss": 1.8318,
+      "step": 15560
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7148807644844055,
+      "learning_rate": 2.0544378383399593e-05,
+      "loss": 1.6204,
+      "step": 15565
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.6687959432601929,
+      "learning_rate": 2.0530544076785617e-05,
+      "loss": 1.9251,
+      "step": 15570
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.5032340288162231,
+      "learning_rate": 2.051671118385625e-05,
+      "loss": 1.9225,
+      "step": 15575
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.197785496711731,
+      "learning_rate": 2.050287970898683e-05,
+      "loss": 1.9162,
+      "step": 15580
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.1399624347686768,
+      "learning_rate": 2.0489049656552236e-05,
+      "loss": 1.8481,
+      "step": 15585
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 1.0693291425704956,
+      "learning_rate": 2.0475221030926888e-05,
+      "loss": 1.919,
+      "step": 15590
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.7608375549316406,
+      "learning_rate": 2.0461393836484776e-05,
+      "loss": 1.8845,
+      "step": 15595
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.8283123970031738,
+      "learning_rate": 2.0447568077599432e-05,
+      "loss": 1.6992,
+      "step": 15600
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.8869904279708862,
+      "learning_rate": 2.043374375864392e-05,
+      "loss": 1.91,
+      "step": 15605
+    },
+    {
+      "epoch": 2.79,
+      "grad_norm": 0.9428117871284485,
+      "learning_rate": 2.0419920883990874e-05,
+      "loss": 2.0304,
+      "step": 15610
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.6648744344711304,
+      "learning_rate": 2.0406099458012438e-05,
+      "loss": 1.5347,
+      "step": 15615
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.004983901977539,
+      "learning_rate": 2.0392279485080323e-05,
+      "loss": 1.6797,
+      "step": 15620
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7050873637199402,
+      "learning_rate": 2.0378460969565782e-05,
+      "loss": 1.8073,
+      "step": 15625
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7096034288406372,
+      "learning_rate": 2.0364643915839587e-05,
+      "loss": 1.694,
+      "step": 15630
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.6012222170829773,
+      "learning_rate": 2.0350828328272076e-05,
+      "loss": 1.8169,
+      "step": 15635
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.9153398275375366,
+      "learning_rate": 2.0337014211233077e-05,
+      "loss": 1.96,
+      "step": 15640
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.016750454902649,
+      "learning_rate": 2.0323201569092003e-05,
+      "loss": 1.9078,
+      "step": 15645
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.7948215007781982,
+      "learning_rate": 2.0309390406217772e-05,
+      "loss": 1.9035,
+      "step": 15650
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.7154805660247803,
+      "learning_rate": 2.029558072697885e-05,
+      "loss": 1.9442,
+      "step": 15655
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.8079678416252136,
+      "learning_rate": 2.028177253574321e-05,
+      "loss": 1.6577,
+      "step": 15660
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.8642460107803345,
+      "learning_rate": 2.026796583687838e-05,
+      "loss": 1.6969,
+      "step": 15665
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.655441164970398,
+      "learning_rate": 2.025416063475141e-05,
+      "loss": 1.854,
+      "step": 15670
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.6256705522537231,
+      "learning_rate": 2.024035693372886e-05,
+      "loss": 1.9581,
+      "step": 15675
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7426342964172363,
+      "learning_rate": 2.0226554738176845e-05,
+      "loss": 1.8966,
+      "step": 15680
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7382630109786987,
+      "learning_rate": 2.0212754052460963e-05,
+      "loss": 1.9278,
+      "step": 15685
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.1039477586746216,
+      "learning_rate": 2.0198954880946376e-05,
+      "loss": 1.8093,
+      "step": 15690
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.5927040576934814,
+      "learning_rate": 2.018515722799774e-05,
+      "loss": 1.8377,
+      "step": 15695
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.4983259439468384,
+      "learning_rate": 2.017136109797924e-05,
+      "loss": 1.9169,
+      "step": 15700
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.9285030364990234,
+      "learning_rate": 2.0157566495254583e-05,
+      "loss": 1.9493,
+      "step": 15705
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.9399527311325073,
+      "learning_rate": 2.014377342418698e-05,
+      "loss": 1.8094,
+      "step": 15710
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 0.7926543951034546,
+      "learning_rate": 2.0129981889139175e-05,
+      "loss": 1.8163,
+      "step": 15715
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.1448842287063599,
+      "learning_rate": 2.0116191894473414e-05,
+      "loss": 1.8834,
+      "step": 15720
+    },
+    {
+      "epoch": 2.81,
+      "grad_norm": 1.2125980854034424,
+      "learning_rate": 2.0102403444551467e-05,
+      "loss": 1.9686,
+      "step": 15725
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.6270211935043335,
+      "learning_rate": 2.008861654373459e-05,
+      "loss": 1.6633,
+      "step": 15730
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.845684826374054,
+      "learning_rate": 2.007483119638358e-05,
+      "loss": 1.5904,
+      "step": 15735
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.9219332337379456,
+      "learning_rate": 2.0061047406858725e-05,
+      "loss": 1.7978,
+      "step": 15740
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.5408844351768494,
+      "learning_rate": 2.0047265179519825e-05,
+      "loss": 1.7812,
+      "step": 15745
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7204087972640991,
+      "learning_rate": 2.0033484518726182e-05,
+      "loss": 2.0534,
+      "step": 15750
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.9531112313270569,
+      "learning_rate": 2.0019705428836604e-05,
+      "loss": 1.6982,
+      "step": 15755
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.7499538064002991,
+      "learning_rate": 2.0005927914209417e-05,
+      "loss": 1.8288,
+      "step": 15760
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.6334372758865356,
+      "learning_rate": 1.9992151979202424e-05,
+      "loss": 1.6551,
+      "step": 15765
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.9548289775848389,
+      "learning_rate": 1.9978377628172938e-05,
+      "loss": 2.0638,
+      "step": 15770
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.6916335225105286,
+      "learning_rate": 1.996460486547779e-05,
+      "loss": 2.0167,
+      "step": 15775
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.789394199848175,
+      "learning_rate": 1.9950833695473277e-05,
+      "loss": 1.9085,
+      "step": 15780
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.42912745475769043,
+      "learning_rate": 1.9937064122515202e-05,
+      "loss": 1.8567,
+      "step": 15785
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.1623276472091675,
+      "learning_rate": 1.992329615095888e-05,
+      "loss": 1.8448,
+      "step": 15790
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.7881998419761658,
+      "learning_rate": 1.9909529785159097e-05,
+      "loss": 1.9888,
+      "step": 15795
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.5520839691162109,
+      "learning_rate": 1.9895765029470148e-05,
+      "loss": 2.0507,
+      "step": 15800
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 1.4102293252944946,
+      "learning_rate": 1.9882001888245816e-05,
+      "loss": 1.9575,
+      "step": 15805
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.5933654308319092,
+      "learning_rate": 1.9868240365839358e-05,
+      "loss": 2.1449,
+      "step": 15810
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.37906643748283386,
+      "learning_rate": 1.9854480466603542e-05,
+      "loss": 1.9846,
+      "step": 15815
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.675449550151825,
+      "learning_rate": 1.9840722194890604e-05,
+      "loss": 2.0511,
+      "step": 15820
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.8554467558860779,
+      "learning_rate": 1.9826965555052274e-05,
+      "loss": 1.7434,
+      "step": 15825
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.68226557970047,
+      "learning_rate": 1.9813210551439756e-05,
+      "loss": 1.9633,
+      "step": 15830
+    },
+    {
+      "epoch": 2.83,
+      "grad_norm": 0.6609245538711548,
+      "learning_rate": 1.9799457188403758e-05,
+      "loss": 2.0091,
+      "step": 15835
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.8956989049911499,
+      "learning_rate": 1.9785705470294447e-05,
+      "loss": 1.775,
+      "step": 15840
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.621740996837616,
+      "learning_rate": 1.9771955401461477e-05,
+      "loss": 1.8911,
+      "step": 15845
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.393864393234253,
+      "learning_rate": 1.975820698625399e-05,
+      "loss": 1.5529,
+      "step": 15850
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.920356273651123,
+      "learning_rate": 1.9744460229020583e-05,
+      "loss": 1.9129,
+      "step": 15855
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.2865760326385498,
+      "learning_rate": 1.973071513410936e-05,
+      "loss": 2.1003,
+      "step": 15860
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.5856181383132935,
+      "learning_rate": 1.971697170586787e-05,
+      "loss": 1.755,
+      "step": 15865
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.4005410969257355,
+      "learning_rate": 1.970322994864315e-05,
+      "loss": 1.992,
+      "step": 15870
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.9305061101913452,
+      "learning_rate": 1.9689489866781695e-05,
+      "loss": 1.9721,
+      "step": 15875
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.9853824377059937,
+      "learning_rate": 1.9675751464629493e-05,
+      "loss": 2.1053,
+      "step": 15880
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.6760913133621216,
+      "learning_rate": 1.9662014746531977e-05,
+      "loss": 1.8466,
+      "step": 15885
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.451269268989563,
+      "learning_rate": 1.9648279716834058e-05,
+      "loss": 2.0028,
+      "step": 15890
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.890579879283905,
+      "learning_rate": 1.9634546379880127e-05,
+      "loss": 1.7298,
+      "step": 15895
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.1749356985092163,
+      "learning_rate": 1.962081474001401e-05,
+      "loss": 2.0342,
+      "step": 15900
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.923694908618927,
+      "learning_rate": 1.9607084801579018e-05,
+      "loss": 2.0843,
+      "step": 15905
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7525226473808289,
+      "learning_rate": 1.9593356568917913e-05,
+      "loss": 2.129,
+      "step": 15910
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.7968067526817322,
+      "learning_rate": 1.9579630046372936e-05,
+      "loss": 2.1804,
+      "step": 15915
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.6558468341827393,
+      "learning_rate": 1.956590523828575e-05,
+      "loss": 1.6117,
+      "step": 15920
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.8354988098144531,
+      "learning_rate": 1.9552182148997513e-05,
+      "loss": 1.7763,
+      "step": 15925
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.8877277374267578,
+      "learning_rate": 1.9538460782848822e-05,
+      "loss": 1.9353,
+      "step": 15930
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.990190327167511,
+      "learning_rate": 1.9524741144179727e-05,
+      "loss": 1.6819,
+      "step": 15935
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 1.0683060884475708,
+      "learning_rate": 1.951102323732975e-05,
+      "loss": 1.7892,
+      "step": 15940
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.5219915509223938,
+      "learning_rate": 1.9497307066637837e-05,
+      "loss": 2.1233,
+      "step": 15945
+    },
+    {
+      "epoch": 2.85,
+      "grad_norm": 0.791999101638794,
+      "learning_rate": 1.948359263644241e-05,
+      "loss": 1.8079,
+      "step": 15950
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.5373420715332031,
+      "learning_rate": 1.9469879951081323e-05,
+      "loss": 1.8237,
+      "step": 15955
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.6582045555114746,
+      "learning_rate": 1.94561690148919e-05,
+      "loss": 1.8346,
+      "step": 15960
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.0421522855758667,
+      "learning_rate": 1.944245983221087e-05,
+      "loss": 2.09,
+      "step": 15965
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.5510540008544922,
+      "learning_rate": 1.942875240737446e-05,
+      "loss": 2.0754,
+      "step": 15970
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7140408158302307,
+      "learning_rate": 1.9415046744718295e-05,
+      "loss": 2.1591,
+      "step": 15975
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.533048689365387,
+      "learning_rate": 1.9401342848577476e-05,
+      "loss": 2.0399,
+      "step": 15980
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.1084396839141846,
+      "learning_rate": 1.9387640723286533e-05,
+      "loss": 1.5784,
+      "step": 15985
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.6885467767715454,
+      "learning_rate": 1.9373940373179424e-05,
+      "loss": 1.6761,
+      "step": 15990
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.7324159145355225,
+      "learning_rate": 1.936024180258957e-05,
+      "loss": 1.9015,
+      "step": 15995
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.0003372430801392,
+      "learning_rate": 1.93465450158498e-05,
+      "loss": 1.5935,
+      "step": 16000
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 0.4618781805038452,
+      "learning_rate": 1.933285001729242e-05,
+      "loss": 2.0161,
+      "step": 16005
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.0251598358154297,
+      "learning_rate": 1.931915681124911e-05,
+      "loss": 1.7638,
+      "step": 16010
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.092031478881836,
+      "learning_rate": 1.9305465402051047e-05,
+      "loss": 1.7572,
+      "step": 16015
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.4748564064502716,
+      "learning_rate": 1.92917757940288e-05,
+      "loss": 1.8665,
+      "step": 16020
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.7850896120071411,
+      "learning_rate": 1.927808799151237e-05,
+      "loss": 1.5633,
+      "step": 16025
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7904292941093445,
+      "learning_rate": 1.9264401998831213e-05,
+      "loss": 1.7375,
+      "step": 16030
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.9015944600105286,
+      "learning_rate": 1.9250717820314182e-05,
+      "loss": 1.5826,
+      "step": 16035
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.7508746981620789,
+      "learning_rate": 1.923703546028958e-05,
+      "loss": 2.1586,
+      "step": 16040
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 4.131949424743652,
+      "learning_rate": 1.922335492308511e-05,
+      "loss": 1.7214,
+      "step": 16045
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 1.2245509624481201,
+      "learning_rate": 1.9209676213027923e-05,
+      "loss": 1.7981,
+      "step": 16050
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.4659061133861542,
+      "learning_rate": 1.919599933444459e-05,
+      "loss": 1.8612,
+      "step": 16055
+    },
+    {
+      "epoch": 2.87,
+      "grad_norm": 0.6791583895683289,
+      "learning_rate": 1.9182324291661084e-05,
+      "loss": 1.8788,
+      "step": 16060
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.2033220529556274,
+      "learning_rate": 1.91686510890028e-05,
+      "loss": 1.7272,
+      "step": 16065
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.5756292343139648,
+      "learning_rate": 1.915497973079457e-05,
+      "loss": 1.9688,
+      "step": 16070
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.4528789222240448,
+      "learning_rate": 1.9141310221360632e-05,
+      "loss": 1.9315,
+      "step": 16075
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.4836488664150238,
+      "learning_rate": 1.9127642565024628e-05,
+      "loss": 1.7283,
+      "step": 16080
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.785624623298645,
+      "learning_rate": 1.9113976766109643e-05,
+      "loss": 1.7548,
+      "step": 16085
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.9146924018859863,
+      "learning_rate": 1.9100312828938135e-05,
+      "loss": 1.8917,
+      "step": 16090
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.9764569401741028,
+      "learning_rate": 1.9086650757832007e-05,
+      "loss": 1.6859,
+      "step": 16095
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.8843137621879578,
+      "learning_rate": 1.9072990557112564e-05,
+      "loss": 1.8221,
+      "step": 16100
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.6954315900802612,
+      "learning_rate": 1.905933223110051e-05,
+      "loss": 1.8598,
+      "step": 16105
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.5043737888336182,
+      "learning_rate": 1.904567578411594e-05,
+      "loss": 2.0398,
+      "step": 16110
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.8501962423324585,
+      "learning_rate": 1.9032021220478405e-05,
+      "loss": 1.8814,
+      "step": 16115
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.6526318192481995,
+      "learning_rate": 1.901836854450681e-05,
+      "loss": 1.9833,
+      "step": 16120
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7955963611602783,
+      "learning_rate": 1.900471776051949e-05,
+      "loss": 1.8694,
+      "step": 16125
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.724168062210083,
+      "learning_rate": 1.8991068872834185e-05,
+      "loss": 2.09,
+      "step": 16130
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7435161471366882,
+      "learning_rate": 1.8977421885768004e-05,
+      "loss": 1.896,
+      "step": 16135
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.8156474828720093,
+      "learning_rate": 1.896377680363749e-05,
+      "loss": 2.0581,
+      "step": 16140
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7331948280334473,
+      "learning_rate": 1.8950133630758582e-05,
+      "loss": 1.7547,
+      "step": 16145
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.7191774845123291,
+      "learning_rate": 1.893649237144658e-05,
+      "loss": 1.6202,
+      "step": 16150
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 1.1362369060516357,
+      "learning_rate": 1.89228530300162e-05,
+      "loss": 1.8106,
+      "step": 16155
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.5437344908714294,
+      "learning_rate": 1.8909215610781566e-05,
+      "loss": 1.9975,
+      "step": 16160
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.9459233283996582,
+      "learning_rate": 1.8895580118056174e-05,
+      "loss": 1.7396,
+      "step": 16165
+    },
+    {
+      "epoch": 2.89,
+      "grad_norm": 0.5802138447761536,
+      "learning_rate": 1.8881946556152918e-05,
+      "loss": 1.8485,
+      "step": 16170
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.6844445466995239,
+      "learning_rate": 1.8868314929384084e-05,
+      "loss": 1.5007,
+      "step": 16175
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7569360136985779,
+      "learning_rate": 1.885468524206134e-05,
+      "loss": 1.5597,
+      "step": 16180
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7909181714057922,
+      "learning_rate": 1.8841057498495736e-05,
+      "loss": 1.8607,
+      "step": 16185
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.9658223986625671,
+      "learning_rate": 1.8827431702997732e-05,
+      "loss": 1.7466,
+      "step": 16190
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.6813267469406128,
+      "learning_rate": 1.8813807859877147e-05,
+      "loss": 1.925,
+      "step": 16195
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.9319490790367126,
+      "learning_rate": 1.8800185973443174e-05,
+      "loss": 1.9117,
+      "step": 16200
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.8836780190467834,
+      "learning_rate": 1.878656604800442e-05,
+      "loss": 2.0741,
+      "step": 16205
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.99202960729599,
+      "learning_rate": 1.8772948087868846e-05,
+      "loss": 1.7016,
+      "step": 16210
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 0.7995858788490295,
+      "learning_rate": 1.8759332097343797e-05,
+      "loss": 2.0791,
+      "step": 16215
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.3566383123397827,
+      "learning_rate": 1.874571808073601e-05,
+      "loss": 1.5975,
+      "step": 16220
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.005405306816101,
+      "learning_rate": 1.8732106042351572e-05,
+      "loss": 1.9958,
+      "step": 16225
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.589972198009491,
+      "learning_rate": 1.8718495986495967e-05,
+      "loss": 2.1223,
+      "step": 16230
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.41853928565979,
+      "learning_rate": 1.8704887917474042e-05,
+      "loss": 2.0237,
+      "step": 16235
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.5799002051353455,
+      "learning_rate": 1.8691281839590018e-05,
+      "loss": 1.8955,
+      "step": 16240
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.6206422448158264,
+      "learning_rate": 1.867767775714747e-05,
+      "loss": 1.9139,
+      "step": 16245
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.6891002655029297,
+      "learning_rate": 1.8664075674449376e-05,
+      "loss": 2.0885,
+      "step": 16250
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.5611288547515869,
+      "learning_rate": 1.865047559579804e-05,
+      "loss": 2.031,
+      "step": 16255
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.0590778589248657,
+      "learning_rate": 1.8636877525495173e-05,
+      "loss": 1.6809,
+      "step": 16260
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.43978947401046753,
+      "learning_rate": 1.8623281467841823e-05,
+      "loss": 1.9269,
+      "step": 16265
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.8819534182548523,
+      "learning_rate": 1.8609687427138407e-05,
+      "loss": 1.8985,
+      "step": 16270
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 1.4660248756408691,
+      "learning_rate": 1.859609540768471e-05,
+      "loss": 1.8683,
+      "step": 16275
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.7070092558860779,
+      "learning_rate": 1.8582505413779885e-05,
+      "loss": 1.7901,
+      "step": 16280
+    },
+    {
+      "epoch": 2.91,
+      "grad_norm": 0.8060388565063477,
+      "learning_rate": 1.8568917449722427e-05,
+      "loss": 1.6429,
+      "step": 16285
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.0174509286880493,
+      "learning_rate": 1.8555331519810188e-05,
+      "loss": 1.8506,
+      "step": 16290
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.0679351091384888,
+      "learning_rate": 1.8541747628340388e-05,
+      "loss": 1.7145,
+      "step": 16295
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.9246571660041809,
+      "learning_rate": 1.8528165779609603e-05,
+      "loss": 1.6584,
+      "step": 16300
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.5674405097961426,
+      "learning_rate": 1.851458597791375e-05,
+      "loss": 1.5784,
+      "step": 16305
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.738040566444397,
+      "learning_rate": 1.8501008227548123e-05,
+      "loss": 1.8847,
+      "step": 16310
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.651881754398346,
+      "learning_rate": 1.8487432532807335e-05,
+      "loss": 1.8891,
+      "step": 16315
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.8285610675811768,
+      "learning_rate": 1.847385889798538e-05,
+      "loss": 1.7669,
+      "step": 16320
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.5357155203819275,
+      "learning_rate": 1.8460287327375576e-05,
+      "loss": 1.9032,
+      "step": 16325
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7268362045288086,
+      "learning_rate": 1.8446717825270595e-05,
+      "loss": 1.841,
+      "step": 16330
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.7002309560775757,
+      "learning_rate": 1.8433150395962473e-05,
+      "loss": 2.0385,
+      "step": 16335
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.3066308498382568,
+      "learning_rate": 1.841958504374256e-05,
+      "loss": 1.8508,
+      "step": 16340
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.6260136961936951,
+      "learning_rate": 1.8406021772901566e-05,
+      "loss": 2.0906,
+      "step": 16345
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.523491382598877,
+      "learning_rate": 1.839246058772954e-05,
+      "loss": 1.9063,
+      "step": 16350
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.4432581663131714,
+      "learning_rate": 1.837890149251588e-05,
+      "loss": 2.143,
+      "step": 16355
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.09247887134552,
+      "learning_rate": 1.83653444915493e-05,
+      "loss": 1.9382,
+      "step": 16360
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.2057552337646484,
+      "learning_rate": 1.835178958911789e-05,
+      "loss": 1.8019,
+      "step": 16365
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.8609539866447449,
+      "learning_rate": 1.8338236789509024e-05,
+      "loss": 1.9375,
+      "step": 16370
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.2171326875686646,
+      "learning_rate": 1.832468609700946e-05,
+      "loss": 1.9367,
+      "step": 16375
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 1.5746159553527832,
+      "learning_rate": 1.831113751590527e-05,
+      "loss": 1.6848,
+      "step": 16380
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.9621313810348511,
+      "learning_rate": 1.8297591050481843e-05,
+      "loss": 1.7429,
+      "step": 16385
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.3991534113883972,
+      "learning_rate": 1.8284046705023915e-05,
+      "loss": 1.6909,
+      "step": 16390
+    },
+    {
+      "epoch": 2.93,
+      "grad_norm": 0.6495162844657898,
+      "learning_rate": 1.8270504483815558e-05,
+      "loss": 1.974,
+      "step": 16395
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.6558966040611267,
+      "learning_rate": 1.8256964391140164e-05,
+      "loss": 1.9515,
+      "step": 16400
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7836574912071228,
+      "learning_rate": 1.824342643128044e-05,
+      "loss": 1.9209,
+      "step": 16405
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.685867190361023,
+      "learning_rate": 1.822989060851844e-05,
+      "loss": 1.6289,
+      "step": 16410
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 2.84834361076355,
+      "learning_rate": 1.8216356927135525e-05,
+      "loss": 1.7064,
+      "step": 16415
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.5934162139892578,
+      "learning_rate": 1.8202825391412393e-05,
+      "loss": 1.7927,
+      "step": 16420
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.5504593253135681,
+      "learning_rate": 1.8192001710585122e-05,
+      "loss": 2.1149,
+      "step": 16425
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.6468228697776794,
+      "learning_rate": 1.817847404783477e-05,
+      "loss": 2.1836,
+      "step": 16430
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.6115560531616211,
+      "learning_rate": 1.8164948542726506e-05,
+      "loss": 1.7916,
+      "step": 16435
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.9618609547615051,
+      "learning_rate": 1.8151425199538447e-05,
+      "loss": 1.7555,
+      "step": 16440
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.8971806168556213,
+      "learning_rate": 1.813790402254801e-05,
+      "loss": 1.837,
+      "step": 16445
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.7342554926872253,
+      "learning_rate": 1.812438501603193e-05,
+      "loss": 1.9009,
+      "step": 16450
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.743880033493042,
+      "learning_rate": 1.811086818426625e-05,
+      "loss": 1.7526,
+      "step": 16455
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.2517614364624023,
+      "learning_rate": 1.8097353531526346e-05,
+      "loss": 1.9244,
+      "step": 16460
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.933478057384491,
+      "learning_rate": 1.8083841062086875e-05,
+      "loss": 1.7212,
+      "step": 16465
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.49430811405181885,
+      "learning_rate": 1.807033078022183e-05,
+      "loss": 1.997,
+      "step": 16470
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.0263806581497192,
+      "learning_rate": 1.8056822690204495e-05,
+      "loss": 1.9313,
+      "step": 16475
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.8995507955551147,
+      "learning_rate": 1.804331679630746e-05,
+      "loss": 1.7927,
+      "step": 16480
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.0173105001449585,
+      "learning_rate": 1.8029813102802624e-05,
+      "loss": 1.6534,
+      "step": 16485
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 1.0232231616973877,
+      "learning_rate": 1.8016311613961206e-05,
+      "loss": 1.9429,
+      "step": 16490
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.8685727715492249,
+      "learning_rate": 1.8002812334053703e-05,
+      "loss": 1.7248,
+      "step": 16495
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.9551222324371338,
+      "learning_rate": 1.7989315267349936e-05,
+      "loss": 1.978,
+      "step": 16500
+    },
+    {
+      "epoch": 2.95,
+      "grad_norm": 0.7746654748916626,
+      "learning_rate": 1.7975820418119e-05,
+      "loss": 2.0558,
+      "step": 16505
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.1619644165039062,
+      "learning_rate": 1.7962327790629308e-05,
+      "loss": 1.8762,
+      "step": 16510
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7726320028305054,
+      "learning_rate": 1.794883738914857e-05,
+      "loss": 1.74,
+      "step": 16515
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.9551023244857788,
+      "learning_rate": 1.793534921794379e-05,
+      "loss": 1.9455,
+      "step": 16520
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.1138999462127686,
+      "learning_rate": 1.792186328128125e-05,
+      "loss": 1.5777,
+      "step": 16525
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.5583162307739258,
+      "learning_rate": 1.7908379583426542e-05,
+      "loss": 1.774,
+      "step": 16530
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.6044538617134094,
+      "learning_rate": 1.7894898128644557e-05,
+      "loss": 1.7541,
+      "step": 16535
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.9106922149658203,
+      "learning_rate": 1.7881418921199454e-05,
+      "loss": 1.956,
+      "step": 16540
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.6646421551704407,
+      "learning_rate": 1.7867941965354706e-05,
+      "loss": 1.8526,
+      "step": 16545
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.8308748006820679,
+      "learning_rate": 1.7854467265373047e-05,
+      "loss": 1.6069,
+      "step": 16550
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.7582127451896667,
+      "learning_rate": 1.7840994825516523e-05,
+      "loss": 1.835,
+      "step": 16555
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.9956608414649963,
+      "learning_rate": 1.7827524650046447e-05,
+      "loss": 1.7829,
+      "step": 16560
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.3976410627365112,
+      "learning_rate": 1.7814056743223424e-05,
+      "loss": 1.8531,
+      "step": 16565
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.6910321712493896,
+      "learning_rate": 1.780059110930735e-05,
+      "loss": 1.8633,
+      "step": 16570
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.205909013748169,
+      "learning_rate": 1.7787127752557378e-05,
+      "loss": 1.9529,
+      "step": 16575
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.323149561882019,
+      "learning_rate": 1.777366667723196e-05,
+      "loss": 1.7019,
+      "step": 16580
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.8329334855079651,
+      "learning_rate": 1.776020788758882e-05,
+      "loss": 1.757,
+      "step": 16585
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 2.090344190597534,
+      "learning_rate": 1.7746751387884962e-05,
+      "loss": 1.7325,
+      "step": 16590
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 1.0283842086791992,
+      "learning_rate": 1.7733297182376663e-05,
+      "loss": 1.8203,
+      "step": 16595
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 2.198268175125122,
+      "learning_rate": 1.7719845275319473e-05,
+      "loss": 1.5753,
+      "step": 16600
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.6753280758857727,
+      "learning_rate": 1.7706395670968225e-05,
+      "loss": 1.7627,
+      "step": 16605
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.7107426524162292,
+      "learning_rate": 1.7692948373577003e-05,
+      "loss": 1.8784,
+      "step": 16610
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.8274734020233154,
+      "learning_rate": 1.7679503387399193e-05,
+      "loss": 2.058,
+      "step": 16615
+    },
+    {
+      "epoch": 2.97,
+      "grad_norm": 0.8373821973800659,
+      "learning_rate": 1.766606071668741e-05,
+      "loss": 2.1438,
+      "step": 16620
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.6570589542388916,
+      "learning_rate": 1.765262036569358e-05,
+      "loss": 2.1135,
+      "step": 16625
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.4979616105556488,
+      "learning_rate": 1.7639182338668847e-05,
+      "loss": 1.8345,
+      "step": 16630
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.0736268758773804,
+      "learning_rate": 1.7625746639863673e-05,
+      "loss": 1.9596,
+      "step": 16635
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7851144075393677,
+      "learning_rate": 1.761231327352773e-05,
+      "loss": 1.8697,
+      "step": 16640
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7080187201499939,
+      "learning_rate": 1.7598882243909997e-05,
+      "loss": 1.5315,
+      "step": 16645
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.8879645466804504,
+      "learning_rate": 1.7585453555258697e-05,
+      "loss": 2.0436,
+      "step": 16650
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.9378961324691772,
+      "learning_rate": 1.75720272118213e-05,
+      "loss": 2.0527,
+      "step": 16655
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.886584460735321,
+      "learning_rate": 1.755860321784456e-05,
+      "loss": 1.8691,
+      "step": 16660
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.5075838565826416,
+      "learning_rate": 1.7545181577574454e-05,
+      "loss": 1.7712,
+      "step": 16665
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.6004984974861145,
+      "learning_rate": 1.7531762295256254e-05,
+      "loss": 1.8251,
+      "step": 16670
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 0.7136458158493042,
+      "learning_rate": 1.7518345375134447e-05,
+      "loss": 1.7217,
+      "step": 16675
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.8066684603691101,
+      "learning_rate": 1.7504930821452808e-05,
+      "loss": 2.0242,
+      "step": 16680
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7061532735824585,
+      "learning_rate": 1.7491518638454336e-05,
+      "loss": 1.7245,
+      "step": 16685
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.7429984211921692,
+      "learning_rate": 1.7478108830381296e-05,
+      "loss": 1.6728,
+      "step": 16690
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.9607851505279541,
+      "learning_rate": 1.74647014014752e-05,
+      "loss": 1.3616,
+      "step": 16695
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.46374931931495667,
+      "learning_rate": 1.74512963559768e-05,
+      "loss": 1.7103,
+      "step": 16700
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.2484031915664673,
+      "learning_rate": 1.7437893698126102e-05,
+      "loss": 1.745,
+      "step": 16705
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.0168793201446533,
+      "learning_rate": 1.7424493432162366e-05,
+      "loss": 1.8125,
+      "step": 16710
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.9517542123794556,
+      "learning_rate": 1.7411095562324062e-05,
+      "loss": 1.7851,
+      "step": 16715
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.2883968353271484,
+      "learning_rate": 1.7397700092848927e-05,
+      "loss": 1.9432,
+      "step": 16720
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 0.5470624566078186,
+      "learning_rate": 1.7384307027973946e-05,
+      "loss": 1.8344,
+      "step": 16725
+    },
+    {
+      "epoch": 2.99,
+      "grad_norm": 1.2388825416564941,
+      "learning_rate": 1.7370916371935322e-05,
+      "loss": 1.8179,
+      "step": 16730
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.9549524784088135,
+      "learning_rate": 1.7357528128968513e-05,
+      "loss": 1.6871,
+      "step": 16735
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.6752102375030518,
+      "learning_rate": 1.7344142303308208e-05,
+      "loss": 1.9497,
+      "step": 16740
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.8145148158073425,
+      "learning_rate": 1.7330758899188322e-05,
+      "loss": 1.8557,
+      "step": 16745
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.9011813402175903,
+      "learning_rate": 1.7317377920842028e-05,
+      "loss": 2.0228,
+      "step": 16750
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.5068954825401306,
+      "learning_rate": 1.7303999372501705e-05,
+      "loss": 1.677,
+      "step": 16755
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.4357677102088928,
+      "learning_rate": 1.729062325839898e-05,
+      "loss": 2.1066,
+      "step": 16760
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.6191133856773376,
+      "learning_rate": 1.7277249582764697e-05,
+      "loss": 1.931,
+      "step": 16765
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.5191856026649475,
+      "learning_rate": 1.726387834982895e-05,
+      "loss": 1.5549,
+      "step": 16770
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7998222708702087,
+      "learning_rate": 1.7250509563821032e-05,
+      "loss": 1.5933,
+      "step": 16775
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.8126819133758545,
+      "learning_rate": 1.7237143228969488e-05,
+      "loss": 2.0985,
+      "step": 16780
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.6312077641487122,
+      "learning_rate": 1.7223779349502073e-05,
+      "loss": 1.9076,
+      "step": 16785
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.9065338969230652,
+      "learning_rate": 1.721041792964577e-05,
+      "loss": 1.7754,
+      "step": 16790
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.9230442047119141,
+      "learning_rate": 1.7197058973626785e-05,
+      "loss": 2.0562,
+      "step": 16795
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.8840682506561279,
+      "learning_rate": 1.7183702485670543e-05,
+      "loss": 1.9154,
+      "step": 16800
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.8318727612495422,
+      "learning_rate": 1.7170348470001688e-05,
+      "loss": 1.9867,
+      "step": 16805
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.6302641034126282,
+      "learning_rate": 1.7156996930844073e-05,
+      "loss": 1.9369,
+      "step": 16810
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.49491533637046814,
+      "learning_rate": 1.7143647872420793e-05,
+      "loss": 1.569,
+      "step": 16815
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 1.2738268375396729,
+      "learning_rate": 1.7130301298954126e-05,
+      "loss": 1.817,
+      "step": 16820
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.8552365899085999,
+      "learning_rate": 1.7116957214665587e-05,
+      "loss": 1.6973,
+      "step": 16825
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 1.1242488622665405,
+      "learning_rate": 1.710361562377591e-05,
+      "loss": 1.721,
+      "step": 16830
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 0.8571417927742004,
+      "learning_rate": 1.7090276530505004e-05,
+      "loss": 1.7216,
+      "step": 16835
+    },
+    {
+      "epoch": 3.01,
+      "grad_norm": 8.810229301452637,
+      "learning_rate": 1.707693993907203e-05,
+      "loss": 2.1472,
+      "step": 16840
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.7794356346130371,
+      "learning_rate": 1.706360585369533e-05,
+      "loss": 1.9728,
+      "step": 16845
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.5598608255386353,
+      "learning_rate": 1.705027427859247e-05,
+      "loss": 1.7381,
+      "step": 16850
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.5534650087356567,
+      "learning_rate": 1.7036945217980205e-05,
+      "loss": 1.9653,
+      "step": 16855
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 1.1248143911361694,
+      "learning_rate": 1.702361867607451e-05,
+      "loss": 1.8914,
+      "step": 16860
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 5.001594543457031,
+      "learning_rate": 1.7010294657090548e-05,
+      "loss": 1.7459,
+      "step": 16865
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.5881880521774292,
+      "learning_rate": 1.6996973165242704e-05,
+      "loss": 1.5695,
+      "step": 16870
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.516708493232727,
+      "learning_rate": 1.6983654204744552e-05,
+      "loss": 1.8567,
+      "step": 16875
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 1.0162904262542725,
+      "learning_rate": 1.6970337779808866e-05,
+      "loss": 1.7251,
+      "step": 16880
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.7917419075965881,
+      "learning_rate": 1.6957023894647614e-05,
+      "loss": 1.7346,
+      "step": 16885
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 0.6009075045585632,
+      "learning_rate": 1.6943712553471966e-05,
+      "loss": 2.0054,
+      "step": 16890
+    },
+    {
+      "epoch": 3.02,
+      "grad_norm": 1.0688451528549194,
+      "learning_rate": 1.6930403760492297e-05,
+      "loss": 1.7544,
+      "step": 16895
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.9357558488845825,
+      "learning_rate": 1.691709751991815e-05,
+      "loss": 1.9594,
+      "step": 16900
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.6222114562988281,
+      "learning_rate": 1.6903793835958283e-05,
+      "loss": 1.9862,
+      "step": 16905
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 1.3084805011749268,
+      "learning_rate": 1.6890492712820636e-05,
+      "loss": 1.7035,
+      "step": 16910
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.9012706279754639,
+      "learning_rate": 1.6877194154712342e-05,
+      "loss": 2.0331,
+      "step": 16915
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 2.868881940841675,
+      "learning_rate": 1.686389816583973e-05,
+      "loss": 1.6892,
+      "step": 16920
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.7443296909332275,
+      "learning_rate": 1.6850604750408296e-05,
+      "loss": 1.6757,
+      "step": 16925
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 1.3445460796356201,
+      "learning_rate": 1.6837313912622748e-05,
+      "loss": 1.9912,
+      "step": 16930
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.43768608570098877,
+      "learning_rate": 1.682402565668695e-05,
+      "loss": 1.843,
+      "step": 16935
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 1.3455616235733032,
+      "learning_rate": 1.6810739986803987e-05,
+      "loss": 1.9418,
+      "step": 16940
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.627327561378479,
+      "learning_rate": 1.6797456907176074e-05,
+      "loss": 1.3489,
+      "step": 16945
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 1.0638902187347412,
+      "learning_rate": 1.678417642200466e-05,
+      "loss": 1.8686,
+      "step": 16950
+    },
+    {
+      "epoch": 3.03,
+      "grad_norm": 0.8032281398773193,
+      "learning_rate": 1.6770898535490333e-05,
+      "loss": 1.6365,
+      "step": 16955
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.4566420912742615,
+      "learning_rate": 1.6757623251832887e-05,
+      "loss": 2.1101,
+      "step": 16960
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.840287446975708,
+      "learning_rate": 1.674435057523128e-05,
+      "loss": 1.5239,
+      "step": 16965
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 1.571700930595398,
+      "learning_rate": 1.673108050988364e-05,
+      "loss": 1.9991,
+      "step": 16970
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.6242066621780396,
+      "learning_rate": 1.6717813059987293e-05,
+      "loss": 2.0588,
+      "step": 16975
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 1.1033291816711426,
+      "learning_rate": 1.6704548229738697e-05,
+      "loss": 1.7984,
+      "step": 16980
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.7740524411201477,
+      "learning_rate": 1.6691286023333517e-05,
+      "loss": 1.5941,
+      "step": 16985
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.677450954914093,
+      "learning_rate": 1.6678026444966587e-05,
+      "loss": 1.9171,
+      "step": 16990
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.6410496234893799,
+      "learning_rate": 1.6664769498831884e-05,
+      "loss": 1.9519,
+      "step": 16995
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 1.0398961305618286,
+      "learning_rate": 1.6651515189122564e-05,
+      "loss": 1.7254,
+      "step": 17000
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.7212021350860596,
+      "learning_rate": 1.663826352003096e-05,
+      "loss": 1.9161,
+      "step": 17005
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.8684729337692261,
+      "learning_rate": 1.6625014495748566e-05,
+      "loss": 1.8468,
+      "step": 17010
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.5007403492927551,
+      "learning_rate": 1.661176812046603e-05,
+      "loss": 2.1323,
+      "step": 17015
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 1.1957733631134033,
+      "learning_rate": 1.659852439837317e-05,
+      "loss": 1.7557,
+      "step": 17020
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.935827910900116,
+      "learning_rate": 1.658528333365896e-05,
+      "loss": 1.9512,
+      "step": 17025
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.5936657786369324,
+      "learning_rate": 1.657204493051153e-05,
+      "loss": 1.8086,
+      "step": 17030
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.7128114700317383,
+      "learning_rate": 1.6558809193118202e-05,
+      "loss": 1.7417,
+      "step": 17035
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 1.5274105072021484,
+      "learning_rate": 1.6545576125665395e-05,
+      "loss": 2.0459,
+      "step": 17040
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 1.2550373077392578,
+      "learning_rate": 1.6532345732338724e-05,
+      "loss": 1.689,
+      "step": 17045
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.6413518786430359,
+      "learning_rate": 1.651911801732296e-05,
+      "loss": 1.9651,
+      "step": 17050
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.7453106045722961,
+      "learning_rate": 1.6505892984802e-05,
+      "loss": 1.8327,
+      "step": 17055
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 0.7208839058876038,
+      "learning_rate": 1.6492670638958924e-05,
+      "loss": 1.7667,
+      "step": 17060
+    },
+    {
+      "epoch": 3.05,
+      "grad_norm": 4.1532416343688965,
+      "learning_rate": 1.6479450983975946e-05,
+      "loss": 1.7473,
+      "step": 17065
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 1.0695805549621582,
+      "learning_rate": 1.646623402403442e-05,
+      "loss": 1.7827,
+      "step": 17070
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.7231548428535461,
+      "learning_rate": 1.6453019763314862e-05,
+      "loss": 1.8391,
+      "step": 17075
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.7367892265319824,
+      "learning_rate": 1.6439808205996942e-05,
+      "loss": 1.6613,
+      "step": 17080
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.9162821769714355,
+      "learning_rate": 1.642659935625945e-05,
+      "loss": 2.0619,
+      "step": 17085
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.6432048678398132,
+      "learning_rate": 1.6413393218280328e-05,
+      "loss": 2.0551,
+      "step": 17090
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 1.1106834411621094,
+      "learning_rate": 1.640018979623668e-05,
+      "loss": 1.7662,
+      "step": 17095
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.7890145778656006,
+      "learning_rate": 1.6386989094304715e-05,
+      "loss": 1.8003,
+      "step": 17100
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.9444851875305176,
+      "learning_rate": 1.6373791116659823e-05,
+      "loss": 1.7911,
+      "step": 17105
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 1.1127926111221313,
+      "learning_rate": 1.6360595867476504e-05,
+      "loss": 1.7545,
+      "step": 17110
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.997949481010437,
+      "learning_rate": 1.63474033509284e-05,
+      "loss": 2.0693,
+      "step": 17115
+    },
+    {
+      "epoch": 3.06,
+      "grad_norm": 0.6662190556526184,
+      "learning_rate": 1.6334213571188288e-05,
+      "loss": 1.7301,
+      "step": 17120
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 1.0271848440170288,
+      "learning_rate": 1.63210265324281e-05,
+      "loss": 1.9007,
+      "step": 17125
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 1.0467654466629028,
+      "learning_rate": 1.6307842238818874e-05,
+      "loss": 1.7783,
+      "step": 17130
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.8379384279251099,
+      "learning_rate": 1.6294660694530776e-05,
+      "loss": 1.629,
+      "step": 17135
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.7403329014778137,
+      "learning_rate": 1.628148190373313e-05,
+      "loss": 1.5475,
+      "step": 17140
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.8592426776885986,
+      "learning_rate": 1.626830587059437e-05,
+      "loss": 1.7707,
+      "step": 17145
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.8583236932754517,
+      "learning_rate": 1.6255132599282065e-05,
+      "loss": 1.785,
+      "step": 17150
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.5623230934143066,
+      "learning_rate": 1.624196209396291e-05,
+      "loss": 1.9508,
+      "step": 17155
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.7288662195205688,
+      "learning_rate": 1.622879435880272e-05,
+      "loss": 1.9152,
+      "step": 17160
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.5778056383132935,
+      "learning_rate": 1.621562939796643e-05,
+      "loss": 1.8435,
+      "step": 17165
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.7465753555297852,
+      "learning_rate": 1.620246721561812e-05,
+      "loss": 1.6645,
+      "step": 17170
+    },
+    {
+      "epoch": 3.07,
+      "grad_norm": 0.6527841687202454,
+      "learning_rate": 1.6189307815920973e-05,
+      "loss": 1.7769,
+      "step": 17175
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.9524030089378357,
+      "learning_rate": 1.617615120303728e-05,
+      "loss": 1.9306,
+      "step": 17180
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.8811773061752319,
+      "learning_rate": 1.6162997381128478e-05,
+      "loss": 1.9073,
+      "step": 17185
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.5965471863746643,
+      "learning_rate": 1.61498463543551e-05,
+      "loss": 2.1639,
+      "step": 17190
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.8437951803207397,
+      "learning_rate": 1.6136698126876806e-05,
+      "loss": 1.8132,
+      "step": 17195
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.6007370948791504,
+      "learning_rate": 1.6123552702852372e-05,
+      "loss": 1.9358,
+      "step": 17200
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.843265950679779,
+      "learning_rate": 1.611041008643968e-05,
+      "loss": 1.9867,
+      "step": 17205
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.7209492325782776,
+      "learning_rate": 1.6097270281795722e-05,
+      "loss": 2.0699,
+      "step": 17210
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 1.3360052108764648,
+      "learning_rate": 1.608413329307662e-05,
+      "loss": 1.8391,
+      "step": 17215
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.6539368033409119,
+      "learning_rate": 1.6070999124437586e-05,
+      "loss": 2.1616,
+      "step": 17220
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.6861870884895325,
+      "learning_rate": 1.605786778003293e-05,
+      "loss": 1.6254,
+      "step": 17225
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.804837703704834,
+      "learning_rate": 1.60447392640161e-05,
+      "loss": 1.6783,
+      "step": 17230
+    },
+    {
+      "epoch": 3.08,
+      "grad_norm": 0.64955735206604,
+      "learning_rate": 1.6031613580539617e-05,
+      "loss": 1.9525,
+      "step": 17235
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 1.0166507959365845,
+      "learning_rate": 1.6018490733755137e-05,
+      "loss": 1.7277,
+      "step": 17240
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.7587583661079407,
+      "learning_rate": 1.60053707278134e-05,
+      "loss": 1.7941,
+      "step": 17245
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.623679518699646,
+      "learning_rate": 1.5992253566864244e-05,
+      "loss": 1.8599,
+      "step": 17250
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.8486801385879517,
+      "learning_rate": 1.597913925505663e-05,
+      "loss": 1.7489,
+      "step": 17255
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.41435787081718445,
+      "learning_rate": 1.5966027796538586e-05,
+      "loss": 1.8594,
+      "step": 17260
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 2.0519394874572754,
+      "learning_rate": 1.595291919545726e-05,
+      "loss": 2.012,
+      "step": 17265
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 1.2812122106552124,
+      "learning_rate": 1.59398134559589e-05,
+      "loss": 2.0921,
+      "step": 17270
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 2.542084217071533,
+      "learning_rate": 1.5926710582188826e-05,
+      "loss": 1.8695,
+      "step": 17275
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.564811646938324,
+      "learning_rate": 1.591361057829146e-05,
+      "loss": 2.3541,
+      "step": 17280
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.9718075394630432,
+      "learning_rate": 1.5900513448410332e-05,
+      "loss": 1.6868,
+      "step": 17285
+    },
+    {
+      "epoch": 3.09,
+      "grad_norm": 0.9781553149223328,
+      "learning_rate": 1.588741919668805e-05,
+      "loss": 1.9755,
+      "step": 17290
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.5526376366615295,
+      "learning_rate": 1.5874327827266304e-05,
+      "loss": 2.0345,
+      "step": 17295
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 3.0677220821380615,
+      "learning_rate": 1.5861239344285893e-05,
+      "loss": 1.9055,
+      "step": 17300
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 1.1272741556167603,
+      "learning_rate": 1.5848153751886685e-05,
+      "loss": 2.0391,
+      "step": 17305
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.8001586198806763,
+      "learning_rate": 1.5835071054207643e-05,
+      "loss": 1.9079,
+      "step": 17310
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.9023197293281555,
+      "learning_rate": 1.582199125538682e-05,
+      "loss": 1.7115,
+      "step": 17315
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.6295259594917297,
+      "learning_rate": 1.5808914359561328e-05,
+      "loss": 1.819,
+      "step": 17320
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.5197964310646057,
+      "learning_rate": 1.579584037086738e-05,
+      "loss": 1.8448,
+      "step": 17325
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.8185873031616211,
+      "learning_rate": 1.578276929344027e-05,
+      "loss": 1.9797,
+      "step": 17330
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 1.1019208431243896,
+      "learning_rate": 1.5769701131414372e-05,
+      "loss": 1.5105,
+      "step": 17335
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.8213723301887512,
+      "learning_rate": 1.575663588892312e-05,
+      "loss": 1.8677,
+      "step": 17340
+    },
+    {
+      "epoch": 3.1,
+      "grad_norm": 0.8821647763252258,
+      "learning_rate": 1.574357357009905e-05,
+      "loss": 1.7404,
+      "step": 17345
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 1.0914586782455444,
+      "learning_rate": 1.5730514179073756e-05,
+      "loss": 1.6639,
+      "step": 17350
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.6956515908241272,
+      "learning_rate": 1.5717457719977906e-05,
+      "loss": 1.9229,
+      "step": 17355
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 1.160091757774353,
+      "learning_rate": 1.5704404196941265e-05,
+      "loss": 1.4537,
+      "step": 17360
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.5331469774246216,
+      "learning_rate": 1.5691353614092627e-05,
+      "loss": 1.842,
+      "step": 17365
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.9176687598228455,
+      "learning_rate": 1.5678305975559882e-05,
+      "loss": 1.5422,
+      "step": 17370
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.5114845633506775,
+      "learning_rate": 1.5665261285469992e-05,
+      "loss": 2.0008,
+      "step": 17375
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.7605366706848145,
+      "learning_rate": 1.5652219547948982e-05,
+      "loss": 1.871,
+      "step": 17380
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 1.129357933998108,
+      "learning_rate": 1.5639180767121938e-05,
+      "loss": 1.7841,
+      "step": 17385
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.8872073292732239,
+      "learning_rate": 1.562614494711301e-05,
+      "loss": 1.8101,
+      "step": 17390
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 1.4235471487045288,
+      "learning_rate": 1.5613112092045418e-05,
+      "loss": 2.1259,
+      "step": 17395
+    },
+    {
+      "epoch": 3.11,
+      "grad_norm": 0.5211644172668457,
+      "learning_rate": 1.5600082206041443e-05,
+      "loss": 2.0976,
+      "step": 17400
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 1.07545804977417,
+      "learning_rate": 1.558705529322243e-05,
+      "loss": 2.0866,
+      "step": 17405
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.8453513383865356,
+      "learning_rate": 1.5574031357708778e-05,
+      "loss": 1.822,
+      "step": 17410
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.5668812394142151,
+      "learning_rate": 1.556101040361993e-05,
+      "loss": 1.8889,
+      "step": 17415
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.6529809236526489,
+      "learning_rate": 1.554799243507441e-05,
+      "loss": 2.2048,
+      "step": 17420
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 1.1768380403518677,
+      "learning_rate": 1.5534977456189797e-05,
+      "loss": 2.0006,
+      "step": 17425
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.7664825320243835,
+      "learning_rate": 1.552196547108271e-05,
+      "loss": 1.8376,
+      "step": 17430
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.5944017767906189,
+      "learning_rate": 1.550895648386883e-05,
+      "loss": 1.7861,
+      "step": 17435
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 1.4491263628005981,
+      "learning_rate": 1.5495950498662877e-05,
+      "loss": 1.8945,
+      "step": 17440
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.961166501045227,
+      "learning_rate": 1.5482947519578645e-05,
+      "loss": 1.9871,
+      "step": 17445
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 1.2184598445892334,
+      "learning_rate": 1.5469947550728958e-05,
+      "loss": 1.7694,
+      "step": 17450
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 1.1210654973983765,
+      "learning_rate": 1.5456950596225703e-05,
+      "loss": 1.6446,
+      "step": 17455
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.3188503682613373,
+      "learning_rate": 1.5443956660179783e-05,
+      "loss": 1.8352,
+      "step": 17460
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.4729178249835968,
+      "learning_rate": 1.5430965746701183e-05,
+      "loss": 1.9481,
+      "step": 17465
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.6622750163078308,
+      "learning_rate": 1.5417977859898914e-05,
+      "loss": 1.9198,
+      "step": 17470
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 2.716395854949951,
+      "learning_rate": 1.5404993003881027e-05,
+      "loss": 1.7338,
+      "step": 17475
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.9567592144012451,
+      "learning_rate": 1.5392011182754627e-05,
+      "loss": 1.5414,
+      "step": 17480
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.6734287738800049,
+      "learning_rate": 1.537903240062585e-05,
+      "loss": 2.1804,
+      "step": 17485
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.9989105463027954,
+      "learning_rate": 1.5366056661599866e-05,
+      "loss": 1.6213,
+      "step": 17490
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 1.0770398378372192,
+      "learning_rate": 1.5353083969780897e-05,
+      "loss": 1.9834,
+      "step": 17495
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 1.3205726146697998,
+      "learning_rate": 1.5340114329272194e-05,
+      "loss": 1.8774,
+      "step": 17500
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.6853746771812439,
+      "learning_rate": 1.532714774417603e-05,
+      "loss": 1.8502,
+      "step": 17505
+    },
+    {
+      "epoch": 3.13,
+      "grad_norm": 0.6563792824745178,
+      "learning_rate": 1.5314184218593724e-05,
+      "loss": 1.6979,
+      "step": 17510
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.8047040104866028,
+      "learning_rate": 1.5301223756625643e-05,
+      "loss": 1.5599,
+      "step": 17515
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 1.3527451753616333,
+      "learning_rate": 1.5288266362371155e-05,
+      "loss": 1.8567,
+      "step": 17520
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.8321057558059692,
+      "learning_rate": 1.527531203992868e-05,
+      "loss": 1.9165,
+      "step": 17525
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.5902025699615479,
+      "learning_rate": 1.526236079339565e-05,
+      "loss": 1.8397,
+      "step": 17530
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.8331001400947571,
+      "learning_rate": 1.5249412626868526e-05,
+      "loss": 1.9709,
+      "step": 17535
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 1.3767504692077637,
+      "learning_rate": 1.5236467544442821e-05,
+      "loss": 1.9877,
+      "step": 17540
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.8942415714263916,
+      "learning_rate": 1.5223525550213034e-05,
+      "loss": 1.8899,
+      "step": 17545
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 1.9923580884933472,
+      "learning_rate": 1.521058664827272e-05,
+      "loss": 1.6486,
+      "step": 17550
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.7297930121421814,
+      "learning_rate": 1.5197650842714426e-05,
+      "loss": 1.8752,
+      "step": 17555
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.7140510082244873,
+      "learning_rate": 1.5184718137629745e-05,
+      "loss": 1.8954,
+      "step": 17560
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.7358191013336182,
+      "learning_rate": 1.5171788537109272e-05,
+      "loss": 2.0138,
+      "step": 17565
+    },
+    {
+      "epoch": 3.14,
+      "grad_norm": 0.8002922534942627,
+      "learning_rate": 1.5158862045242633e-05,
+      "loss": 2.0375,
+      "step": 17570
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.6648169755935669,
+      "learning_rate": 1.5145938666118459e-05,
+      "loss": 1.9708,
+      "step": 17575
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 1.1531585454940796,
+      "learning_rate": 1.5133018403824411e-05,
+      "loss": 1.72,
+      "step": 17580
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.5088279843330383,
+      "learning_rate": 1.512010126244715e-05,
+      "loss": 1.7109,
+      "step": 17585
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.825671911239624,
+      "learning_rate": 1.5107187246072357e-05,
+      "loss": 1.7681,
+      "step": 17590
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 1.1927964687347412,
+      "learning_rate": 1.5094276358784728e-05,
+      "loss": 1.897,
+      "step": 17595
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.9928845763206482,
+      "learning_rate": 1.5081368604667956e-05,
+      "loss": 1.8343,
+      "step": 17600
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.9306197166442871,
+      "learning_rate": 1.5068463987804748e-05,
+      "loss": 1.9001,
+      "step": 17605
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.7192040681838989,
+      "learning_rate": 1.5055562512276827e-05,
+      "loss": 1.8987,
+      "step": 17610
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 2.9316582679748535,
+      "learning_rate": 1.5042664182164926e-05,
+      "loss": 1.9867,
+      "step": 17615
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 0.7092540860176086,
+      "learning_rate": 1.502976900154876e-05,
+      "loss": 1.9289,
+      "step": 17620
+    },
+    {
+      "epoch": 3.15,
+      "grad_norm": 3.189732313156128,
+      "learning_rate": 1.5016876974507065e-05,
+      "loss": 1.6191,
+      "step": 17625
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.6250412464141846,
+      "learning_rate": 1.500398810511759e-05,
+      "loss": 1.7705,
+      "step": 17630
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.9590062499046326,
+      "learning_rate": 1.4991102397457058e-05,
+      "loss": 1.6913,
+      "step": 17635
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.5470601916313171,
+      "learning_rate": 1.4978219855601216e-05,
+      "loss": 1.912,
+      "step": 17640
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.9383836984634399,
+      "learning_rate": 1.496534048362479e-05,
+      "loss": 1.8818,
+      "step": 17645
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.911783754825592,
+      "learning_rate": 1.4952464285601515e-05,
+      "loss": 1.9046,
+      "step": 17650
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 2.2841453552246094,
+      "learning_rate": 1.493959126560412e-05,
+      "loss": 1.5841,
+      "step": 17655
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.586314857006073,
+      "learning_rate": 1.4926721427704335e-05,
+      "loss": 1.8118,
+      "step": 17660
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.8905709981918335,
+      "learning_rate": 1.4913854775972869e-05,
+      "loss": 1.8675,
+      "step": 17665
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 1.4882047176361084,
+      "learning_rate": 1.4900991314479434e-05,
+      "loss": 1.5654,
+      "step": 17670
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 0.8829624652862549,
+      "learning_rate": 1.4888131047292742e-05,
+      "loss": 1.697,
+      "step": 17675
+    },
+    {
+      "epoch": 3.16,
+      "grad_norm": 1.1058586835861206,
+      "learning_rate": 1.4875273978480464e-05,
+      "loss": 1.7776,
+      "step": 17680
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 1.0767732858657837,
+      "learning_rate": 1.4862420112109293e-05,
+      "loss": 1.9039,
+      "step": 17685
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.7137855291366577,
+      "learning_rate": 1.48495694522449e-05,
+      "loss": 1.7499,
+      "step": 17690
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.44608253240585327,
+      "learning_rate": 1.4836722002951908e-05,
+      "loss": 1.8009,
+      "step": 17695
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 1.0002092123031616,
+      "learning_rate": 1.4823877768293975e-05,
+      "loss": 1.4692,
+      "step": 17700
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.8534446358680725,
+      "learning_rate": 1.4811036752333718e-05,
+      "loss": 1.6299,
+      "step": 17705
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.4848932921886444,
+      "learning_rate": 1.4798198959132733e-05,
+      "loss": 2.0872,
+      "step": 17710
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.5030261874198914,
+      "learning_rate": 1.4785364392751606e-05,
+      "loss": 2.0136,
+      "step": 17715
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.644991397857666,
+      "learning_rate": 1.4772533057249898e-05,
+      "loss": 1.8112,
+      "step": 17720
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.925753653049469,
+      "learning_rate": 1.4759704956686144e-05,
+      "loss": 1.55,
+      "step": 17725
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 0.7059409618377686,
+      "learning_rate": 1.4746880095117866e-05,
+      "loss": 1.8275,
+      "step": 17730
+    },
+    {
+      "epoch": 3.17,
+      "grad_norm": 1.2741386890411377,
+      "learning_rate": 1.4734058476601553e-05,
+      "loss": 1.9445,
+      "step": 17735
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.7439374923706055,
+      "learning_rate": 1.4721240105192663e-05,
+      "loss": 1.7957,
+      "step": 17740
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.9652694463729858,
+      "learning_rate": 1.470842498494564e-05,
+      "loss": 1.8712,
+      "step": 17745
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.7117319107055664,
+      "learning_rate": 1.4695613119913903e-05,
+      "loss": 1.6265,
+      "step": 17750
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 1.2906383275985718,
+      "learning_rate": 1.4682804514149818e-05,
+      "loss": 1.7699,
+      "step": 17755
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 1.508944034576416,
+      "learning_rate": 1.466999917170474e-05,
+      "loss": 1.7239,
+      "step": 17760
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.925870954990387,
+      "learning_rate": 1.4657197096628994e-05,
+      "loss": 2.3444,
+      "step": 17765
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 1.065234899520874,
+      "learning_rate": 1.464439829297185e-05,
+      "loss": 1.5987,
+      "step": 17770
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.7614918351173401,
+      "learning_rate": 1.4631602764781572e-05,
+      "loss": 2.0636,
+      "step": 17775
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.5512299537658691,
+      "learning_rate": 1.4618810516105371e-05,
+      "loss": 1.9394,
+      "step": 17780
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.7986428141593933,
+      "learning_rate": 1.460602155098941e-05,
+      "loss": 1.8252,
+      "step": 17785
+    },
+    {
+      "epoch": 3.18,
+      "grad_norm": 0.6706057190895081,
+      "learning_rate": 1.4593235873478839e-05,
+      "loss": 1.8949,
+      "step": 17790
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.6929710507392883,
+      "learning_rate": 1.4580453487617745e-05,
+      "loss": 1.8602,
+      "step": 17795
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.9884799122810364,
+      "learning_rate": 1.456767439744919e-05,
+      "loss": 1.6815,
+      "step": 17800
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 1.1615020036697388,
+      "learning_rate": 1.455489860701519e-05,
+      "loss": 1.8135,
+      "step": 17805
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.9147149920463562,
+      "learning_rate": 1.4542126120356714e-05,
+      "loss": 1.7469,
+      "step": 17810
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.612957239151001,
+      "learning_rate": 1.4529356941513697e-05,
+      "loss": 1.8893,
+      "step": 17815
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.8941155076026917,
+      "learning_rate": 1.4516591074524999e-05,
+      "loss": 2.1368,
+      "step": 17820
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.798802375793457,
+      "learning_rate": 1.4503828523428458e-05,
+      "loss": 1.7414,
+      "step": 17825
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 1.113351821899414,
+      "learning_rate": 1.4491069292260868e-05,
+      "loss": 1.9364,
+      "step": 17830
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.9882722496986389,
+      "learning_rate": 1.4478313385057946e-05,
+      "loss": 1.75,
+      "step": 17835
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.7189344167709351,
+      "learning_rate": 1.4465560805854378e-05,
+      "loss": 1.9663,
+      "step": 17840
+    },
+    {
+      "epoch": 3.19,
+      "grad_norm": 0.7250414490699768,
+      "learning_rate": 1.4452811558683804e-05,
+      "loss": 1.8795,
+      "step": 17845
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.8930559754371643,
+      "learning_rate": 1.4440065647578777e-05,
+      "loss": 1.9817,
+      "step": 17850
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.7246281504631042,
+      "learning_rate": 1.4427323076570831e-05,
+      "loss": 1.9332,
+      "step": 17855
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.7635440230369568,
+      "learning_rate": 1.4414583849690428e-05,
+      "loss": 1.8095,
+      "step": 17860
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.7749634981155396,
+      "learning_rate": 1.4401847970966966e-05,
+      "loss": 1.7567,
+      "step": 17865
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.7701842784881592,
+      "learning_rate": 1.4389115444428811e-05,
+      "loss": 1.8668,
+      "step": 17870
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 2.8975706100463867,
+      "learning_rate": 1.4376386274103231e-05,
+      "loss": 1.7546,
+      "step": 17875
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 1.0762674808502197,
+      "learning_rate": 1.436366046401645e-05,
+      "loss": 1.8692,
+      "step": 17880
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.6440033316612244,
+      "learning_rate": 1.4350938018193626e-05,
+      "loss": 1.6009,
+      "step": 17885
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.7783496975898743,
+      "learning_rate": 1.433821894065887e-05,
+      "loss": 1.7383,
+      "step": 17890
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.7856084704399109,
+      "learning_rate": 1.4325503235435207e-05,
+      "loss": 2.0661,
+      "step": 17895
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.7685439586639404,
+      "learning_rate": 1.4312790906544598e-05,
+      "loss": 1.596,
+      "step": 17900
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 0.8082401156425476,
+      "learning_rate": 1.430008195800796e-05,
+      "loss": 1.8206,
+      "step": 17905
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.9814727306365967,
+      "learning_rate": 1.4287376393845095e-05,
+      "loss": 1.61,
+      "step": 17910
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.779301106929779,
+      "learning_rate": 1.4274674218074775e-05,
+      "loss": 1.5867,
+      "step": 17915
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 1.0549635887145996,
+      "learning_rate": 1.4261975434714686e-05,
+      "loss": 1.9271,
+      "step": 17920
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.9940875768661499,
+      "learning_rate": 1.4249280047781435e-05,
+      "loss": 2.0389,
+      "step": 17925
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.7739433646202087,
+      "learning_rate": 1.4236588061290563e-05,
+      "loss": 1.9533,
+      "step": 17930
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.6601684093475342,
+      "learning_rate": 1.4223899479256542e-05,
+      "loss": 1.6776,
+      "step": 17935
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.9212928414344788,
+      "learning_rate": 1.4211214305692744e-05,
+      "loss": 1.8223,
+      "step": 17940
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.4190042018890381,
+      "learning_rate": 1.419853254461148e-05,
+      "loss": 1.9103,
+      "step": 17945
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 1.1490883827209473,
+      "learning_rate": 1.4185854200023984e-05,
+      "loss": 1.5747,
+      "step": 17950
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 0.8107929825782776,
+      "learning_rate": 1.4173179275940401e-05,
+      "loss": 1.9681,
+      "step": 17955
+    },
+    {
+      "epoch": 3.21,
+      "grad_norm": 1.9732346534729004,
+      "learning_rate": 1.4160507776369813e-05,
+      "loss": 1.5155,
+      "step": 17960
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.7424911856651306,
+      "learning_rate": 1.4147839705320177e-05,
+      "loss": 1.8159,
+      "step": 17965
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 1.0739436149597168,
+      "learning_rate": 1.4135175066798412e-05,
+      "loss": 1.7462,
+      "step": 17970
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.9837630987167358,
+      "learning_rate": 1.4122513864810317e-05,
+      "loss": 1.5915,
+      "step": 17975
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.7608397006988525,
+      "learning_rate": 1.410985610336062e-05,
+      "loss": 1.9469,
+      "step": 17980
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.3973992168903351,
+      "learning_rate": 1.4097201786452963e-05,
+      "loss": 1.9932,
+      "step": 17985
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.6849148273468018,
+      "learning_rate": 1.4084550918089895e-05,
+      "loss": 1.4885,
+      "step": 17990
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.8641477227210999,
+      "learning_rate": 1.4071903502272882e-05,
+      "loss": 1.9552,
+      "step": 17995
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 1.1732696294784546,
+      "learning_rate": 1.4059259543002267e-05,
+      "loss": 1.7839,
+      "step": 18000
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.6170725226402283,
+      "learning_rate": 1.4046619044277337e-05,
+      "loss": 2.0093,
+      "step": 18005
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.7883812785148621,
+      "learning_rate": 1.4033982010096269e-05,
+      "loss": 1.5738,
+      "step": 18010
+    },
+    {
+      "epoch": 3.22,
+      "grad_norm": 0.8948926329612732,
+      "learning_rate": 1.402134844445615e-05,
+      "loss": 1.7868,
+      "step": 18015
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.8349456787109375,
+      "learning_rate": 1.400871835135295e-05,
+      "loss": 1.7468,
+      "step": 18020
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.6893956661224365,
+      "learning_rate": 1.3996091734781575e-05,
+      "loss": 1.8127,
+      "step": 18025
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.509495198726654,
+      "learning_rate": 1.3983468598735786e-05,
+      "loss": 1.6751,
+      "step": 18030
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.7317885756492615,
+      "learning_rate": 1.3970848947208284e-05,
+      "loss": 2.0498,
+      "step": 18035
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.7369408011436462,
+      "learning_rate": 1.395823278419065e-05,
+      "loss": 1.6929,
+      "step": 18040
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 1.6330751180648804,
+      "learning_rate": 1.3945620113673369e-05,
+      "loss": 1.5587,
+      "step": 18045
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.796711802482605,
+      "learning_rate": 1.3933010939645821e-05,
+      "loss": 1.7163,
+      "step": 18050
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 1.373434066772461,
+      "learning_rate": 1.392040526609626e-05,
+      "loss": 1.7293,
+      "step": 18055
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.7607307434082031,
+      "learning_rate": 1.3907803097011862e-05,
+      "loss": 1.6402,
+      "step": 18060
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 1.729021668434143,
+      "learning_rate": 1.3895204436378667e-05,
+      "loss": 2.0188,
+      "step": 18065
+    },
+    {
+      "epoch": 3.23,
+      "grad_norm": 0.6117586493492126,
+      "learning_rate": 1.3882609288181628e-05,
+      "loss": 1.5768,
+      "step": 18070
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 1.12637460231781,
+      "learning_rate": 1.3870017656404577e-05,
+      "loss": 1.6816,
+      "step": 18075
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 1.0976755619049072,
+      "learning_rate": 1.3857429545030246e-05,
+      "loss": 1.8734,
+      "step": 18080
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.9964720010757446,
+      "learning_rate": 1.384484495804022e-05,
+      "loss": 1.8126,
+      "step": 18085
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.9928443431854248,
+      "learning_rate": 1.3832263899415006e-05,
+      "loss": 1.8391,
+      "step": 18090
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 1.7136353254318237,
+      "learning_rate": 1.381968637313398e-05,
+      "loss": 1.5318,
+      "step": 18095
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.8411396741867065,
+      "learning_rate": 1.3807112383175396e-05,
+      "loss": 1.89,
+      "step": 18100
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.641628623008728,
+      "learning_rate": 1.3794541933516402e-05,
+      "loss": 1.9114,
+      "step": 18105
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.8874943852424622,
+      "learning_rate": 1.3781975028133027e-05,
+      "loss": 2.0295,
+      "step": 18110
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.7927294373512268,
+      "learning_rate": 1.376941167100016e-05,
+      "loss": 1.7748,
+      "step": 18115
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 1.0050328969955444,
+      "learning_rate": 1.3756851866091566e-05,
+      "loss": 2.0182,
+      "step": 18120
+    },
+    {
+      "epoch": 3.24,
+      "grad_norm": 0.8478461503982544,
+      "learning_rate": 1.3744295617379916e-05,
+      "loss": 1.7941,
+      "step": 18125
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.9578462243080139,
+      "learning_rate": 1.3731742928836732e-05,
+      "loss": 1.8209,
+      "step": 18130
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.9452106356620789,
+      "learning_rate": 1.3719193804432423e-05,
+      "loss": 1.8019,
+      "step": 18135
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.9227070808410645,
+      "learning_rate": 1.3706648248136272e-05,
+      "loss": 2.0451,
+      "step": 18140
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.7945765256881714,
+      "learning_rate": 1.3694106263916406e-05,
+      "loss": 1.8797,
+      "step": 18145
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.7952302098274231,
+      "learning_rate": 1.3681567855739857e-05,
+      "loss": 1.8476,
+      "step": 18150
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.5315759778022766,
+      "learning_rate": 1.3669033027572518e-05,
+      "loss": 1.8467,
+      "step": 18155
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.5972190499305725,
+      "learning_rate": 1.3656501783379122e-05,
+      "loss": 1.7655,
+      "step": 18160
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.8763524293899536,
+      "learning_rate": 1.3643974127123298e-05,
+      "loss": 1.7197,
+      "step": 18165
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.714412271976471,
+      "learning_rate": 1.363145006276755e-05,
+      "loss": 1.8869,
+      "step": 18170
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.8367454409599304,
+      "learning_rate": 1.36189295942732e-05,
+      "loss": 1.7424,
+      "step": 18175
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.7740775942802429,
+      "learning_rate": 1.3606412725600471e-05,
+      "loss": 1.7942,
+      "step": 18180
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.6157845854759216,
+      "learning_rate": 1.359389946070844e-05,
+      "loss": 2.0431,
+      "step": 18185
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.6158450245857239,
+      "learning_rate": 1.3581389803555036e-05,
+      "loss": 2.0421,
+      "step": 18190
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 1.7718108892440796,
+      "learning_rate": 1.3568883758097056e-05,
+      "loss": 1.6756,
+      "step": 18195
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 1.0088618993759155,
+      "learning_rate": 1.3556381328290154e-05,
+      "loss": 1.6387,
+      "step": 18200
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 2.629340171813965,
+      "learning_rate": 1.3543882518088835e-05,
+      "loss": 1.8785,
+      "step": 18205
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 2.2919623851776123,
+      "learning_rate": 1.3531387331446444e-05,
+      "loss": 1.9667,
+      "step": 18210
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.5759145617485046,
+      "learning_rate": 1.3518895772315207e-05,
+      "loss": 1.8605,
+      "step": 18215
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.9002760648727417,
+      "learning_rate": 1.3506407844646196e-05,
+      "loss": 1.8257,
+      "step": 18220
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.7235890030860901,
+      "learning_rate": 1.3493923552389326e-05,
+      "loss": 1.8972,
+      "step": 18225
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.6796399354934692,
+      "learning_rate": 1.3481442899493382e-05,
+      "loss": 1.7873,
+      "step": 18230
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 2.119182586669922,
+      "learning_rate": 1.346896588990596e-05,
+      "loss": 1.9575,
+      "step": 18235
+    },
+    {
+      "epoch": 3.26,
+      "grad_norm": 0.8993787169456482,
+      "learning_rate": 1.3456492527573533e-05,
+      "loss": 1.5285,
+      "step": 18240
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 1.0154492855072021,
+      "learning_rate": 1.3444022816441414e-05,
+      "loss": 1.89,
+      "step": 18245
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.5175526738166809,
+      "learning_rate": 1.3431556760453778e-05,
+      "loss": 1.8722,
+      "step": 18250
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.8413906097412109,
+      "learning_rate": 1.3419094363553597e-05,
+      "loss": 1.6884,
+      "step": 18255
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.7729244232177734,
+      "learning_rate": 1.3406635629682743e-05,
+      "loss": 1.8852,
+      "step": 18260
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.7230870127677917,
+      "learning_rate": 1.3394180562781877e-05,
+      "loss": 1.9346,
+      "step": 18265
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.7597714066505432,
+      "learning_rate": 1.3381729166790535e-05,
+      "loss": 1.7774,
+      "step": 18270
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 2.0124258995056152,
+      "learning_rate": 1.336928144564708e-05,
+      "loss": 1.8878,
+      "step": 18275
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.7441228628158569,
+      "learning_rate": 1.3356837403288719e-05,
+      "loss": 1.9835,
+      "step": 18280
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.8413273692131042,
+      "learning_rate": 1.3344397043651496e-05,
+      "loss": 1.7068,
+      "step": 18285
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 0.9278790950775146,
+      "learning_rate": 1.333196037067027e-05,
+      "loss": 1.572,
+      "step": 18290
+    },
+    {
+      "epoch": 3.27,
+      "grad_norm": 1.6057875156402588,
+      "learning_rate": 1.3319527388278764e-05,
+      "loss": 1.6434,
+      "step": 18295
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.7727483510971069,
+      "learning_rate": 1.33070981004095e-05,
+      "loss": 1.8625,
+      "step": 18300
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.9005012512207031,
+      "learning_rate": 1.329467251099386e-05,
+      "loss": 1.8556,
+      "step": 18305
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.4766554832458496,
+      "learning_rate": 1.3282250623962045e-05,
+      "loss": 1.7374,
+      "step": 18310
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 1.3196171522140503,
+      "learning_rate": 1.3269832443243085e-05,
+      "loss": 1.6436,
+      "step": 18315
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.6074075698852539,
+      "learning_rate": 1.3257417972764852e-05,
+      "loss": 1.7342,
+      "step": 18320
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 1.836259126663208,
+      "learning_rate": 1.3245007216454008e-05,
+      "loss": 1.7356,
+      "step": 18325
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 1.7985725402832031,
+      "learning_rate": 1.3232600178236077e-05,
+      "loss": 2.0223,
+      "step": 18330
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 1.469251275062561,
+      "learning_rate": 1.3220196862035388e-05,
+      "loss": 1.8288,
+      "step": 18335
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.7703099250793457,
+      "learning_rate": 1.3207797271775113e-05,
+      "loss": 1.6407,
+      "step": 18340
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 4.6768717765808105,
+      "learning_rate": 1.319540141137721e-05,
+      "loss": 1.9388,
+      "step": 18345
+    },
+    {
+      "epoch": 3.28,
+      "grad_norm": 0.891442596912384,
+      "learning_rate": 1.3183009284762492e-05,
+      "loss": 1.8492,
+      "step": 18350
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.5014522075653076,
+      "learning_rate": 1.3170620895850563e-05,
+      "loss": 1.8215,
+      "step": 18355
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 2.594623327255249,
+      "learning_rate": 1.3158236248559866e-05,
+      "loss": 1.6727,
+      "step": 18360
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 2.9839723110198975,
+      "learning_rate": 1.3145855346807651e-05,
+      "loss": 1.3427,
+      "step": 18365
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.809145450592041,
+      "learning_rate": 1.313347819450999e-05,
+      "loss": 2.0794,
+      "step": 18370
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 1.4444146156311035,
+      "learning_rate": 1.3121104795581774e-05,
+      "loss": 1.8111,
+      "step": 18375
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 3.0015711784362793,
+      "learning_rate": 1.3108735153936669e-05,
+      "loss": 2.0326,
+      "step": 18380
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.8941876292228699,
+      "learning_rate": 1.3096369273487202e-05,
+      "loss": 1.8635,
+      "step": 18385
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.9546188116073608,
+      "learning_rate": 1.308400715814469e-05,
+      "loss": 1.6841,
+      "step": 18390
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 2.428107500076294,
+      "learning_rate": 1.3071648811819248e-05,
+      "loss": 1.8043,
+      "step": 18395
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 1.22321617603302,
+      "learning_rate": 1.3059294238419809e-05,
+      "loss": 1.7878,
+      "step": 18400
+    },
+    {
+      "epoch": 3.29,
+      "grad_norm": 0.8854243755340576,
+      "learning_rate": 1.3046943441854115e-05,
+      "loss": 1.8578,
+      "step": 18405
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.663280189037323,
+      "learning_rate": 1.3034596426028727e-05,
+      "loss": 1.9011,
+      "step": 18410
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.9913401007652283,
+      "learning_rate": 1.3022253194848968e-05,
+      "loss": 1.7046,
+      "step": 18415
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.7245838046073914,
+      "learning_rate": 1.3009913752219006e-05,
+      "loss": 2.0733,
+      "step": 18420
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.7399362921714783,
+      "learning_rate": 1.2997578102041785e-05,
+      "loss": 1.7638,
+      "step": 18425
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.6854051351547241,
+      "learning_rate": 1.2985246248219068e-05,
+      "loss": 1.6586,
+      "step": 18430
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 3.867100715637207,
+      "learning_rate": 1.2972918194651417e-05,
+      "loss": 1.7975,
+      "step": 18435
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.889519214630127,
+      "learning_rate": 1.2960593945238173e-05,
+      "loss": 1.6503,
+      "step": 18440
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 1.0203913450241089,
+      "learning_rate": 1.2948273503877473e-05,
+      "loss": 1.8388,
+      "step": 18445
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.6218311786651611,
+      "learning_rate": 1.293595687446627e-05,
+      "loss": 1.8764,
+      "step": 18450
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 0.6152927875518799,
+      "learning_rate": 1.2923644060900305e-05,
+      "loss": 1.722,
+      "step": 18455
+    },
+    {
+      "epoch": 3.3,
+      "grad_norm": 1.706099033355713,
+      "learning_rate": 1.2911335067074107e-05,
+      "loss": 1.5927,
+      "step": 18460
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.9180427193641663,
+      "learning_rate": 1.289902989688101e-05,
+      "loss": 1.608,
+      "step": 18465
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.6481319665908813,
+      "learning_rate": 1.2886728554213107e-05,
+      "loss": 1.6192,
+      "step": 18470
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.6985259056091309,
+      "learning_rate": 1.2874431042961311e-05,
+      "loss": 2.0233,
+      "step": 18475
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.6424618363380432,
+      "learning_rate": 1.2862137367015312e-05,
+      "loss": 1.8963,
+      "step": 18480
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.7724943161010742,
+      "learning_rate": 1.2849847530263598e-05,
+      "loss": 1.8944,
+      "step": 18485
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.7090172171592712,
+      "learning_rate": 1.2837561536593414e-05,
+      "loss": 1.8762,
+      "step": 18490
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.8090688586235046,
+      "learning_rate": 1.2825279389890817e-05,
+      "loss": 1.7738,
+      "step": 18495
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.6650318503379822,
+      "learning_rate": 1.2813001094040649e-05,
+      "loss": 1.6909,
+      "step": 18500
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.4980815351009369,
+      "learning_rate": 1.2800726652926503e-05,
+      "loss": 1.7088,
+      "step": 18505
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.8817710280418396,
+      "learning_rate": 1.2788456070430782e-05,
+      "loss": 1.6356,
+      "step": 18510
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 0.5760951638221741,
+      "learning_rate": 1.2776189350434659e-05,
+      "loss": 1.7639,
+      "step": 18515
+    },
+    {
+      "epoch": 3.31,
+      "grad_norm": 3.9502806663513184,
+      "learning_rate": 1.2763926496818088e-05,
+      "loss": 1.4146,
+      "step": 18520
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 1.6004568338394165,
+      "learning_rate": 1.2751667513459802e-05,
+      "loss": 2.014,
+      "step": 18525
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.7599541544914246,
+      "learning_rate": 1.2739412404237306e-05,
+      "loss": 1.9472,
+      "step": 18530
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 1.0068447589874268,
+      "learning_rate": 1.2727161173026861e-05,
+      "loss": 2.1016,
+      "step": 18535
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.5630433559417725,
+      "learning_rate": 1.2714913823703534e-05,
+      "loss": 1.8729,
+      "step": 18540
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 0.7253677845001221,
+      "learning_rate": 1.2702670360141144e-05,
+      "loss": 1.6124,
+      "step": 18545
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 1.3994909524917603,
+      "learning_rate": 1.2690430786212292e-05,
+      "loss": 1.65,
+      "step": 18550
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 1.1676279306411743,
+      "learning_rate": 1.2678195105788353e-05,
+      "loss": 1.5819,
+      "step": 18555
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 1.0580713748931885,
+      "learning_rate": 1.2665963322739439e-05,
+      "loss": 1.9325,
+      "step": 18560
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 1.0381492376327515,
+      "learning_rate": 1.2653735440934464e-05,
+      "loss": 2.1058,
+      "step": 18565
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 1.4881700277328491,
+      "learning_rate": 1.264151146424109e-05,
+      "loss": 1.8533,
+      "step": 18570
+    },
+    {
+      "epoch": 3.32,
+      "grad_norm": 1.4090508222579956,
+      "learning_rate": 1.2629291396525767e-05,
+      "loss": 1.5407,
+      "step": 18575
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 1.4334440231323242,
+      "learning_rate": 1.2617075241653665e-05,
+      "loss": 1.7129,
+      "step": 18580
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 1.431878685951233,
+      "learning_rate": 1.2604863003488752e-05,
+      "loss": 1.6889,
+      "step": 18585
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 1.138646125793457,
+      "learning_rate": 1.2592654685893757e-05,
+      "loss": 2.1111,
+      "step": 18590
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 2.590043306350708,
+      "learning_rate": 1.2580450292730142e-05,
+      "loss": 1.9615,
+      "step": 18595
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 1.139367699623108,
+      "learning_rate": 1.2568249827858153e-05,
+      "loss": 1.7285,
+      "step": 18600
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 1.8558638095855713,
+      "learning_rate": 1.2556053295136785e-05,
+      "loss": 1.6827,
+      "step": 18605
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.39854827523231506,
+      "learning_rate": 1.2543860698423792e-05,
+      "loss": 1.9213,
+      "step": 18610
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 1.0664751529693604,
+      "learning_rate": 1.2531672041575688e-05,
+      "loss": 1.8553,
+      "step": 18615
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.6911807656288147,
+      "learning_rate": 1.2519487328447715e-05,
+      "loss": 1.8099,
+      "step": 18620
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.5039658546447754,
+      "learning_rate": 1.2507306562893903e-05,
+      "loss": 1.9478,
+      "step": 18625
+    },
+    {
+      "epoch": 3.33,
+      "grad_norm": 0.6167506575584412,
+      "learning_rate": 1.2495129748766998e-05,
+      "loss": 1.7913,
+      "step": 18630
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.8103507161140442,
+      "learning_rate": 1.2482956889918529e-05,
+      "loss": 1.8247,
+      "step": 18635
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 1.3701515197753906,
+      "learning_rate": 1.2470787990198755e-05,
+      "loss": 1.7334,
+      "step": 18640
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.8373657464981079,
+      "learning_rate": 1.2458623053456697e-05,
+      "loss": 1.7793,
+      "step": 18645
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.7973329424858093,
+      "learning_rate": 1.2446462083540094e-05,
+      "loss": 1.5442,
+      "step": 18650
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.8437389731407166,
+      "learning_rate": 1.2434305084295458e-05,
+      "loss": 1.6464,
+      "step": 18655
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.8967138528823853,
+      "learning_rate": 1.242215205956804e-05,
+      "loss": 1.9497,
+      "step": 18660
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.665662407875061,
+      "learning_rate": 1.2410003013201826e-05,
+      "loss": 1.7312,
+      "step": 18665
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.7674746513366699,
+      "learning_rate": 1.2397857949039557e-05,
+      "loss": 1.9874,
+      "step": 18670
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 1.0936229228973389,
+      "learning_rate": 1.2385716870922698e-05,
+      "loss": 1.9003,
+      "step": 18675
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.903713047504425,
+      "learning_rate": 1.237357978269145e-05,
+      "loss": 1.6076,
+      "step": 18680
+    },
+    {
+      "epoch": 3.34,
+      "grad_norm": 0.5833523273468018,
+      "learning_rate": 1.2361446688184772e-05,
+      "loss": 2.0229,
+      "step": 18685
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 1.014012098312378,
+      "learning_rate": 1.234931759124035e-05,
+      "loss": 1.9358,
+      "step": 18690
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.6757218241691589,
+      "learning_rate": 1.2337192495694605e-05,
+      "loss": 1.7718,
+      "step": 18695
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 1.0655688047409058,
+      "learning_rate": 1.2325071405382693e-05,
+      "loss": 1.7798,
+      "step": 18700
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.6979892253875732,
+      "learning_rate": 1.2312954324138514e-05,
+      "loss": 1.7062,
+      "step": 18705
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.6946207880973816,
+      "learning_rate": 1.2300841255794668e-05,
+      "loss": 1.9337,
+      "step": 18710
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.827663242816925,
+      "learning_rate": 1.2288732204182529e-05,
+      "loss": 1.7177,
+      "step": 18715
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.8608758449554443,
+      "learning_rate": 1.2276627173132158e-05,
+      "loss": 1.7821,
+      "step": 18720
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 1.8402502536773682,
+      "learning_rate": 1.2264526166472373e-05,
+      "loss": 1.7818,
+      "step": 18725
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.789306104183197,
+      "learning_rate": 1.2252429188030709e-05,
+      "loss": 2.2084,
+      "step": 18730
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.7342821359634399,
+      "learning_rate": 1.2240336241633443e-05,
+      "loss": 1.8832,
+      "step": 18735
+    },
+    {
+      "epoch": 3.35,
+      "grad_norm": 0.7892515659332275,
+      "learning_rate": 1.2228247331105541e-05,
+      "loss": 1.6436,
+      "step": 18740
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 1.011290431022644,
+      "learning_rate": 1.2216162460270725e-05,
+      "loss": 1.8126,
+      "step": 18745
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.9589811563491821,
+      "learning_rate": 1.2204081632951425e-05,
+      "loss": 1.8529,
+      "step": 18750
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.4915600121021271,
+      "learning_rate": 1.2192004852968797e-05,
+      "loss": 2.1086,
+      "step": 18755
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.8448190689086914,
+      "learning_rate": 1.217993212414272e-05,
+      "loss": 1.8251,
+      "step": 18760
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.5597540140151978,
+      "learning_rate": 1.2167863450291786e-05,
+      "loss": 1.8408,
+      "step": 18765
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.7109097242355347,
+      "learning_rate": 1.2155798835233286e-05,
+      "loss": 2.3664,
+      "step": 18770
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.5528181195259094,
+      "learning_rate": 1.2143738282783263e-05,
+      "loss": 2.0561,
+      "step": 18775
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.8648270964622498,
+      "learning_rate": 1.2131681796756452e-05,
+      "loss": 1.7651,
+      "step": 18780
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 1.4095818996429443,
+      "learning_rate": 1.2119629380966313e-05,
+      "loss": 1.7121,
+      "step": 18785
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.9153962135314941,
+      "learning_rate": 1.2107581039225014e-05,
+      "loss": 1.9327,
+      "step": 18790
+    },
+    {
+      "epoch": 3.36,
+      "grad_norm": 0.4868113398551941,
+      "learning_rate": 1.2095536775343436e-05,
+      "loss": 1.9,
+      "step": 18795
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.7522435784339905,
+      "learning_rate": 1.2083496593131155e-05,
+      "loss": 1.6748,
+      "step": 18800
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.6917173266410828,
+      "learning_rate": 1.2071460496396481e-05,
+      "loss": 1.7236,
+      "step": 18805
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.5699201822280884,
+      "learning_rate": 1.2059428488946423e-05,
+      "loss": 1.7056,
+      "step": 18810
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 1.4349066019058228,
+      "learning_rate": 1.2047400574586676e-05,
+      "loss": 1.6144,
+      "step": 18815
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.8018911480903625,
+      "learning_rate": 1.203537675712167e-05,
+      "loss": 2.1065,
+      "step": 18820
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.7882276773452759,
+      "learning_rate": 1.2023357040354535e-05,
+      "loss": 1.9657,
+      "step": 18825
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 1.0160123109817505,
+      "learning_rate": 1.2011341428087077e-05,
+      "loss": 2.0429,
+      "step": 18830
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 1.0329535007476807,
+      "learning_rate": 1.1999329924119832e-05,
+      "loss": 1.856,
+      "step": 18835
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 1.044445514678955,
+      "learning_rate": 1.1987322532252024e-05,
+      "loss": 1.3516,
+      "step": 18840
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.7211053371429443,
+      "learning_rate": 1.1975319256281586e-05,
+      "loss": 1.9731,
+      "step": 18845
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 6.382941246032715,
+      "learning_rate": 1.1963320100005146e-05,
+      "loss": 1.7734,
+      "step": 18850
+    },
+    {
+      "epoch": 3.37,
+      "grad_norm": 0.6544294357299805,
+      "learning_rate": 1.195132506721802e-05,
+      "loss": 1.6961,
+      "step": 18855
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 1.0940073728561401,
+      "learning_rate": 1.1939334161714216e-05,
+      "loss": 1.8632,
+      "step": 18860
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.816892147064209,
+      "learning_rate": 1.1927347387286452e-05,
+      "loss": 1.9797,
+      "step": 18865
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.7033507227897644,
+      "learning_rate": 1.1915364747726135e-05,
+      "loss": 2.0543,
+      "step": 18870
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 1.3776378631591797,
+      "learning_rate": 1.1903386246823361e-05,
+      "loss": 1.7737,
+      "step": 18875
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 1.1460137367248535,
+      "learning_rate": 1.1891411888366933e-05,
+      "loss": 1.607,
+      "step": 18880
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.7334376573562622,
+      "learning_rate": 1.1879441676144303e-05,
+      "loss": 1.8029,
+      "step": 18885
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 1.3489125967025757,
+      "learning_rate": 1.1867475613941653e-05,
+      "loss": 1.8603,
+      "step": 18890
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 1.415173053741455,
+      "learning_rate": 1.1855513705543833e-05,
+      "loss": 1.7612,
+      "step": 18895
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 2.525381088256836,
+      "learning_rate": 1.1843555954734387e-05,
+      "loss": 1.7866,
+      "step": 18900
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 0.555586040019989,
+      "learning_rate": 1.1831602365295545e-05,
+      "loss": 1.9931,
+      "step": 18905
+    },
+    {
+      "epoch": 3.38,
+      "grad_norm": 1.590984582901001,
+      "learning_rate": 1.1819652941008197e-05,
+      "loss": 1.6122,
+      "step": 18910
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.7002542614936829,
+      "learning_rate": 1.1807707685651959e-05,
+      "loss": 1.6119,
+      "step": 18915
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.8759981989860535,
+      "learning_rate": 1.179576660300508e-05,
+      "loss": 1.8901,
+      "step": 18920
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.4278295338153839,
+      "learning_rate": 1.1783829696844522e-05,
+      "loss": 1.7944,
+      "step": 18925
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.7735504508018494,
+      "learning_rate": 1.177189697094592e-05,
+      "loss": 1.6411,
+      "step": 18930
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.8619164824485779,
+      "learning_rate": 1.175996842908358e-05,
+      "loss": 2.0436,
+      "step": 18935
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.7432487607002258,
+      "learning_rate": 1.1748044075030504e-05,
+      "loss": 1.924,
+      "step": 18940
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.8288931846618652,
+      "learning_rate": 1.1736123912558327e-05,
+      "loss": 1.9304,
+      "step": 18945
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 2.158874273300171,
+      "learning_rate": 1.172420794543741e-05,
+      "loss": 1.8055,
+      "step": 18950
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.8662157654762268,
+      "learning_rate": 1.171229617743674e-05,
+      "loss": 1.7861,
+      "step": 18955
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.9218710064888,
+      "learning_rate": 1.170038861232401e-05,
+      "loss": 2.0901,
+      "step": 18960
+    },
+    {
+      "epoch": 3.39,
+      "grad_norm": 0.7493339776992798,
+      "learning_rate": 1.1688485253865567e-05,
+      "loss": 1.9827,
+      "step": 18965
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.8299612402915955,
+      "learning_rate": 1.1676586105826442e-05,
+      "loss": 1.7931,
+      "step": 18970
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 1.1126405000686646,
+      "learning_rate": 1.166469117197031e-05,
+      "loss": 1.918,
+      "step": 18975
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.9959370493888855,
+      "learning_rate": 1.1652800456059532e-05,
+      "loss": 1.8812,
+      "step": 18980
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.6793087124824524,
+      "learning_rate": 1.1640913961855132e-05,
+      "loss": 2.0555,
+      "step": 18985
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.6176539659500122,
+      "learning_rate": 1.1629031693116796e-05,
+      "loss": 1.706,
+      "step": 18990
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.9862648248672485,
+      "learning_rate": 1.1617153653602883e-05,
+      "loss": 1.55,
+      "step": 18995
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.554785966873169,
+      "learning_rate": 1.1605279847070388e-05,
+      "loss": 2.1433,
+      "step": 19000
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.6524524688720703,
+      "learning_rate": 1.1593410277275003e-05,
+      "loss": 1.7382,
+      "step": 19005
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.8985192179679871,
+      "learning_rate": 1.158154494797104e-05,
+      "loss": 2.1055,
+      "step": 19010
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.636359453201294,
+      "learning_rate": 1.1569683862911507e-05,
+      "loss": 1.927,
+      "step": 19015
+    },
+    {
+      "epoch": 3.4,
+      "grad_norm": 0.5719195604324341,
+      "learning_rate": 1.1557827025848047e-05,
+      "loss": 1.6219,
+      "step": 19020
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.8190729022026062,
+      "learning_rate": 1.1545974440530968e-05,
+      "loss": 1.5272,
+      "step": 19025
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.7199324369430542,
+      "learning_rate": 1.1534126110709246e-05,
+      "loss": 1.9192,
+      "step": 19030
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.5589447021484375,
+      "learning_rate": 1.152228204013047e-05,
+      "loss": 1.9119,
+      "step": 19035
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.8806366324424744,
+      "learning_rate": 1.151044223254092e-05,
+      "loss": 1.6215,
+      "step": 19040
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.65036940574646,
+      "learning_rate": 1.1498606691685526e-05,
+      "loss": 1.8842,
+      "step": 19045
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.6128280162811279,
+      "learning_rate": 1.1486775421307842e-05,
+      "loss": 1.9723,
+      "step": 19050
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.8853219747543335,
+      "learning_rate": 1.1474948425150092e-05,
+      "loss": 1.8919,
+      "step": 19055
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 1.1091639995574951,
+      "learning_rate": 1.1463125706953156e-05,
+      "loss": 2.0356,
+      "step": 19060
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.6628566980361938,
+      "learning_rate": 1.145130727045653e-05,
+      "loss": 1.9267,
+      "step": 19065
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 1.4559956789016724,
+      "learning_rate": 1.1439493119398381e-05,
+      "loss": 1.7764,
+      "step": 19070
+    },
+    {
+      "epoch": 3.41,
+      "grad_norm": 0.5751055479049683,
+      "learning_rate": 1.1427683257515515e-05,
+      "loss": 1.5434,
+      "step": 19075
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.7229131460189819,
+      "learning_rate": 1.1415877688543378e-05,
+      "loss": 1.4448,
+      "step": 19080
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.6522566676139832,
+      "learning_rate": 1.1404076416216072e-05,
+      "loss": 1.9578,
+      "step": 19085
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 1.1385724544525146,
+      "learning_rate": 1.139227944426631e-05,
+      "loss": 1.794,
+      "step": 19090
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 1.248424768447876,
+      "learning_rate": 1.1380486776425477e-05,
+      "loss": 1.6419,
+      "step": 19095
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.9762007594108582,
+      "learning_rate": 1.136869841642357e-05,
+      "loss": 1.6544,
+      "step": 19100
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.4511853754520416,
+      "learning_rate": 1.135691436798924e-05,
+      "loss": 1.8262,
+      "step": 19105
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.8737003803253174,
+      "learning_rate": 1.1345134634849775e-05,
+      "loss": 1.9587,
+      "step": 19110
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.5674211382865906,
+      "learning_rate": 1.133335922073109e-05,
+      "loss": 1.9214,
+      "step": 19115
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 0.7211077213287354,
+      "learning_rate": 1.1321588129357746e-05,
+      "loss": 1.7593,
+      "step": 19120
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 1.222629427909851,
+      "learning_rate": 1.130982136445291e-05,
+      "loss": 1.971,
+      "step": 19125
+    },
+    {
+      "epoch": 3.42,
+      "grad_norm": 1.0511137247085571,
+      "learning_rate": 1.1298058929738411e-05,
+      "loss": 1.7639,
+      "step": 19130
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 1.5264891386032104,
+      "learning_rate": 1.1286300828934702e-05,
+      "loss": 1.7966,
+      "step": 19135
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 1.0392403602600098,
+      "learning_rate": 1.1274547065760838e-05,
+      "loss": 1.5846,
+      "step": 19140
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 1.030380368232727,
+      "learning_rate": 1.1262797643934537e-05,
+      "loss": 1.7171,
+      "step": 19145
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 1.143563985824585,
+      "learning_rate": 1.1251052567172135e-05,
+      "loss": 2.1491,
+      "step": 19150
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.6801722049713135,
+      "learning_rate": 1.1239311839188571e-05,
+      "loss": 1.9382,
+      "step": 19155
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 1.0771989822387695,
+      "learning_rate": 1.122757546369744e-05,
+      "loss": 1.6079,
+      "step": 19160
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 1.1643553972244263,
+      "learning_rate": 1.121584344441094e-05,
+      "loss": 1.8147,
+      "step": 19165
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 1.202305555343628,
+      "learning_rate": 1.1204115785039901e-05,
+      "loss": 1.631,
+      "step": 19170
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.6472001075744629,
+      "learning_rate": 1.1192392489293777e-05,
+      "loss": 1.9842,
+      "step": 19175
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.7353944182395935,
+      "learning_rate": 1.1180673560880619e-05,
+      "loss": 1.6733,
+      "step": 19180
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.6678537726402283,
+      "learning_rate": 1.1168959003507129e-05,
+      "loss": 1.8173,
+      "step": 19185
+    },
+    {
+      "epoch": 3.43,
+      "grad_norm": 0.6647874116897583,
+      "learning_rate": 1.1157248820878596e-05,
+      "loss": 1.7445,
+      "step": 19190
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.5945031046867371,
+      "learning_rate": 1.1145543016698946e-05,
+      "loss": 1.9277,
+      "step": 19195
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 2.9035160541534424,
+      "learning_rate": 1.1133841594670711e-05,
+      "loss": 2.0099,
+      "step": 19200
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.5868393182754517,
+      "learning_rate": 1.1122144558495043e-05,
+      "loss": 1.6714,
+      "step": 19205
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.993786633014679,
+      "learning_rate": 1.111045191187171e-05,
+      "loss": 1.7378,
+      "step": 19210
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.8114561438560486,
+      "learning_rate": 1.109876365849907e-05,
+      "loss": 2.2337,
+      "step": 19215
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.6973251700401306,
+      "learning_rate": 1.1087079802074112e-05,
+      "loss": 2.1974,
+      "step": 19220
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 1.055870771408081,
+      "learning_rate": 1.1075400346292433e-05,
+      "loss": 1.596,
+      "step": 19225
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 1.0499026775360107,
+      "learning_rate": 1.1063725294848238e-05,
+      "loss": 1.7357,
+      "step": 19230
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 1.200201392173767,
+      "learning_rate": 1.1052054651434318e-05,
+      "loss": 1.8395,
+      "step": 19235
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.6308977007865906,
+      "learning_rate": 1.1040388419742106e-05,
+      "loss": 2.024,
+      "step": 19240
+    },
+    {
+      "epoch": 3.44,
+      "grad_norm": 0.704451322555542,
+      "learning_rate": 1.1028726603461597e-05,
+      "loss": 1.8261,
+      "step": 19245
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 1.1138906478881836,
+      "learning_rate": 1.1017069206281427e-05,
+      "loss": 1.6242,
+      "step": 19250
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.6243778467178345,
+      "learning_rate": 1.1005416231888819e-05,
+      "loss": 1.9646,
+      "step": 19255
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.723716676235199,
+      "learning_rate": 1.0993767683969594e-05,
+      "loss": 1.9183,
+      "step": 19260
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.4877179265022278,
+      "learning_rate": 1.0982123566208185e-05,
+      "loss": 1.7803,
+      "step": 19265
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 1.7348027229309082,
+      "learning_rate": 1.0970483882287602e-05,
+      "loss": 1.7823,
+      "step": 19270
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 1.0215563774108887,
+      "learning_rate": 1.0958848635889482e-05,
+      "loss": 1.6437,
+      "step": 19275
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.9767959117889404,
+      "learning_rate": 1.0947217830694023e-05,
+      "loss": 1.6133,
+      "step": 19280
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 1.0302644968032837,
+      "learning_rate": 1.0935591470380046e-05,
+      "loss": 1.8479,
+      "step": 19285
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.5490931868553162,
+      "learning_rate": 1.0923969558624959e-05,
+      "loss": 1.6583,
+      "step": 19290
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 1.0158607959747314,
+      "learning_rate": 1.0912352099104762e-05,
+      "loss": 1.7078,
+      "step": 19295
+    },
+    {
+      "epoch": 3.45,
+      "grad_norm": 0.7652471661567688,
+      "learning_rate": 1.0900739095494053e-05,
+      "loss": 1.9895,
+      "step": 19300
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.7485408782958984,
+      "learning_rate": 1.0889130551465999e-05,
+      "loss": 1.7783,
+      "step": 19305
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.9016578197479248,
+      "learning_rate": 1.0877526470692378e-05,
+      "loss": 2.0336,
+      "step": 19310
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.8303315043449402,
+      "learning_rate": 1.086592685684355e-05,
+      "loss": 2.0914,
+      "step": 19315
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.6593153476715088,
+      "learning_rate": 1.0854331713588461e-05,
+      "loss": 1.8333,
+      "step": 19320
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.5653439164161682,
+      "learning_rate": 1.0842741044594656e-05,
+      "loss": 1.7719,
+      "step": 19325
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.8699783682823181,
+      "learning_rate": 1.0831154853528238e-05,
+      "loss": 2.05,
+      "step": 19330
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 2.1618435382843018,
+      "learning_rate": 1.0819573144053904e-05,
+      "loss": 1.7721,
+      "step": 19335
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.6418623924255371,
+      "learning_rate": 1.0807995919834946e-05,
+      "loss": 2.0052,
+      "step": 19340
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 0.4572185277938843,
+      "learning_rate": 1.0796423184533224e-05,
+      "loss": 2.1008,
+      "step": 19345
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 2.1507296562194824,
+      "learning_rate": 1.0784854941809192e-05,
+      "loss": 1.8341,
+      "step": 19350
+    },
+    {
+      "epoch": 3.46,
+      "grad_norm": 1.2364892959594727,
+      "learning_rate": 1.0773291195321872e-05,
+      "loss": 1.9426,
+      "step": 19355
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.9573560357093811,
+      "learning_rate": 1.0761731948728857e-05,
+      "loss": 1.9652,
+      "step": 19360
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 1.1250308752059937,
+      "learning_rate": 1.075017720568633e-05,
+      "loss": 1.8002,
+      "step": 19365
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 3.7866458892822266,
+      "learning_rate": 1.0738626969849056e-05,
+      "loss": 1.7255,
+      "step": 19370
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 1.0425763130187988,
+      "learning_rate": 1.0727081244870343e-05,
+      "loss": 1.7018,
+      "step": 19375
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.9461259245872498,
+      "learning_rate": 1.0715540034402105e-05,
+      "loss": 1.9527,
+      "step": 19380
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 1.1015312671661377,
+      "learning_rate": 1.0704003342094823e-05,
+      "loss": 1.7186,
+      "step": 19385
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.7580620646476746,
+      "learning_rate": 1.0692471171597524e-05,
+      "loss": 1.8424,
+      "step": 19390
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 1.06261146068573,
+      "learning_rate": 1.0680943526557829e-05,
+      "loss": 1.8829,
+      "step": 19395
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.9854909777641296,
+      "learning_rate": 1.0669420410621928e-05,
+      "loss": 1.7523,
+      "step": 19400
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.6900574564933777,
+      "learning_rate": 1.0657901827434563e-05,
+      "loss": 1.9449,
+      "step": 19405
+    },
+    {
+      "epoch": 3.47,
+      "grad_norm": 0.6487289667129517,
+      "learning_rate": 1.0646387780639055e-05,
+      "loss": 1.8822,
+      "step": 19410
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.5597279667854309,
+      "learning_rate": 1.0634878273877297e-05,
+      "loss": 2.0047,
+      "step": 19415
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.49827390909194946,
+      "learning_rate": 1.0623373310789722e-05,
+      "loss": 1.9747,
+      "step": 19420
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.751218855381012,
+      "learning_rate": 1.0611872895015328e-05,
+      "loss": 1.7458,
+      "step": 19425
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 1.029354214668274,
+      "learning_rate": 1.0600377030191701e-05,
+      "loss": 1.8894,
+      "step": 19430
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 1.2062971591949463,
+      "learning_rate": 1.0588885719954971e-05,
+      "loss": 1.876,
+      "step": 19435
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.9020774960517883,
+      "learning_rate": 1.0577398967939822e-05,
+      "loss": 1.855,
+      "step": 19440
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.6630207300186157,
+      "learning_rate": 1.0565916777779519e-05,
+      "loss": 2.1588,
+      "step": 19445
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.8399782180786133,
+      "learning_rate": 1.0554439153105844e-05,
+      "loss": 1.6518,
+      "step": 19450
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.8747025728225708,
+      "learning_rate": 1.054296609754917e-05,
+      "loss": 1.7769,
+      "step": 19455
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 0.9904999732971191,
+      "learning_rate": 1.0531497614738414e-05,
+      "loss": 1.6096,
+      "step": 19460
+    },
+    {
+      "epoch": 3.48,
+      "grad_norm": 1.0800423622131348,
+      "learning_rate": 1.0520033708301056e-05,
+      "loss": 1.9036,
+      "step": 19465
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 2.292233467102051,
+      "learning_rate": 1.0508574381863095e-05,
+      "loss": 1.7432,
+      "step": 19470
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 2.433584451675415,
+      "learning_rate": 1.049711963904913e-05,
+      "loss": 1.6728,
+      "step": 19475
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.8242534399032593,
+      "learning_rate": 1.0485669483482266e-05,
+      "loss": 1.6268,
+      "step": 19480
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.3501177728176117,
+      "learning_rate": 1.0474223918784178e-05,
+      "loss": 1.8424,
+      "step": 19485
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.7704662084579468,
+      "learning_rate": 1.0462782948575098e-05,
+      "loss": 1.5104,
+      "step": 19490
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 1.05587637424469,
+      "learning_rate": 1.0451346576473787e-05,
+      "loss": 2.0312,
+      "step": 19495
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.6842361688613892,
+      "learning_rate": 1.0439914806097559e-05,
+      "loss": 1.8268,
+      "step": 19500
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 1.027362585067749,
+      "learning_rate": 1.042848764106228e-05,
+      "loss": 2.1638,
+      "step": 19505
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 1.474015235900879,
+      "learning_rate": 1.0417065084982346e-05,
+      "loss": 1.6934,
+      "step": 19510
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.5891749858856201,
+      "learning_rate": 1.040564714147069e-05,
+      "loss": 1.8321,
+      "step": 19515
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 0.8095927238464355,
+      "learning_rate": 1.0394233814138804e-05,
+      "loss": 1.7532,
+      "step": 19520
+    },
+    {
+      "epoch": 3.49,
+      "grad_norm": 1.7392970323562622,
+      "learning_rate": 1.0382825106596713e-05,
+      "loss": 2.0015,
+      "step": 19525
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 1.0401033163070679,
+      "learning_rate": 1.0371421022452981e-05,
+      "loss": 2.0158,
+      "step": 19530
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 1.081748127937317,
+      "learning_rate": 1.036002156531472e-05,
+      "loss": 2.0205,
+      "step": 19535
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.47808757424354553,
+      "learning_rate": 1.0348626738787548e-05,
+      "loss": 1.4998,
+      "step": 19540
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.770613968372345,
+      "learning_rate": 1.0337236546475642e-05,
+      "loss": 1.8638,
+      "step": 19545
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.6831596493721008,
+      "learning_rate": 1.0325850991981715e-05,
+      "loss": 1.833,
+      "step": 19550
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.8213655352592468,
+      "learning_rate": 1.0314470078907013e-05,
+      "loss": 1.8172,
+      "step": 19555
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.9885990619659424,
+      "learning_rate": 1.0303093810851288e-05,
+      "loss": 1.7939,
+      "step": 19560
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 1.137285590171814,
+      "learning_rate": 1.0291722191412865e-05,
+      "loss": 1.7968,
+      "step": 19565
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 1.3749232292175293,
+      "learning_rate": 1.0280355224188556e-05,
+      "loss": 1.8718,
+      "step": 19570
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.9204155802726746,
+      "learning_rate": 1.0268992912773732e-05,
+      "loss": 1.6898,
+      "step": 19575
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.7502628564834595,
+      "learning_rate": 1.025763526076228e-05,
+      "loss": 1.669,
+      "step": 19580
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 1.0295674800872803,
+      "learning_rate": 1.0246282271746612e-05,
+      "loss": 1.7433,
+      "step": 19585
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.7040165662765503,
+      "learning_rate": 1.0234933949317683e-05,
+      "loss": 1.6466,
+      "step": 19590
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.6841382384300232,
+      "learning_rate": 1.0223590297064933e-05,
+      "loss": 2.0555,
+      "step": 19595
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.6829349994659424,
+      "learning_rate": 1.0212251318576357e-05,
+      "loss": 1.7514,
+      "step": 19600
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 1.0354279279708862,
+      "learning_rate": 1.0200917017438478e-05,
+      "loss": 1.7173,
+      "step": 19605
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.7070810794830322,
+      "learning_rate": 1.01895873972363e-05,
+      "loss": 2.1088,
+      "step": 19610
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 1.9190095663070679,
+      "learning_rate": 1.017826246155338e-05,
+      "loss": 2.2198,
+      "step": 19615
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.7562488913536072,
+      "learning_rate": 1.0166942213971792e-05,
+      "loss": 1.7641,
+      "step": 19620
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 0.9604626893997192,
+      "learning_rate": 1.015562665807212e-05,
+      "loss": 1.9173,
+      "step": 19625
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 2.0906529426574707,
+      "learning_rate": 1.0144315797433449e-05,
+      "loss": 1.9204,
+      "step": 19630
+    },
+    {
+      "epoch": 3.51,
+      "grad_norm": 1.8429858684539795,
+      "learning_rate": 1.0133009635633402e-05,
+      "loss": 1.7915,
+      "step": 19635
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 1.0076621770858765,
+      "learning_rate": 1.0121708176248107e-05,
+      "loss": 1.6617,
+      "step": 19640
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.9586327075958252,
+      "learning_rate": 1.0110411422852206e-05,
+      "loss": 1.8037,
+      "step": 19645
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.7485795617103577,
+      "learning_rate": 1.0099119379018856e-05,
+      "loss": 2.0374,
+      "step": 19650
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.6466397643089294,
+      "learning_rate": 1.0087832048319716e-05,
+      "loss": 1.795,
+      "step": 19655
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 1.1926106214523315,
+      "learning_rate": 1.0076549434324944e-05,
+      "loss": 1.8342,
+      "step": 19660
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.621414840221405,
+      "learning_rate": 1.0065271540603236e-05,
+      "loss": 1.7218,
+      "step": 19665
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 1.1308784484863281,
+      "learning_rate": 1.005399837072177e-05,
+      "loss": 1.7901,
+      "step": 19670
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.8960298299789429,
+      "learning_rate": 1.0042729928246245e-05,
+      "loss": 1.8848,
+      "step": 19675
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.4410213232040405,
+      "learning_rate": 1.0031466216740865e-05,
+      "loss": 1.7415,
+      "step": 19680
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.7294899821281433,
+      "learning_rate": 1.0020207239768314e-05,
+      "loss": 2.1899,
+      "step": 19685
+    },
+    {
+      "epoch": 3.52,
+      "grad_norm": 0.6985868215560913,
+      "learning_rate": 1.0008953000889803e-05,
+      "loss": 1.848,
+      "step": 19690
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.5390515327453613,
+      "learning_rate": 9.997703503665051e-06,
+      "loss": 2.0746,
+      "step": 19695
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.9798139929771423,
+      "learning_rate": 9.986458751652238e-06,
+      "loss": 1.8011,
+      "step": 19700
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.7954207062721252,
+      "learning_rate": 9.975218748408083e-06,
+      "loss": 1.9957,
+      "step": 19705
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 1.3126145601272583,
+      "learning_rate": 9.963983497487783e-06,
+      "loss": 1.8887,
+      "step": 19710
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.8574180006980896,
+      "learning_rate": 9.95275300244505e-06,
+      "loss": 1.6767,
+      "step": 19715
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.6159719228744507,
+      "learning_rate": 9.941527266832062e-06,
+      "loss": 2.1028,
+      "step": 19720
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.8486490249633789,
+      "learning_rate": 9.930306294199512e-06,
+      "loss": 1.8083,
+      "step": 19725
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 2.6029884815216064,
+      "learning_rate": 9.919090088096589e-06,
+      "loss": 1.7197,
+      "step": 19730
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.9481103420257568,
+      "learning_rate": 9.907878652070962e-06,
+      "loss": 1.7045,
+      "step": 19735
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.3375113606452942,
+      "learning_rate": 9.896671989668813e-06,
+      "loss": 1.8396,
+      "step": 19740
+    },
+    {
+      "epoch": 3.53,
+      "grad_norm": 0.7272607684135437,
+      "learning_rate": 9.885470104434783e-06,
+      "loss": 1.7523,
+      "step": 19745
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 1.5993422269821167,
+      "learning_rate": 9.874272999912013e-06,
+      "loss": 2.0852,
+      "step": 19750
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.6138380169868469,
+      "learning_rate": 9.863080679642143e-06,
+      "loss": 1.8729,
+      "step": 19755
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.8599143028259277,
+      "learning_rate": 9.851893147165295e-06,
+      "loss": 2.0843,
+      "step": 19760
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.7874997854232788,
+      "learning_rate": 9.840710406020072e-06,
+      "loss": 1.6064,
+      "step": 19765
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.9153751730918884,
+      "learning_rate": 9.829532459743579e-06,
+      "loss": 1.6868,
+      "step": 19770
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.6881905794143677,
+      "learning_rate": 9.818359311871367e-06,
+      "loss": 1.6391,
+      "step": 19775
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 1.0899308919906616,
+      "learning_rate": 9.807190965937499e-06,
+      "loss": 1.9639,
+      "step": 19780
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 1.196337342262268,
+      "learning_rate": 9.796027425474515e-06,
+      "loss": 1.7011,
+      "step": 19785
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 1.8400663137435913,
+      "learning_rate": 9.784868694013444e-06,
+      "loss": 1.9638,
+      "step": 19790
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.783830463886261,
+      "learning_rate": 9.773714775083759e-06,
+      "loss": 1.7963,
+      "step": 19795
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.5852648615837097,
+      "learning_rate": 9.762565672213444e-06,
+      "loss": 1.6803,
+      "step": 19800
+    },
+    {
+      "epoch": 3.54,
+      "grad_norm": 0.6797320246696472,
+      "learning_rate": 9.751421388928961e-06,
+      "loss": 1.9709,
+      "step": 19805
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.8105748295783997,
+      "learning_rate": 9.740281928755218e-06,
+      "loss": 2.0418,
+      "step": 19810
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.5812113285064697,
+      "learning_rate": 9.72914729521562e-06,
+      "loss": 1.8629,
+      "step": 19815
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.9974383115768433,
+      "learning_rate": 9.718017491832046e-06,
+      "loss": 2.231,
+      "step": 19820
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 1.628359079360962,
+      "learning_rate": 9.706892522124839e-06,
+      "loss": 1.7025,
+      "step": 19825
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 2.6423473358154297,
+      "learning_rate": 9.695772389612826e-06,
+      "loss": 1.7726,
+      "step": 19830
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.6964113712310791,
+      "learning_rate": 9.684657097813282e-06,
+      "loss": 1.5745,
+      "step": 19835
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 1.0182359218597412,
+      "learning_rate": 9.673546650241963e-06,
+      "loss": 2.1207,
+      "step": 19840
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.8013817071914673,
+      "learning_rate": 9.662441050413093e-06,
+      "loss": 1.7262,
+      "step": 19845
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.9642962217330933,
+      "learning_rate": 9.651340301839367e-06,
+      "loss": 1.9043,
+      "step": 19850
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.8014324307441711,
+      "learning_rate": 9.64024440803194e-06,
+      "loss": 1.8698,
+      "step": 19855
+    },
+    {
+      "epoch": 3.55,
+      "grad_norm": 0.7242090106010437,
+      "learning_rate": 9.62915337250044e-06,
+      "loss": 1.6975,
+      "step": 19860
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.8667596578598022,
+      "learning_rate": 9.618067198752937e-06,
+      "loss": 1.7597,
+      "step": 19865
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.5412402749061584,
+      "learning_rate": 9.606985890295984e-06,
+      "loss": 1.7658,
+      "step": 19870
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 1.8349127769470215,
+      "learning_rate": 9.595909450634586e-06,
+      "loss": 1.7718,
+      "step": 19875
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 1.097348928451538,
+      "learning_rate": 9.58483788327221e-06,
+      "loss": 2.0921,
+      "step": 19880
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.8967131972312927,
+      "learning_rate": 9.573771191710795e-06,
+      "loss": 1.8385,
+      "step": 19885
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 7.748429298400879,
+      "learning_rate": 9.562709379450705e-06,
+      "loss": 1.4055,
+      "step": 19890
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.7774618268013,
+      "learning_rate": 9.551652449990797e-06,
+      "loss": 2.0261,
+      "step": 19895
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.6186455488204956,
+      "learning_rate": 9.540600406828354e-06,
+      "loss": 1.8112,
+      "step": 19900
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.5887669324874878,
+      "learning_rate": 9.529553253459132e-06,
+      "loss": 1.9148,
+      "step": 19905
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.9519124627113342,
+      "learning_rate": 9.518510993377331e-06,
+      "loss": 1.7988,
+      "step": 19910
+    },
+    {
+      "epoch": 3.56,
+      "grad_norm": 0.5737485885620117,
+      "learning_rate": 9.507473630075613e-06,
+      "loss": 1.8965,
+      "step": 19915
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 1.0009839534759521,
+      "learning_rate": 9.496441167045095e-06,
+      "loss": 1.9366,
+      "step": 19920
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.9404077529907227,
+      "learning_rate": 9.485413607775314e-06,
+      "loss": 1.7763,
+      "step": 19925
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.8102321624755859,
+      "learning_rate": 9.474390955754293e-06,
+      "loss": 1.6678,
+      "step": 19930
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 1.2429808378219604,
+      "learning_rate": 9.46337321446847e-06,
+      "loss": 1.8401,
+      "step": 19935
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.66902095079422,
+      "learning_rate": 9.452360387402756e-06,
+      "loss": 1.6209,
+      "step": 19940
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.5885403752326965,
+      "learning_rate": 9.441352478040497e-06,
+      "loss": 2.0758,
+      "step": 19945
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.7154830694198608,
+      "learning_rate": 9.430349489863493e-06,
+      "loss": 1.9128,
+      "step": 19950
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.5632511973381042,
+      "learning_rate": 9.419351426351958e-06,
+      "loss": 2.2468,
+      "step": 19955
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 2.063032388687134,
+      "learning_rate": 9.40835829098458e-06,
+      "loss": 1.6646,
+      "step": 19960
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.9876983165740967,
+      "learning_rate": 9.397370087238483e-06,
+      "loss": 1.8661,
+      "step": 19965
+    },
+    {
+      "epoch": 3.57,
+      "grad_norm": 0.797542154788971,
+      "learning_rate": 9.386386818589212e-06,
+      "loss": 2.0163,
+      "step": 19970
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.8260084390640259,
+      "learning_rate": 9.375408488510787e-06,
+      "loss": 1.7415,
+      "step": 19975
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.849568247795105,
+      "learning_rate": 9.364435100475625e-06,
+      "loss": 1.648,
+      "step": 19980
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.8121174573898315,
+      "learning_rate": 9.35346665795459e-06,
+      "loss": 2.2069,
+      "step": 19985
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.9171432256698608,
+      "learning_rate": 9.342503164417005e-06,
+      "loss": 2.1348,
+      "step": 19990
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.9130405783653259,
+      "learning_rate": 9.331544623330606e-06,
+      "loss": 2.0652,
+      "step": 19995
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.814951479434967,
+      "learning_rate": 9.320591038161574e-06,
+      "loss": 1.6998,
+      "step": 20000
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 1.7941076755523682,
+      "learning_rate": 9.309642412374514e-06,
+      "loss": 1.7472,
+      "step": 20005
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.9372169971466064,
+      "learning_rate": 9.298698749432474e-06,
+      "loss": 1.759,
+      "step": 20010
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.6178033351898193,
+      "learning_rate": 9.287760052796909e-06,
+      "loss": 1.8053,
+      "step": 20015
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.8956480622291565,
+      "learning_rate": 9.27682632592773e-06,
+      "loss": 1.8945,
+      "step": 20020
+    },
+    {
+      "epoch": 3.58,
+      "grad_norm": 0.7049922347068787,
+      "learning_rate": 9.265897572283263e-06,
+      "loss": 1.9184,
+      "step": 20025
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.6929253935813904,
+      "learning_rate": 9.254973795320254e-06,
+      "loss": 1.9086,
+      "step": 20030
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.7709065675735474,
+      "learning_rate": 9.244054998493886e-06,
+      "loss": 1.7822,
+      "step": 20035
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.817094624042511,
+      "learning_rate": 9.23314118525778e-06,
+      "loss": 1.7846,
+      "step": 20040
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.9673941135406494,
+      "learning_rate": 9.22223235906394e-06,
+      "loss": 1.8459,
+      "step": 20045
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.639333188533783,
+      "learning_rate": 9.211328523362828e-06,
+      "loss": 1.8805,
+      "step": 20050
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 1.8863416910171509,
+      "learning_rate": 9.200429681603318e-06,
+      "loss": 1.7241,
+      "step": 20055
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 1.824625849723816,
+      "learning_rate": 9.189535837232701e-06,
+      "loss": 1.8988,
+      "step": 20060
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.7470659017562866,
+      "learning_rate": 9.178646993696702e-06,
+      "loss": 1.3453,
+      "step": 20065
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 1.1714149713516235,
+      "learning_rate": 9.167763154439437e-06,
+      "loss": 2.1238,
+      "step": 20070
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 0.8737668991088867,
+      "learning_rate": 9.156884322903452e-06,
+      "loss": 1.9031,
+      "step": 20075
+    },
+    {
+      "epoch": 3.59,
+      "grad_norm": 1.0085196495056152,
+      "learning_rate": 9.146010502529717e-06,
+      "loss": 1.58,
+      "step": 20080
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.6037978529930115,
+      "learning_rate": 9.13514169675761e-06,
+      "loss": 1.8138,
+      "step": 20085
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 1.0545234680175781,
+      "learning_rate": 9.124277909024923e-06,
+      "loss": 1.7113,
+      "step": 20090
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.7123447060585022,
+      "learning_rate": 9.113419142767864e-06,
+      "loss": 1.7098,
+      "step": 20095
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.9203841090202332,
+      "learning_rate": 9.102565401421056e-06,
+      "loss": 1.6357,
+      "step": 20100
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.8946130871772766,
+      "learning_rate": 9.091716688417515e-06,
+      "loss": 2.1125,
+      "step": 20105
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.6853888630867004,
+      "learning_rate": 9.080873007188681e-06,
+      "loss": 1.9502,
+      "step": 20110
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.7109887003898621,
+      "learning_rate": 9.070034361164412e-06,
+      "loss": 1.7719,
+      "step": 20115
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 1.7917437553405762,
+      "learning_rate": 9.059200753772942e-06,
+      "loss": 1.8331,
+      "step": 20120
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.8030440211296082,
+      "learning_rate": 9.048372188440942e-06,
+      "loss": 1.8035,
+      "step": 20125
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.9415981769561768,
+      "learning_rate": 9.037548668593482e-06,
+      "loss": 2.0076,
+      "step": 20130
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 1.300551176071167,
+      "learning_rate": 9.026730197654018e-06,
+      "loss": 1.7933,
+      "step": 20135
+    },
+    {
+      "epoch": 3.6,
+      "grad_norm": 0.8916707038879395,
+      "learning_rate": 9.015916779044428e-06,
+      "loss": 1.9135,
+      "step": 20140
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.4446997046470642,
+      "learning_rate": 9.005108416184985e-06,
+      "loss": 1.8592,
+      "step": 20145
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.7679259777069092,
+      "learning_rate": 8.994305112494364e-06,
+      "loss": 1.8635,
+      "step": 20150
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 1.084495186805725,
+      "learning_rate": 8.983506871389648e-06,
+      "loss": 1.6656,
+      "step": 20155
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.5544742941856384,
+      "learning_rate": 8.972713696286297e-06,
+      "loss": 2.0044,
+      "step": 20160
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 1.2430366277694702,
+      "learning_rate": 8.961925590598196e-06,
+      "loss": 1.9011,
+      "step": 20165
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.7293367385864258,
+      "learning_rate": 8.951142557737597e-06,
+      "loss": 1.8652,
+      "step": 20170
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.6293432116508484,
+      "learning_rate": 8.940364601115167e-06,
+      "loss": 2.0155,
+      "step": 20175
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.6995351910591125,
+      "learning_rate": 8.929591724139966e-06,
+      "loss": 1.5788,
+      "step": 20180
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 1.0130141973495483,
+      "learning_rate": 8.918823930219455e-06,
+      "loss": 2.1183,
+      "step": 20185
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 0.8194257020950317,
+      "learning_rate": 8.908061222759457e-06,
+      "loss": 1.8322,
+      "step": 20190
+    },
+    {
+      "epoch": 3.61,
+      "grad_norm": 3.565441608428955,
+      "learning_rate": 8.899454721330733e-06,
+      "loss": 1.7912,
+      "step": 20195
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.643989622592926,
+      "learning_rate": 8.888701178077265e-06,
+      "loss": 1.8781,
+      "step": 20200
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.9238978624343872,
+      "learning_rate": 8.877952730812116e-06,
+      "loss": 1.8105,
+      "step": 20205
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.7382200360298157,
+      "learning_rate": 8.867209382935022e-06,
+      "loss": 1.6843,
+      "step": 20210
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 1.2155033349990845,
+      "learning_rate": 8.856471137844071e-06,
+      "loss": 1.6075,
+      "step": 20215
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 12.773721694946289,
+      "learning_rate": 8.845737998935779e-06,
+      "loss": 1.8195,
+      "step": 20220
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 1.3359897136688232,
+      "learning_rate": 8.835009969605012e-06,
+      "loss": 2.098,
+      "step": 20225
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.8200415968894958,
+      "learning_rate": 8.824287053245046e-06,
+      "loss": 1.639,
+      "step": 20230
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 1.8653521537780762,
+      "learning_rate": 8.813569253247522e-06,
+      "loss": 1.714,
+      "step": 20235
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.7212638258934021,
+      "learning_rate": 8.802856573002474e-06,
+      "loss": 1.5678,
+      "step": 20240
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.6007484793663025,
+      "learning_rate": 8.792149015898321e-06,
+      "loss": 2.021,
+      "step": 20245
+    },
+    {
+      "epoch": 3.62,
+      "grad_norm": 0.8171059489250183,
+      "learning_rate": 8.781446585321836e-06,
+      "loss": 1.644,
+      "step": 20250
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.5657952427864075,
+      "learning_rate": 8.770749284658198e-06,
+      "loss": 1.7338,
+      "step": 20255
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.8464135527610779,
+      "learning_rate": 8.760057117290956e-06,
+      "loss": 1.8144,
+      "step": 20260
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.8611558079719543,
+      "learning_rate": 8.749370086602022e-06,
+      "loss": 1.9741,
+      "step": 20265
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.8753948211669922,
+      "learning_rate": 8.738688195971698e-06,
+      "loss": 2.034,
+      "step": 20270
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.4790874421596527,
+      "learning_rate": 8.728011448778652e-06,
+      "loss": 1.8537,
+      "step": 20275
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.5300187468528748,
+      "learning_rate": 8.717339848399947e-06,
+      "loss": 1.6698,
+      "step": 20280
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 1.3581863641738892,
+      "learning_rate": 8.706673398210973e-06,
+      "loss": 1.9417,
+      "step": 20285
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 0.47221723198890686,
+      "learning_rate": 8.696012101585533e-06,
+      "loss": 2.0264,
+      "step": 20290
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 1.0025262832641602,
+      "learning_rate": 8.685355961895784e-06,
+      "loss": 1.7893,
+      "step": 20295
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 1.1813714504241943,
+      "learning_rate": 8.67470498251225e-06,
+      "loss": 1.8123,
+      "step": 20300
+    },
+    {
+      "epoch": 3.63,
+      "grad_norm": 2.1088335514068604,
+      "learning_rate": 8.664059166803834e-06,
+      "loss": 1.8788,
+      "step": 20305
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.9035912752151489,
+      "learning_rate": 8.653418518137792e-06,
+      "loss": 2.0769,
+      "step": 20310
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.8217169046401978,
+      "learning_rate": 8.642783039879742e-06,
+      "loss": 1.7871,
+      "step": 20315
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 1.7109413146972656,
+      "learning_rate": 8.632152735393682e-06,
+      "loss": 1.7824,
+      "step": 20320
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 1.142771601676941,
+      "learning_rate": 8.621527608041969e-06,
+      "loss": 1.7042,
+      "step": 20325
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.5646505355834961,
+      "learning_rate": 8.610907661185316e-06,
+      "loss": 1.8401,
+      "step": 20330
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.9124100208282471,
+      "learning_rate": 8.600292898182815e-06,
+      "loss": 2.1127,
+      "step": 20335
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.7633247971534729,
+      "learning_rate": 8.589683322391889e-06,
+      "loss": 1.9485,
+      "step": 20340
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.6282685399055481,
+      "learning_rate": 8.579078937168344e-06,
+      "loss": 1.8386,
+      "step": 20345
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.593157172203064,
+      "learning_rate": 8.568479745866344e-06,
+      "loss": 1.806,
+      "step": 20350
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 0.9064671397209167,
+      "learning_rate": 8.557885751838384e-06,
+      "loss": 1.8968,
+      "step": 20355
+    },
+    {
+      "epoch": 3.64,
+      "grad_norm": 1.2595146894454956,
+      "learning_rate": 8.54729695843535e-06,
+      "loss": 1.8831,
+      "step": 20360
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.8819836378097534,
+      "learning_rate": 8.53671336900647e-06,
+      "loss": 1.7677,
+      "step": 20365
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.46707066893577576,
+      "learning_rate": 8.526134986899304e-06,
+      "loss": 1.7894,
+      "step": 20370
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 3.5591564178466797,
+      "learning_rate": 8.5155618154598e-06,
+      "loss": 1.6036,
+      "step": 20375
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.959492564201355,
+      "learning_rate": 8.504993858032234e-06,
+      "loss": 1.9008,
+      "step": 20380
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.6960819959640503,
+      "learning_rate": 8.494431117959247e-06,
+      "loss": 1.6686,
+      "step": 20385
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 1.4117478132247925,
+      "learning_rate": 8.48387359858182e-06,
+      "loss": 1.6402,
+      "step": 20390
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 1.1412084102630615,
+      "learning_rate": 8.473321303239296e-06,
+      "loss": 1.7246,
+      "step": 20395
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.6889460682868958,
+      "learning_rate": 8.462774235269348e-06,
+      "loss": 1.9935,
+      "step": 20400
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 3.105839490890503,
+      "learning_rate": 8.452232398007993e-06,
+      "loss": 1.9629,
+      "step": 20405
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 1.2569962739944458,
+      "learning_rate": 8.44169579478961e-06,
+      "loss": 1.8072,
+      "step": 20410
+    },
+    {
+      "epoch": 3.65,
+      "grad_norm": 0.9013957977294922,
+      "learning_rate": 8.431164428946927e-06,
+      "loss": 2.0294,
+      "step": 20415
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 1.8932322263717651,
+      "learning_rate": 8.420638303810993e-06,
+      "loss": 1.7084,
+      "step": 20420
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.6387518644332886,
+      "learning_rate": 8.410117422711227e-06,
+      "loss": 1.9067,
+      "step": 20425
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.5416659712791443,
+      "learning_rate": 8.399601788975354e-06,
+      "loss": 1.9721,
+      "step": 20430
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 1.3782933950424194,
+      "learning_rate": 8.389091405929467e-06,
+      "loss": 1.9061,
+      "step": 20435
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 2.072787046432495,
+      "learning_rate": 8.37858627689799e-06,
+      "loss": 1.9161,
+      "step": 20440
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.5644092559814453,
+      "learning_rate": 8.368086405203696e-06,
+      "loss": 1.7495,
+      "step": 20445
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 1.1922764778137207,
+      "learning_rate": 8.357591794167666e-06,
+      "loss": 1.8278,
+      "step": 20450
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.7478966116905212,
+      "learning_rate": 8.34710244710935e-06,
+      "loss": 1.8286,
+      "step": 20455
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.6147327423095703,
+      "learning_rate": 8.336618367346505e-06,
+      "loss": 1.615,
+      "step": 20460
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 1.372970700263977,
+      "learning_rate": 8.326139558195242e-06,
+      "loss": 1.9454,
+      "step": 20465
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.6343486905097961,
+      "learning_rate": 8.315666022969998e-06,
+      "loss": 1.8179,
+      "step": 20470
+    },
+    {
+      "epoch": 3.66,
+      "grad_norm": 0.6342081427574158,
+      "learning_rate": 8.30519776498354e-06,
+      "loss": 1.7307,
+      "step": 20475
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 1.3767012357711792,
+      "learning_rate": 8.294734787546975e-06,
+      "loss": 1.8573,
+      "step": 20480
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.8521339297294617,
+      "learning_rate": 8.284277093969737e-06,
+      "loss": 1.9258,
+      "step": 20485
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.3690893054008484,
+      "learning_rate": 8.273824687559573e-06,
+      "loss": 1.9283,
+      "step": 20490
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.61278235912323,
+      "learning_rate": 8.263377571622563e-06,
+      "loss": 1.6369,
+      "step": 20495
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.7713987827301025,
+      "learning_rate": 8.252935749463132e-06,
+      "loss": 1.9637,
+      "step": 20500
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.4912928640842438,
+      "learning_rate": 8.242499224384012e-06,
+      "loss": 1.9876,
+      "step": 20505
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 1.3459941148757935,
+      "learning_rate": 8.232067999686269e-06,
+      "loss": 1.5705,
+      "step": 20510
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.9344834089279175,
+      "learning_rate": 8.221642078669296e-06,
+      "loss": 1.7796,
+      "step": 20515
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 1.331256628036499,
+      "learning_rate": 8.211221464630788e-06,
+      "loss": 1.85,
+      "step": 20520
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.6549345850944519,
+      "learning_rate": 8.200806160866784e-06,
+      "loss": 1.7886,
+      "step": 20525
+    },
+    {
+      "epoch": 3.67,
+      "grad_norm": 0.852634847164154,
+      "learning_rate": 8.19039617067163e-06,
+      "loss": 1.6675,
+      "step": 20530
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.6992741227149963,
+      "learning_rate": 8.179991497337997e-06,
+      "loss": 1.9973,
+      "step": 20535
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.8704904317855835,
+      "learning_rate": 8.169592144156885e-06,
+      "loss": 1.7932,
+      "step": 20540
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 4.262266635894775,
+      "learning_rate": 8.159198114417591e-06,
+      "loss": 1.679,
+      "step": 20545
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.6595445275306702,
+      "learning_rate": 8.148809411407723e-06,
+      "loss": 1.8552,
+      "step": 20550
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.7737083435058594,
+      "learning_rate": 8.13842603841323e-06,
+      "loss": 1.8383,
+      "step": 20555
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.8114953637123108,
+      "learning_rate": 8.128047998718363e-06,
+      "loss": 1.8305,
+      "step": 20560
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.9654433131217957,
+      "learning_rate": 8.117675295605685e-06,
+      "loss": 1.7202,
+      "step": 20565
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 1.0531843900680542,
+      "learning_rate": 8.10730793235608e-06,
+      "loss": 1.437,
+      "step": 20570
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.9250448942184448,
+      "learning_rate": 8.096945912248721e-06,
+      "loss": 1.4706,
+      "step": 20575
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.9229045510292053,
+      "learning_rate": 8.08658923856111e-06,
+      "loss": 1.8187,
+      "step": 20580
+    },
+    {
+      "epoch": 3.68,
+      "grad_norm": 0.9853529930114746,
+      "learning_rate": 8.076237914569063e-06,
+      "loss": 1.7397,
+      "step": 20585
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.8005128502845764,
+      "learning_rate": 8.065891943546675e-06,
+      "loss": 1.9982,
+      "step": 20590
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.5977509021759033,
+      "learning_rate": 8.055551328766378e-06,
+      "loss": 1.8997,
+      "step": 20595
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 1.766401767730713,
+      "learning_rate": 8.045216073498896e-06,
+      "loss": 1.6968,
+      "step": 20600
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 1.244626760482788,
+      "learning_rate": 8.03488618101327e-06,
+      "loss": 1.7645,
+      "step": 20605
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.5766518115997314,
+      "learning_rate": 8.02456165457682e-06,
+      "loss": 1.8566,
+      "step": 20610
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.8886863589286804,
+      "learning_rate": 8.014242497455191e-06,
+      "loss": 1.7796,
+      "step": 20615
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.8238760232925415,
+      "learning_rate": 8.00392871291232e-06,
+      "loss": 2.0147,
+      "step": 20620
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.5203852653503418,
+      "learning_rate": 7.993620304210447e-06,
+      "loss": 2.0413,
+      "step": 20625
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.9186719059944153,
+      "learning_rate": 7.983317274610125e-06,
+      "loss": 1.8603,
+      "step": 20630
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 1.096434473991394,
+      "learning_rate": 7.973019627370182e-06,
+      "loss": 1.6407,
+      "step": 20635
+    },
+    {
+      "epoch": 3.69,
+      "grad_norm": 0.8249112963676453,
+      "learning_rate": 7.96272736574774e-06,
+      "loss": 1.7465,
+      "step": 20640
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 1.604615330696106,
+      "learning_rate": 7.95244049299825e-06,
+      "loss": 1.7763,
+      "step": 20645
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.730076789855957,
+      "learning_rate": 7.942159012375434e-06,
+      "loss": 1.9924,
+      "step": 20650
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 1.068292260169983,
+      "learning_rate": 7.93188292713131e-06,
+      "loss": 2.1643,
+      "step": 20655
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 1.0101828575134277,
+      "learning_rate": 7.92161224051621e-06,
+      "loss": 1.4485,
+      "step": 20660
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.4581923186779022,
+      "learning_rate": 7.911346955778722e-06,
+      "loss": 1.9251,
+      "step": 20665
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.5387589335441589,
+      "learning_rate": 7.901087076165756e-06,
+      "loss": 1.9034,
+      "step": 20670
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.7266547679901123,
+      "learning_rate": 7.8908326049225e-06,
+      "loss": 1.8593,
+      "step": 20675
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 2.4284040927886963,
+      "learning_rate": 7.880583545292441e-06,
+      "loss": 1.5051,
+      "step": 20680
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.9465983510017395,
+      "learning_rate": 7.870339900517335e-06,
+      "loss": 1.8325,
+      "step": 20685
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.9285613894462585,
+      "learning_rate": 7.86010167383724e-06,
+      "loss": 1.8667,
+      "step": 20690
+    },
+    {
+      "epoch": 3.7,
+      "grad_norm": 0.7407609820365906,
+      "learning_rate": 7.849868868490506e-06,
+      "loss": 1.8178,
+      "step": 20695
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 5.604633331298828,
+      "learning_rate": 7.839641487713745e-06,
+      "loss": 1.994,
+      "step": 20700
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.8995658159255981,
+      "learning_rate": 7.829419534741875e-06,
+      "loss": 1.9699,
+      "step": 20705
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 1.138853907585144,
+      "learning_rate": 7.819203012808091e-06,
+      "loss": 1.9048,
+      "step": 20710
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.8742569088935852,
+      "learning_rate": 7.808991925143869e-06,
+      "loss": 2.0381,
+      "step": 20715
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.40514862537384033,
+      "learning_rate": 7.798786274978969e-06,
+      "loss": 2.1029,
+      "step": 20720
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 1.0639851093292236,
+      "learning_rate": 7.788586065541431e-06,
+      "loss": 1.6771,
+      "step": 20725
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 2.2875404357910156,
+      "learning_rate": 7.778391300057553e-06,
+      "loss": 1.8035,
+      "step": 20730
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.6936883330345154,
+      "learning_rate": 7.768201981751944e-06,
+      "loss": 1.8963,
+      "step": 20735
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.7434184551239014,
+      "learning_rate": 7.758018113847473e-06,
+      "loss": 1.6205,
+      "step": 20740
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 1.422929048538208,
+      "learning_rate": 7.747839699565288e-06,
+      "loss": 2.0673,
+      "step": 20745
+    },
+    {
+      "epoch": 3.71,
+      "grad_norm": 0.9202067852020264,
+      "learning_rate": 7.737666742124816e-06,
+      "loss": 1.6292,
+      "step": 20750
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.48007720708847046,
+      "learning_rate": 7.727499244743747e-06,
+      "loss": 1.8354,
+      "step": 20755
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 1.398300051689148,
+      "learning_rate": 7.717337210638049e-06,
+      "loss": 1.7383,
+      "step": 20760
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 1.650846004486084,
+      "learning_rate": 7.707180643021969e-06,
+      "loss": 1.8777,
+      "step": 20765
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 3.9538509845733643,
+      "learning_rate": 7.697029545108025e-06,
+      "loss": 1.6718,
+      "step": 20770
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.817905843257904,
+      "learning_rate": 7.686883920106986e-06,
+      "loss": 1.6816,
+      "step": 20775
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.4926592707633972,
+      "learning_rate": 7.676743771227906e-06,
+      "loss": 1.7673,
+      "step": 20780
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.5661248564720154,
+      "learning_rate": 7.666609101678121e-06,
+      "loss": 1.7792,
+      "step": 20785
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.8012621402740479,
+      "learning_rate": 7.656479914663195e-06,
+      "loss": 1.7374,
+      "step": 20790
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 1.1758484840393066,
+      "learning_rate": 7.646356213386988e-06,
+      "loss": 1.5582,
+      "step": 20795
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 1.1315603256225586,
+      "learning_rate": 7.63623800105162e-06,
+      "loss": 2.1121,
+      "step": 20800
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.9538705348968506,
+      "learning_rate": 7.626125280857471e-06,
+      "loss": 1.5779,
+      "step": 20805
+    },
+    {
+      "epoch": 3.72,
+      "grad_norm": 0.8569843769073486,
+      "learning_rate": 7.616018056003191e-06,
+      "loss": 1.8019,
+      "step": 20810
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 1.2387229204177856,
+      "learning_rate": 7.605916329685672e-06,
+      "loss": 1.6668,
+      "step": 20815
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.8599191308021545,
+      "learning_rate": 7.595820105100094e-06,
+      "loss": 1.7848,
+      "step": 20820
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.7321486473083496,
+      "learning_rate": 7.585729385439869e-06,
+      "loss": 1.7338,
+      "step": 20825
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.7330595254898071,
+      "learning_rate": 7.575644173896687e-06,
+      "loss": 1.9163,
+      "step": 20830
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.7383427619934082,
+      "learning_rate": 7.5655644736604945e-06,
+      "loss": 2.0798,
+      "step": 20835
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.7931711673736572,
+      "learning_rate": 7.555490287919497e-06,
+      "loss": 1.9642,
+      "step": 20840
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.7924222350120544,
+      "learning_rate": 7.545421619860133e-06,
+      "loss": 1.6436,
+      "step": 20845
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 1.1156904697418213,
+      "learning_rate": 7.53535847266712e-06,
+      "loss": 1.853,
+      "step": 20850
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.7994991540908813,
+      "learning_rate": 7.5253008495234255e-06,
+      "loss": 1.7786,
+      "step": 20855
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 1.3960700035095215,
+      "learning_rate": 7.515248753610266e-06,
+      "loss": 1.9,
+      "step": 20860
+    },
+    {
+      "epoch": 3.73,
+      "grad_norm": 0.9603045582771301,
+      "learning_rate": 7.505202188107113e-06,
+      "loss": 1.8195,
+      "step": 20865
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 1.181504487991333,
+      "learning_rate": 7.495161156191679e-06,
+      "loss": 1.6414,
+      "step": 20870
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.9459575414657593,
+      "learning_rate": 7.485125661039927e-06,
+      "loss": 1.9267,
+      "step": 20875
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 1.0895607471466064,
+      "learning_rate": 7.47509570582608e-06,
+      "loss": 1.8476,
+      "step": 20880
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.9939213395118713,
+      "learning_rate": 7.465071293722606e-06,
+      "loss": 1.8669,
+      "step": 20885
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.4708957076072693,
+      "learning_rate": 7.455052427900217e-06,
+      "loss": 1.7555,
+      "step": 20890
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.8547396659851074,
+      "learning_rate": 7.445039111527866e-06,
+      "loss": 1.8426,
+      "step": 20895
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.9265462160110474,
+      "learning_rate": 7.435031347772767e-06,
+      "loss": 2.0446,
+      "step": 20900
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.7409059405326843,
+      "learning_rate": 7.425029139800349e-06,
+      "loss": 2.0207,
+      "step": 20905
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.8008315563201904,
+      "learning_rate": 7.415032490774318e-06,
+      "loss": 1.6952,
+      "step": 20910
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.6471626162528992,
+      "learning_rate": 7.4050414038565856e-06,
+      "loss": 1.807,
+      "step": 20915
+    },
+    {
+      "epoch": 3.74,
+      "grad_norm": 0.8940064907073975,
+      "learning_rate": 7.3950558822073354e-06,
+      "loss": 1.6712,
+      "step": 20920
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.8687788844108582,
+      "learning_rate": 7.3850759289849745e-06,
+      "loss": 1.9076,
+      "step": 20925
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.44189202785491943,
+      "learning_rate": 7.3751015473461634e-06,
+      "loss": 1.7675,
+      "step": 20930
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.8087650537490845,
+      "learning_rate": 7.3671260556953974e-06,
+      "loss": 1.9859,
+      "step": 20935
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 1.3803483247756958,
+      "learning_rate": 7.357161710856081e-06,
+      "loss": 2.1581,
+      "step": 20940
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.845194935798645,
+      "learning_rate": 7.347202946429546e-06,
+      "loss": 1.6597,
+      "step": 20945
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 1.7201052904129028,
+      "learning_rate": 7.337249765565737e-06,
+      "loss": 1.9686,
+      "step": 20950
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.8273900747299194,
+      "learning_rate": 7.327302171412848e-06,
+      "loss": 1.7416,
+      "step": 20955
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.9852076768875122,
+      "learning_rate": 7.317360167117274e-06,
+      "loss": 1.5727,
+      "step": 20960
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.5173888802528381,
+      "learning_rate": 7.307423755823678e-06,
+      "loss": 1.8316,
+      "step": 20965
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.5297005772590637,
+      "learning_rate": 7.2974929406749246e-06,
+      "loss": 1.5954,
+      "step": 20970
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.7687706351280212,
+      "learning_rate": 7.287567724812125e-06,
+      "loss": 2.2233,
+      "step": 20975
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.9378607869148254,
+      "learning_rate": 7.277648111374616e-06,
+      "loss": 1.8363,
+      "step": 20980
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.6281090378761292,
+      "learning_rate": 7.2677341034999625e-06,
+      "loss": 1.5648,
+      "step": 20985
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 1.1043148040771484,
+      "learning_rate": 7.257825704323964e-06,
+      "loss": 1.8477,
+      "step": 20990
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 1.0617643594741821,
+      "learning_rate": 7.247922916980624e-06,
+      "loss": 1.6808,
+      "step": 20995
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 2.293015480041504,
+      "learning_rate": 7.238025744602186e-06,
+      "loss": 1.8726,
+      "step": 21000
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.5096902847290039,
+      "learning_rate": 7.22813419031913e-06,
+      "loss": 1.7613,
+      "step": 21005
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.8138590455055237,
+      "learning_rate": 7.218248257260127e-06,
+      "loss": 1.6788,
+      "step": 21010
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.5406817197799683,
+      "learning_rate": 7.2083679485520974e-06,
+      "loss": 1.8645,
+      "step": 21015
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.9503873586654663,
+      "learning_rate": 7.198493267320178e-06,
+      "loss": 1.6199,
+      "step": 21020
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 1.1343449354171753,
+      "learning_rate": 7.1886242166877084e-06,
+      "loss": 1.4947,
+      "step": 21025
+    },
+    {
+      "epoch": 3.76,
+      "grad_norm": 0.753250002861023,
+      "learning_rate": 7.1787607997762654e-06,
+      "loss": 2.0704,
+      "step": 21030
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.606606662273407,
+      "learning_rate": 7.168903019705642e-06,
+      "loss": 1.5729,
+      "step": 21035
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 1.2045537233352661,
+      "learning_rate": 7.159050879593843e-06,
+      "loss": 1.6266,
+      "step": 21040
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 1.3641705513000488,
+      "learning_rate": 7.149204382557095e-06,
+      "loss": 1.6646,
+      "step": 21045
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 1.94281804561615,
+      "learning_rate": 7.139363531709825e-06,
+      "loss": 1.7295,
+      "step": 21050
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.4692668318748474,
+      "learning_rate": 7.129528330164695e-06,
+      "loss": 1.7785,
+      "step": 21055
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 1.879866123199463,
+      "learning_rate": 7.11969878103256e-06,
+      "loss": 1.866,
+      "step": 21060
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.6057711243629456,
+      "learning_rate": 7.109874887422499e-06,
+      "loss": 1.9315,
+      "step": 21065
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.5978086590766907,
+      "learning_rate": 7.1000566524418055e-06,
+      "loss": 1.9369,
+      "step": 21070
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 1.631386637687683,
+      "learning_rate": 7.090244079195976e-06,
+      "loss": 1.6194,
+      "step": 21075
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.8467076420783997,
+      "learning_rate": 7.080437170788723e-06,
+      "loss": 1.9249,
+      "step": 21080
+    },
+    {
+      "epoch": 3.77,
+      "grad_norm": 0.6821494698524475,
+      "learning_rate": 7.070635930321948e-06,
+      "loss": 1.7521,
+      "step": 21085
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.9568500518798828,
+      "learning_rate": 7.060840360895785e-06,
+      "loss": 1.7341,
+      "step": 21090
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 1.729252576828003,
+      "learning_rate": 7.051050465608558e-06,
+      "loss": 1.6962,
+      "step": 21095
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 1.2900949716567993,
+      "learning_rate": 7.041266247556813e-06,
+      "loss": 2.2104,
+      "step": 21100
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 1.797668695449829,
+      "learning_rate": 7.031487709835272e-06,
+      "loss": 1.7064,
+      "step": 21105
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.8635187745094299,
+      "learning_rate": 7.021714855536893e-06,
+      "loss": 1.7889,
+      "step": 21110
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.6783146858215332,
+      "learning_rate": 7.011947687752804e-06,
+      "loss": 1.8269,
+      "step": 21115
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 1.27726411819458,
+      "learning_rate": 7.002186209572359e-06,
+      "loss": 1.9167,
+      "step": 21120
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.873021125793457,
+      "learning_rate": 6.9924304240831045e-06,
+      "loss": 1.7524,
+      "step": 21125
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.6796855926513672,
+      "learning_rate": 6.982680334370786e-06,
+      "loss": 1.899,
+      "step": 21130
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 1.1195898056030273,
+      "learning_rate": 6.972935943519354e-06,
+      "loss": 2.2136,
+      "step": 21135
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 1.5191144943237305,
+      "learning_rate": 6.963197254610937e-06,
+      "loss": 2.0849,
+      "step": 21140
+    },
+    {
+      "epoch": 3.78,
+      "grad_norm": 0.978111982345581,
+      "learning_rate": 6.9534642707258875e-06,
+      "loss": 1.9883,
+      "step": 21145
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 1.8795119524002075,
+      "learning_rate": 6.943736994942721e-06,
+      "loss": 1.6622,
+      "step": 21150
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 1.2708956003189087,
+      "learning_rate": 6.934015430338176e-06,
+      "loss": 1.9199,
+      "step": 21155
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.7446703910827637,
+      "learning_rate": 6.924299579987173e-06,
+      "loss": 1.885,
+      "step": 21160
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 1.003434658050537,
+      "learning_rate": 6.914589446962824e-06,
+      "loss": 1.8679,
+      "step": 21165
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.9047329425811768,
+      "learning_rate": 6.904885034336448e-06,
+      "loss": 1.7717,
+      "step": 21170
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.6826780438423157,
+      "learning_rate": 6.895186345177521e-06,
+      "loss": 1.9632,
+      "step": 21175
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.5355373024940491,
+      "learning_rate": 6.885493382553737e-06,
+      "loss": 2.0283,
+      "step": 21180
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.8014163374900818,
+      "learning_rate": 6.875806149530975e-06,
+      "loss": 1.7532,
+      "step": 21185
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 0.9986274242401123,
+      "learning_rate": 6.866124649173295e-06,
+      "loss": 1.9851,
+      "step": 21190
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 1.0653488636016846,
+      "learning_rate": 6.8564488845429515e-06,
+      "loss": 2.1604,
+      "step": 21195
+    },
+    {
+      "epoch": 3.79,
+      "grad_norm": 1.1097596883773804,
+      "learning_rate": 6.846778858700373e-06,
+      "loss": 1.6555,
+      "step": 21200
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 15.828474044799805,
+      "learning_rate": 6.837114574704176e-06,
+      "loss": 1.6847,
+      "step": 21205
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 3.438732385635376,
+      "learning_rate": 6.827456035611166e-06,
+      "loss": 1.7038,
+      "step": 21210
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.7653161287307739,
+      "learning_rate": 6.817803244476331e-06,
+      "loss": 1.8852,
+      "step": 21215
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 1.15346360206604,
+      "learning_rate": 6.8081562043528445e-06,
+      "loss": 1.6198,
+      "step": 21220
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.6364823579788208,
+      "learning_rate": 6.7985149182920564e-06,
+      "loss": 1.8619,
+      "step": 21225
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 1.149183988571167,
+      "learning_rate": 6.788879389343486e-06,
+      "loss": 1.9567,
+      "step": 21230
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.8027864694595337,
+      "learning_rate": 6.7792496205548475e-06,
+      "loss": 1.9675,
+      "step": 21235
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.7436734437942505,
+      "learning_rate": 6.769625614972036e-06,
+      "loss": 1.8813,
+      "step": 21240
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 1.2169238328933716,
+      "learning_rate": 6.760007375639099e-06,
+      "loss": 1.8129,
+      "step": 21245
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 0.6234614253044128,
+      "learning_rate": 6.750394905598284e-06,
+      "loss": 1.8025,
+      "step": 21250
+    },
+    {
+      "epoch": 3.8,
+      "grad_norm": 2.2738096714019775,
+      "learning_rate": 6.740788207890017e-06,
+      "loss": 2.0495,
+      "step": 21255
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 4.580897808074951,
+      "learning_rate": 6.731187285552865e-06,
+      "loss": 1.4265,
+      "step": 21260
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 1.344910740852356,
+      "learning_rate": 6.721592141623606e-06,
+      "loss": 1.9222,
+      "step": 21265
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 1.4023634195327759,
+      "learning_rate": 6.712002779137169e-06,
+      "loss": 1.7676,
+      "step": 21270
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 2.361953020095825,
+      "learning_rate": 6.702419201126661e-06,
+      "loss": 1.6256,
+      "step": 21275
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.7495207190513611,
+      "learning_rate": 6.69284141062336e-06,
+      "loss": 1.9166,
+      "step": 21280
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.5383853316307068,
+      "learning_rate": 6.68326941065672e-06,
+      "loss": 1.8247,
+      "step": 21285
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.7718889713287354,
+      "learning_rate": 6.67370320425435e-06,
+      "loss": 1.959,
+      "step": 21290
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.666534423828125,
+      "learning_rate": 6.6641427944420185e-06,
+      "loss": 1.9477,
+      "step": 21295
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 1.1692932844161987,
+      "learning_rate": 6.654588184243682e-06,
+      "loss": 1.7154,
+      "step": 21300
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 0.8114296793937683,
+      "learning_rate": 6.645039376681459e-06,
+      "loss": 1.6824,
+      "step": 21305
+    },
+    {
+      "epoch": 3.81,
+      "grad_norm": 1.2504725456237793,
+      "learning_rate": 6.635496374775626e-06,
+      "loss": 1.7564,
+      "step": 21310
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 1.4884883165359497,
+      "learning_rate": 6.6259591815446335e-06,
+      "loss": 1.4289,
+      "step": 21315
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.705243706703186,
+      "learning_rate": 6.616427800005068e-06,
+      "loss": 1.8026,
+      "step": 21320
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.7473030090332031,
+      "learning_rate": 6.606902233171711e-06,
+      "loss": 1.9331,
+      "step": 21325
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.8204745054244995,
+      "learning_rate": 6.597382484057482e-06,
+      "loss": 1.798,
+      "step": 21330
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.48442649841308594,
+      "learning_rate": 6.587868555673479e-06,
+      "loss": 1.8777,
+      "step": 21335
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 1.1391490697860718,
+      "learning_rate": 6.5783604510289365e-06,
+      "loss": 1.8687,
+      "step": 21340
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.8502964973449707,
+      "learning_rate": 6.56885817313127e-06,
+      "loss": 1.8788,
+      "step": 21345
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 2.4996724128723145,
+      "learning_rate": 6.559361724986029e-06,
+      "loss": 2.0761,
+      "step": 21350
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.40672042965888977,
+      "learning_rate": 6.549871109596939e-06,
+      "loss": 1.8876,
+      "step": 21355
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 1.889902949333191,
+      "learning_rate": 6.5403863299658665e-06,
+      "loss": 1.8512,
+      "step": 21360
+    },
+    {
+      "epoch": 3.82,
+      "grad_norm": 0.7777634263038635,
+      "learning_rate": 6.530907389092842e-06,
+      "loss": 2.0392,
+      "step": 21365
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.6818479299545288,
+      "learning_rate": 6.521434289976047e-06,
+      "loss": 1.6088,
+      "step": 21370
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.5815435647964478,
+      "learning_rate": 6.511967035611816e-06,
+      "loss": 1.9384,
+      "step": 21375
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 1.1326960325241089,
+      "learning_rate": 6.5025056289946315e-06,
+      "loss": 2.0075,
+      "step": 21380
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.7850456833839417,
+      "learning_rate": 6.493050073117116e-06,
+      "loss": 1.8053,
+      "step": 21385
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 1.2165858745574951,
+      "learning_rate": 6.483600370970061e-06,
+      "loss": 1.6863,
+      "step": 21390
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 1.1443709135055542,
+      "learning_rate": 6.474156525542399e-06,
+      "loss": 1.9593,
+      "step": 21395
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.9057660698890686,
+      "learning_rate": 6.4647185398212075e-06,
+      "loss": 1.7776,
+      "step": 21400
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 1.2972538471221924,
+      "learning_rate": 6.45528641679172e-06,
+      "loss": 1.7882,
+      "step": 21405
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 1.1421040296554565,
+      "learning_rate": 6.445860159437295e-06,
+      "loss": 1.4755,
+      "step": 21410
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 1.4928812980651855,
+      "learning_rate": 6.4364397707394555e-06,
+      "loss": 1.8019,
+      "step": 21415
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 1.1022720336914062,
+      "learning_rate": 6.427025253677857e-06,
+      "loss": 2.0423,
+      "step": 21420
+    },
+    {
+      "epoch": 3.83,
+      "grad_norm": 0.6213718056678772,
+      "learning_rate": 6.417616611230318e-06,
+      "loss": 1.8774,
+      "step": 21425
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 1.03056800365448,
+      "learning_rate": 6.40821384637276e-06,
+      "loss": 1.995,
+      "step": 21430
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.6993535757064819,
+      "learning_rate": 6.398816962079287e-06,
+      "loss": 2.0732,
+      "step": 21435
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 1.2051020860671997,
+      "learning_rate": 6.3894259613221095e-06,
+      "loss": 1.5418,
+      "step": 21440
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.8141611218452454,
+      "learning_rate": 6.380040847071603e-06,
+      "loss": 1.536,
+      "step": 21445
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.784510612487793,
+      "learning_rate": 6.370661622296261e-06,
+      "loss": 1.8054,
+      "step": 21450
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 1.0735670328140259,
+      "learning_rate": 6.361288289962733e-06,
+      "loss": 1.9683,
+      "step": 21455
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.6221914887428284,
+      "learning_rate": 6.351920853035795e-06,
+      "loss": 1.829,
+      "step": 21460
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.5578571557998657,
+      "learning_rate": 6.34255931447835e-06,
+      "loss": 1.6748,
+      "step": 21465
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.4949614703655243,
+      "learning_rate": 6.333203677251445e-06,
+      "loss": 1.9189,
+      "step": 21470
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 1.253214716911316,
+      "learning_rate": 6.323853944314267e-06,
+      "loss": 1.5789,
+      "step": 21475
+    },
+    {
+      "epoch": 3.84,
+      "grad_norm": 0.5360049605369568,
+      "learning_rate": 6.3145101186241175e-06,
+      "loss": 1.7498,
+      "step": 21480
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.6865981221199036,
+      "learning_rate": 6.305172203136447e-06,
+      "loss": 1.8506,
+      "step": 21485
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.7552772760391235,
+      "learning_rate": 6.295840200804823e-06,
+      "loss": 1.9033,
+      "step": 21490
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.9978206157684326,
+      "learning_rate": 6.2865141145809604e-06,
+      "loss": 2.001,
+      "step": 21495
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.6220951080322266,
+      "learning_rate": 6.277193947414678e-06,
+      "loss": 1.809,
+      "step": 21500
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 1.6220732927322388,
+      "learning_rate": 6.267879702253937e-06,
+      "loss": 1.789,
+      "step": 21505
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.6907435059547424,
+      "learning_rate": 6.258571382044831e-06,
+      "loss": 2.0546,
+      "step": 21510
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.738203227519989,
+      "learning_rate": 6.2492689897315685e-06,
+      "loss": 2.0193,
+      "step": 21515
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.5000423192977905,
+      "learning_rate": 6.239972528256497e-06,
+      "loss": 1.666,
+      "step": 21520
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.42941489815711975,
+      "learning_rate": 6.23068200056007e-06,
+      "loss": 1.9748,
+      "step": 21525
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.4890025854110718,
+      "learning_rate": 6.221397409580864e-06,
+      "loss": 2.0923,
+      "step": 21530
+    },
+    {
+      "epoch": 3.85,
+      "grad_norm": 0.8558998107910156,
+      "learning_rate": 6.212118758255595e-06,
+      "loss": 1.7637,
+      "step": 21535
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.7938655614852905,
+      "learning_rate": 6.202846049519093e-06,
+      "loss": 1.752,
+      "step": 21540
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 1.0123392343521118,
+      "learning_rate": 6.193579286304302e-06,
+      "loss": 1.6815,
+      "step": 21545
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 3.046292781829834,
+      "learning_rate": 6.184318471542303e-06,
+      "loss": 1.7829,
+      "step": 21550
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.652489423751831,
+      "learning_rate": 6.175063608162266e-06,
+      "loss": 1.903,
+      "step": 21555
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.838995099067688,
+      "learning_rate": 6.165814699091499e-06,
+      "loss": 1.789,
+      "step": 21560
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.7666052579879761,
+      "learning_rate": 6.156571747255435e-06,
+      "loss": 1.6717,
+      "step": 21565
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.5006678700447083,
+      "learning_rate": 6.147334755577596e-06,
+      "loss": 1.9561,
+      "step": 21570
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.8589296340942383,
+      "learning_rate": 6.138103726979641e-06,
+      "loss": 1.9194,
+      "step": 21575
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.4786919951438904,
+      "learning_rate": 6.128878664381332e-06,
+      "loss": 1.9861,
+      "step": 21580
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 1.460368275642395,
+      "learning_rate": 6.119659570700559e-06,
+      "loss": 1.6571,
+      "step": 21585
+    },
+    {
+      "epoch": 3.86,
+      "grad_norm": 0.6441243290901184,
+      "learning_rate": 6.110446448853293e-06,
+      "loss": 2.1022,
+      "step": 21590
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.36151692271232605,
+      "learning_rate": 6.10123930175365e-06,
+      "loss": 2.0311,
+      "step": 21595
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.6082046031951904,
+      "learning_rate": 6.092038132313835e-06,
+      "loss": 1.7961,
+      "step": 21600
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.5249208211898804,
+      "learning_rate": 6.082842943444172e-06,
+      "loss": 1.7438,
+      "step": 21605
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.9609794020652771,
+      "learning_rate": 6.073653738053098e-06,
+      "loss": 2.2195,
+      "step": 21610
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.8128685355186462,
+      "learning_rate": 6.064470519047141e-06,
+      "loss": 1.6784,
+      "step": 21615
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 1.8232717514038086,
+      "learning_rate": 6.0552932893309374e-06,
+      "loss": 1.8214,
+      "step": 21620
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 1.522959589958191,
+      "learning_rate": 6.046122051807246e-06,
+      "loss": 1.5712,
+      "step": 21625
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.9963710904121399,
+      "learning_rate": 6.036956809376915e-06,
+      "loss": 1.7489,
+      "step": 21630
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 1.037346601486206,
+      "learning_rate": 6.027797564938906e-06,
+      "loss": 1.5724,
+      "step": 21635
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 1.6422029733657837,
+      "learning_rate": 6.018644321390288e-06,
+      "loss": 1.7118,
+      "step": 21640
+    },
+    {
+      "epoch": 3.87,
+      "grad_norm": 0.9936122298240662,
+      "learning_rate": 6.009497081626203e-06,
+      "loss": 1.8194,
+      "step": 21645
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 1.9703397750854492,
+      "learning_rate": 6.000355848539926e-06,
+      "loss": 1.6014,
+      "step": 21650
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.7200276851654053,
+      "learning_rate": 5.99122062502282e-06,
+      "loss": 1.9046,
+      "step": 21655
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.6927645802497864,
+      "learning_rate": 5.982091413964352e-06,
+      "loss": 1.6747,
+      "step": 21660
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 1.2343568801879883,
+      "learning_rate": 5.97296821825207e-06,
+      "loss": 1.6779,
+      "step": 21665
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.5954350829124451,
+      "learning_rate": 5.9638510407716394e-06,
+      "loss": 1.791,
+      "step": 21670
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 1.0342341661453247,
+      "learning_rate": 5.954739884406821e-06,
+      "loss": 1.7339,
+      "step": 21675
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 1.0114802122116089,
+      "learning_rate": 5.94563475203945e-06,
+      "loss": 1.9951,
+      "step": 21680
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 2.355144500732422,
+      "learning_rate": 5.936535646549479e-06,
+      "loss": 1.668,
+      "step": 21685
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.5509099960327148,
+      "learning_rate": 5.927442570814948e-06,
+      "loss": 1.819,
+      "step": 21690
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.8965473175048828,
+      "learning_rate": 5.918355527711983e-06,
+      "loss": 1.8374,
+      "step": 21695
+    },
+    {
+      "epoch": 3.88,
+      "grad_norm": 0.8329721093177795,
+      "learning_rate": 5.9092745201148215e-06,
+      "loss": 1.7613,
+      "step": 21700
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.7440474033355713,
+      "learning_rate": 5.900199550895763e-06,
+      "loss": 1.7955,
+      "step": 21705
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.6702372431755066,
+      "learning_rate": 5.891130622925209e-06,
+      "loss": 1.5468,
+      "step": 21710
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 1.7360087633132935,
+      "learning_rate": 5.882067739071659e-06,
+      "loss": 1.8326,
+      "step": 21715
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.6325078010559082,
+      "learning_rate": 5.873010902201695e-06,
+      "loss": 1.8646,
+      "step": 21720
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.45870426297187805,
+      "learning_rate": 5.863960115179984e-06,
+      "loss": 1.8511,
+      "step": 21725
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 1.7255089282989502,
+      "learning_rate": 5.854915380869286e-06,
+      "loss": 1.8966,
+      "step": 21730
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.8191224932670593,
+      "learning_rate": 5.845876702130434e-06,
+      "loss": 1.8138,
+      "step": 21735
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 1.0439651012420654,
+      "learning_rate": 5.836844081822357e-06,
+      "loss": 1.6274,
+      "step": 21740
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.7920172810554504,
+      "learning_rate": 5.827817522802065e-06,
+      "loss": 1.9206,
+      "step": 21745
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 1.2962287664413452,
+      "learning_rate": 5.818797027924647e-06,
+      "loss": 1.9135,
+      "step": 21750
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 0.9365116357803345,
+      "learning_rate": 5.809782600043287e-06,
+      "loss": 1.7649,
+      "step": 21755
+    },
+    {
+      "epoch": 3.89,
+      "grad_norm": 3.8692009449005127,
+      "learning_rate": 5.800774242009227e-06,
+      "loss": 1.8553,
+      "step": 21760
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 2.046795606613159,
+      "learning_rate": 5.7917719566718095e-06,
+      "loss": 1.7213,
+      "step": 21765
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.8467017412185669,
+      "learning_rate": 5.782775746878444e-06,
+      "loss": 1.9483,
+      "step": 21770
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.7216947078704834,
+      "learning_rate": 5.773785615474622e-06,
+      "loss": 2.0848,
+      "step": 21775
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.9281513690948486,
+      "learning_rate": 5.764801565303918e-06,
+      "loss": 1.9586,
+      "step": 21780
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 3.031953811645508,
+      "learning_rate": 5.755823599207977e-06,
+      "loss": 1.7133,
+      "step": 21785
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.6615840196609497,
+      "learning_rate": 5.7468517200265276e-06,
+      "loss": 2.0092,
+      "step": 21790
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 1.0913735628128052,
+      "learning_rate": 5.737885930597351e-06,
+      "loss": 1.6966,
+      "step": 21795
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 1.035690188407898,
+      "learning_rate": 5.728926233756338e-06,
+      "loss": 1.7959,
+      "step": 21800
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.7709459662437439,
+      "learning_rate": 5.719972632337409e-06,
+      "loss": 1.9455,
+      "step": 21805
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.6541898250579834,
+      "learning_rate": 5.711025129172595e-06,
+      "loss": 1.6532,
+      "step": 21810
+    },
+    {
+      "epoch": 3.9,
+      "grad_norm": 0.6813231110572815,
+      "learning_rate": 5.702083727091978e-06,
+      "loss": 1.9995,
+      "step": 21815
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.7361324429512024,
+      "learning_rate": 5.693148428923722e-06,
+      "loss": 1.9465,
+      "step": 21820
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 1.0799798965454102,
+      "learning_rate": 5.68421923749404e-06,
+      "loss": 1.8866,
+      "step": 21825
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.9929239749908447,
+      "learning_rate": 5.675296155627236e-06,
+      "loss": 1.7591,
+      "step": 21830
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.9932503700256348,
+      "learning_rate": 5.666379186145668e-06,
+      "loss": 1.7421,
+      "step": 21835
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.8309764862060547,
+      "learning_rate": 5.6574683318697665e-06,
+      "loss": 1.9331,
+      "step": 21840
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.8258433938026428,
+      "learning_rate": 5.6485635956180365e-06,
+      "loss": 1.8122,
+      "step": 21845
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 1.11664617061615,
+      "learning_rate": 5.639664980207024e-06,
+      "loss": 1.8511,
+      "step": 21850
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.7466225624084473,
+      "learning_rate": 5.63077248845135e-06,
+      "loss": 1.9555,
+      "step": 21855
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.9185487031936646,
+      "learning_rate": 5.621886123163708e-06,
+      "loss": 1.6394,
+      "step": 21860
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 1.338221788406372,
+      "learning_rate": 5.613005887154845e-06,
+      "loss": 2.0797,
+      "step": 21865
+    },
+    {
+      "epoch": 3.91,
+      "grad_norm": 0.5035430192947388,
+      "learning_rate": 5.604131783233574e-06,
+      "loss": 1.6932,
+      "step": 21870
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 1.3614168167114258,
+      "learning_rate": 5.595263814206761e-06,
+      "loss": 1.7473,
+      "step": 21875
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 1.0385257005691528,
+      "learning_rate": 5.586401982879347e-06,
+      "loss": 1.7702,
+      "step": 21880
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.7399263381958008,
+      "learning_rate": 5.5775462920543035e-06,
+      "loss": 1.8132,
+      "step": 21885
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.5867356061935425,
+      "learning_rate": 5.568696744532687e-06,
+      "loss": 1.7692,
+      "step": 21890
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 1.5751367807388306,
+      "learning_rate": 5.5598533431136055e-06,
+      "loss": 2.0297,
+      "step": 21895
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.8599883317947388,
+      "learning_rate": 5.551016090594208e-06,
+      "loss": 1.8222,
+      "step": 21900
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.6561763882637024,
+      "learning_rate": 5.542184989769711e-06,
+      "loss": 1.8176,
+      "step": 21905
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 1.0254570245742798,
+      "learning_rate": 5.533360043433394e-06,
+      "loss": 1.8785,
+      "step": 21910
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 1.0083361864089966,
+      "learning_rate": 5.524541254376564e-06,
+      "loss": 1.8599,
+      "step": 21915
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.6952515244483948,
+      "learning_rate": 5.515728625388603e-06,
+      "loss": 1.6481,
+      "step": 21920
+    },
+    {
+      "epoch": 3.92,
+      "grad_norm": 0.6888887286186218,
+      "learning_rate": 5.506922159256933e-06,
+      "loss": 1.9352,
+      "step": 21925
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 1.0823732614517212,
+      "learning_rate": 5.498121858767036e-06,
+      "loss": 1.7055,
+      "step": 21930
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 1.0809731483459473,
+      "learning_rate": 5.489327726702445e-06,
+      "loss": 2.1196,
+      "step": 21935
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 2.2783148288726807,
+      "learning_rate": 5.480539765844725e-06,
+      "loss": 1.6809,
+      "step": 21940
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 1.1743180751800537,
+      "learning_rate": 5.471757978973496e-06,
+      "loss": 1.5346,
+      "step": 21945
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.7503143548965454,
+      "learning_rate": 5.462982368866435e-06,
+      "loss": 1.7871,
+      "step": 21950
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.5664401650428772,
+      "learning_rate": 5.454212938299258e-06,
+      "loss": 1.9305,
+      "step": 21955
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.5282362103462219,
+      "learning_rate": 5.445449690045725e-06,
+      "loss": 2.0377,
+      "step": 21960
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.9779241681098938,
+      "learning_rate": 5.436692626877648e-06,
+      "loss": 1.4596,
+      "step": 21965
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 1.3879930973052979,
+      "learning_rate": 5.427941751564886e-06,
+      "loss": 1.8408,
+      "step": 21970
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 0.5937609672546387,
+      "learning_rate": 5.419197066875312e-06,
+      "loss": 1.9729,
+      "step": 21975
+    },
+    {
+      "epoch": 3.93,
+      "grad_norm": 1.0135959386825562,
+      "learning_rate": 5.41045857557487e-06,
+      "loss": 1.7513,
+      "step": 21980
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.7260029315948486,
+      "learning_rate": 5.401726280427541e-06,
+      "loss": 1.919,
+      "step": 21985
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.8399494290351868,
+      "learning_rate": 5.393000184195343e-06,
+      "loss": 1.6943,
+      "step": 21990
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 1.7650916576385498,
+      "learning_rate": 5.384280289638322e-06,
+      "loss": 1.5243,
+      "step": 21995
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.5501816868782043,
+      "learning_rate": 5.375566599514581e-06,
+      "loss": 2.116,
+      "step": 22000
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.9165323376655579,
+      "learning_rate": 5.366859116580244e-06,
+      "loss": 1.7613,
+      "step": 22005
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.5392761826515198,
+      "learning_rate": 5.358157843589484e-06,
+      "loss": 1.7364,
+      "step": 22010
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.6568076014518738,
+      "learning_rate": 5.349462783294504e-06,
+      "loss": 1.7346,
+      "step": 22015
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 1.0636271238327026,
+      "learning_rate": 5.340773938445545e-06,
+      "loss": 1.8723,
+      "step": 22020
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.8600271344184875,
+      "learning_rate": 5.332091311790885e-06,
+      "loss": 1.5156,
+      "step": 22025
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.602660596370697,
+      "learning_rate": 5.323414906076821e-06,
+      "loss": 1.7696,
+      "step": 22030
+    },
+    {
+      "epoch": 3.94,
+      "grad_norm": 0.6825761198997498,
+      "learning_rate": 5.314744724047702e-06,
+      "loss": 1.6636,
+      "step": 22035
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.6560852527618408,
+      "learning_rate": 5.306080768445881e-06,
+      "loss": 1.8196,
+      "step": 22040
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.8872010111808777,
+      "learning_rate": 5.297423042011774e-06,
+      "loss": 1.9585,
+      "step": 22045
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.6532019972801208,
+      "learning_rate": 5.288771547483806e-06,
+      "loss": 1.7353,
+      "step": 22050
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 1.1290751695632935,
+      "learning_rate": 5.280126287598444e-06,
+      "loss": 1.617,
+      "step": 22055
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.7250970602035522,
+      "learning_rate": 5.271487265090163e-06,
+      "loss": 1.6928,
+      "step": 22060
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.8637431859970093,
+      "learning_rate": 5.2628544826914846e-06,
+      "loss": 1.5799,
+      "step": 22065
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 3.6587893962860107,
+      "learning_rate": 5.254227943132948e-06,
+      "loss": 1.6546,
+      "step": 22070
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.8665122985839844,
+      "learning_rate": 5.245607649143119e-06,
+      "loss": 1.8547,
+      "step": 22075
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.883367657661438,
+      "learning_rate": 5.236993603448595e-06,
+      "loss": 1.7483,
+      "step": 22080
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 2.016577959060669,
+      "learning_rate": 5.2283858087739806e-06,
+      "loss": 1.5007,
+      "step": 22085
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 0.9476630687713623,
+      "learning_rate": 5.2197842678419204e-06,
+      "loss": 1.8659,
+      "step": 22090
+    },
+    {
+      "epoch": 3.95,
+      "grad_norm": 1.0685213804244995,
+      "learning_rate": 5.211188983373067e-06,
+      "loss": 1.8435,
+      "step": 22095
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 1.0245375633239746,
+      "learning_rate": 5.202599958086102e-06,
+      "loss": 1.6491,
+      "step": 22100
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 1.622461199760437,
+      "learning_rate": 5.194017194697728e-06,
+      "loss": 1.7717,
+      "step": 22105
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.615947961807251,
+      "learning_rate": 5.185440695922664e-06,
+      "loss": 1.9252,
+      "step": 22110
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.5909544229507446,
+      "learning_rate": 5.176870464473657e-06,
+      "loss": 2.1178,
+      "step": 22115
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.8277536034584045,
+      "learning_rate": 5.168306503061446e-06,
+      "loss": 1.8633,
+      "step": 22120
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.8978816270828247,
+      "learning_rate": 5.159748814394813e-06,
+      "loss": 1.622,
+      "step": 22125
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.6580782532691956,
+      "learning_rate": 5.151197401180552e-06,
+      "loss": 1.6254,
+      "step": 22130
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.612777590751648,
+      "learning_rate": 5.142652266123454e-06,
+      "loss": 1.7793,
+      "step": 22135
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.7944098114967346,
+      "learning_rate": 5.1341134119263476e-06,
+      "loss": 1.5077,
+      "step": 22140
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 0.791584849357605,
+      "learning_rate": 5.125580841290062e-06,
+      "loss": 1.7463,
+      "step": 22145
+    },
+    {
+      "epoch": 3.96,
+      "grad_norm": 1.069318175315857,
+      "learning_rate": 5.1170545569134365e-06,
+      "loss": 1.7364,
+      "step": 22150
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.9720897078514099,
+      "learning_rate": 5.108534561493328e-06,
+      "loss": 1.8411,
+      "step": 22155
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.4824497699737549,
+      "learning_rate": 5.100020857724605e-06,
+      "loss": 1.7391,
+      "step": 22160
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 1.237189531326294,
+      "learning_rate": 5.091513448300144e-06,
+      "loss": 1.7938,
+      "step": 22165
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.8964689373970032,
+      "learning_rate": 5.083012335910828e-06,
+      "loss": 1.8008,
+      "step": 22170
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.6493983864784241,
+      "learning_rate": 5.074517523245561e-06,
+      "loss": 2.0167,
+      "step": 22175
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 1.8859643936157227,
+      "learning_rate": 5.066029012991236e-06,
+      "loss": 1.8854,
+      "step": 22180
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.8033244609832764,
+      "learning_rate": 5.05754680783275e-06,
+      "loss": 1.9866,
+      "step": 22185
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 1.5448501110076904,
+      "learning_rate": 5.04907091045303e-06,
+      "loss": 1.7147,
+      "step": 22190
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 2.1923139095306396,
+      "learning_rate": 5.04060132353299e-06,
+      "loss": 1.752,
+      "step": 22195
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.8507521152496338,
+      "learning_rate": 5.03213804975155e-06,
+      "loss": 1.8768,
+      "step": 22200
+    },
+    {
+      "epoch": 3.97,
+      "grad_norm": 0.5235335826873779,
+      "learning_rate": 5.0236810917856485e-06,
+      "loss": 1.9388,
+      "step": 22205
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.9762120842933655,
+      "learning_rate": 5.015230452310193e-06,
+      "loss": 1.7633,
+      "step": 22210
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.9791060090065002,
+      "learning_rate": 5.006786133998126e-06,
+      "loss": 1.5675,
+      "step": 22215
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 1.307426929473877,
+      "learning_rate": 4.998348139520378e-06,
+      "loss": 1.5626,
+      "step": 22220
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 2.5995423793792725,
+      "learning_rate": 4.989916471545869e-06,
+      "loss": 1.8907,
+      "step": 22225
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 1.0652718544006348,
+      "learning_rate": 4.981491132741531e-06,
+      "loss": 2.1655,
+      "step": 22230
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.8676015734672546,
+      "learning_rate": 4.973072125772302e-06,
+      "loss": 1.9071,
+      "step": 22235
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.7337488532066345,
+      "learning_rate": 4.9646594533010875e-06,
+      "loss": 1.6795,
+      "step": 22240
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.6141374111175537,
+      "learning_rate": 4.956253117988816e-06,
+      "loss": 2.0898,
+      "step": 22245
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.5068919062614441,
+      "learning_rate": 4.947853122494406e-06,
+      "loss": 1.964,
+      "step": 22250
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 1.0111421346664429,
+      "learning_rate": 4.939459469474764e-06,
+      "loss": 1.8229,
+      "step": 22255
+    },
+    {
+      "epoch": 3.98,
+      "grad_norm": 0.9964427947998047,
+      "learning_rate": 4.931072161584802e-06,
+      "loss": 1.886,
+      "step": 22260
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.7016617059707642,
+      "learning_rate": 4.922691201477406e-06,
+      "loss": 1.7212,
+      "step": 22265
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.8478853106498718,
+      "learning_rate": 4.914316591803475e-06,
+      "loss": 1.7993,
+      "step": 22270
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.7734459042549133,
+      "learning_rate": 4.905948335211882e-06,
+      "loss": 2.101,
+      "step": 22275
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.6338139176368713,
+      "learning_rate": 4.897586434349502e-06,
+      "loss": 1.8311,
+      "step": 22280
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 2.914543867111206,
+      "learning_rate": 4.889230891861196e-06,
+      "loss": 1.9766,
+      "step": 22285
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.9734477400779724,
+      "learning_rate": 4.880881710389817e-06,
+      "loss": 1.8301,
+      "step": 22290
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 1.7465955018997192,
+      "learning_rate": 4.8725388925762064e-06,
+      "loss": 1.8134,
+      "step": 22295
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.780785322189331,
+      "learning_rate": 4.864202441059176e-06,
+      "loss": 1.8292,
+      "step": 22300
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.9263495206832886,
+      "learning_rate": 4.855872358475547e-06,
+      "loss": 1.8219,
+      "step": 22305
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.8709249496459961,
+      "learning_rate": 4.847548647460112e-06,
+      "loss": 1.8204,
+      "step": 22310
+    },
+    {
+      "epoch": 3.99,
+      "grad_norm": 0.9024534821510315,
+      "learning_rate": 4.839231310645664e-06,
+      "loss": 1.7256,
+      "step": 22315
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.1584830284118652,
+      "learning_rate": 4.830920350662954e-06,
+      "loss": 1.6609,
+      "step": 22320
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.5752809643745422,
+      "learning_rate": 4.822615770140745e-06,
+      "loss": 1.6129,
+      "step": 22325
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.2050139904022217,
+      "learning_rate": 4.8143175717057505e-06,
+      "loss": 1.5614,
+      "step": 22330
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.9474218487739563,
+      "learning_rate": 4.806025757982693e-06,
+      "loss": 1.9284,
+      "step": 22335
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.6500478982925415,
+      "learning_rate": 4.797740331594264e-06,
+      "loss": 1.9486,
+      "step": 22340
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.0756356716156006,
+      "learning_rate": 4.789461295161138e-06,
+      "loss": 1.9763,
+      "step": 22345
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.6653231382369995,
+      "learning_rate": 4.781188651301969e-06,
+      "loss": 2.0301,
+      "step": 22350
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.7933474779129028,
+      "learning_rate": 4.772922402633373e-06,
+      "loss": 1.9962,
+      "step": 22355
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 2.7801687717437744,
+      "learning_rate": 4.764662551769975e-06,
+      "loss": 1.9394,
+      "step": 22360
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.8057134747505188,
+      "learning_rate": 4.756409101324338e-06,
+      "loss": 1.7502,
+      "step": 22365
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 2.9395413398742676,
+      "learning_rate": 4.748162053907035e-06,
+      "loss": 1.8419,
+      "step": 22370
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 2.3610217571258545,
+      "learning_rate": 4.739921412126591e-06,
+      "loss": 1.9091,
+      "step": 22375
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 0.6767581105232239,
+      "learning_rate": 4.731687178589517e-06,
+      "loss": 1.7989,
+      "step": 22380
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 0.8745439052581787,
+      "learning_rate": 4.7234593559003e-06,
+      "loss": 1.8339,
+      "step": 22385
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 0.7209843993186951,
+      "learning_rate": 4.715237946661377e-06,
+      "loss": 1.8103,
+      "step": 22390
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 2.5547735691070557,
+      "learning_rate": 4.707022953473184e-06,
+      "loss": 2.2534,
+      "step": 22395
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 0.7963310480117798,
+      "learning_rate": 4.698814378934108e-06,
+      "loss": 1.888,
+      "step": 22400
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 1.050917148590088,
+      "learning_rate": 4.690612225640517e-06,
+      "loss": 1.9726,
+      "step": 22405
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 0.7625211477279663,
+      "learning_rate": 4.68241649618675e-06,
+      "loss": 1.9173,
+      "step": 22410
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 0.531043291091919,
+      "learning_rate": 4.674227193165107e-06,
+      "loss": 2.0842,
+      "step": 22415
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 0.671322226524353,
+      "learning_rate": 4.666044319165841e-06,
+      "loss": 1.6739,
+      "step": 22420
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 0.8251287341117859,
+      "learning_rate": 4.6578678767771975e-06,
+      "loss": 1.9518,
+      "step": 22425
+    },
+    {
+      "epoch": 4.01,
+      "grad_norm": 1.6571036577224731,
+      "learning_rate": 4.6496978685853786e-06,
+      "loss": 1.6041,
+      "step": 22430
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 0.7980236411094666,
+      "learning_rate": 4.641534297174549e-06,
+      "loss": 1.4048,
+      "step": 22435
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 1.075265884399414,
+      "learning_rate": 4.633377165126848e-06,
+      "loss": 1.9821,
+      "step": 22440
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 1.2659225463867188,
+      "learning_rate": 4.625226475022351e-06,
+      "loss": 1.6767,
+      "step": 22445
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 3.332648754119873,
+      "learning_rate": 4.6170822294391205e-06,
+      "loss": 1.6765,
+      "step": 22450
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 1.1672558784484863,
+      "learning_rate": 4.608944430953185e-06,
+      "loss": 1.6675,
+      "step": 22455
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 0.6216001510620117,
+      "learning_rate": 4.6008130821385025e-06,
+      "loss": 1.7697,
+      "step": 22460
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 0.9190794825553894,
+      "learning_rate": 4.592688185567023e-06,
+      "loss": 1.9444,
+      "step": 22465
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 0.6193316578865051,
+      "learning_rate": 4.58456974380864e-06,
+      "loss": 1.942,
+      "step": 22470
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 0.9209474921226501,
+      "learning_rate": 4.576457759431216e-06,
+      "loss": 1.99,
+      "step": 22475
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 0.9471963047981262,
+      "learning_rate": 4.568352235000553e-06,
+      "loss": 1.6965,
+      "step": 22480
+    },
+    {
+      "epoch": 4.02,
+      "grad_norm": 1.2862259149551392,
+      "learning_rate": 4.560253173080423e-06,
+      "loss": 1.7552,
+      "step": 22485
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 0.8248801231384277,
+      "learning_rate": 4.552160576232553e-06,
+      "loss": 1.9429,
+      "step": 22490
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 0.5932010412216187,
+      "learning_rate": 4.544074447016628e-06,
+      "loss": 1.6949,
+      "step": 22495
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 1.4545879364013672,
+      "learning_rate": 4.535994787990283e-06,
+      "loss": 1.9368,
+      "step": 22500
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 0.8596073985099792,
+      "learning_rate": 4.527921601709104e-06,
+      "loss": 1.8986,
+      "step": 22505
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 1.0523921251296997,
+      "learning_rate": 4.519854890726624e-06,
+      "loss": 1.7516,
+      "step": 22510
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 0.9680181741714478,
+      "learning_rate": 4.511794657594343e-06,
+      "loss": 1.6559,
+      "step": 22515
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 1.0432225465774536,
+      "learning_rate": 4.503740904861703e-06,
+      "loss": 1.6328,
+      "step": 22520
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 1.3763519525527954,
+      "learning_rate": 4.495693635076101e-06,
+      "loss": 1.6849,
+      "step": 22525
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 0.7782979607582092,
+      "learning_rate": 4.487652850782886e-06,
+      "loss": 1.6029,
+      "step": 22530
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 0.760043203830719,
+      "learning_rate": 4.479618554525339e-06,
+      "loss": 1.9747,
+      "step": 22535
+    },
+    {
+      "epoch": 4.03,
+      "grad_norm": 0.7995366454124451,
+      "learning_rate": 4.471590748844703e-06,
+      "loss": 1.8518,
+      "step": 22540
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 1.1295702457427979,
+      "learning_rate": 4.4635694362801684e-06,
+      "loss": 2.1671,
+      "step": 22545
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 0.6298494338989258,
+      "learning_rate": 4.4555546193688735e-06,
+      "loss": 1.894,
+      "step": 22550
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 0.6293065547943115,
+      "learning_rate": 4.4475463006458855e-06,
+      "loss": 1.6044,
+      "step": 22555
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 1.1865392923355103,
+      "learning_rate": 4.4395444826442395e-06,
+      "loss": 1.8025,
+      "step": 22560
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 0.8114657998085022,
+      "learning_rate": 4.431549167894891e-06,
+      "loss": 1.973,
+      "step": 22565
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 0.7298006415367126,
+      "learning_rate": 4.423560358926756e-06,
+      "loss": 1.809,
+      "step": 22570
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 1.246186375617981,
+      "learning_rate": 4.415578058266687e-06,
+      "loss": 1.733,
+      "step": 22575
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 0.7696637511253357,
+      "learning_rate": 4.407602268439479e-06,
+      "loss": 1.481,
+      "step": 22580
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 0.4412114918231964,
+      "learning_rate": 4.3996329919678666e-06,
+      "loss": 2.2101,
+      "step": 22585
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 0.5162980556488037,
+      "learning_rate": 4.3916702313725276e-06,
+      "loss": 2.1751,
+      "step": 22590
+    },
+    {
+      "epoch": 4.04,
+      "grad_norm": 1.8025277853012085,
+      "learning_rate": 4.3837139891720755e-06,
+      "loss": 1.7814,
+      "step": 22595
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 0.36622488498687744,
+      "learning_rate": 4.375764267883048e-06,
+      "loss": 2.1449,
+      "step": 22600
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 1.1410011053085327,
+      "learning_rate": 4.367821070019948e-06,
+      "loss": 1.7696,
+      "step": 22605
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 0.6294861435890198,
+      "learning_rate": 4.359884398095196e-06,
+      "loss": 1.8894,
+      "step": 22610
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 1.1744203567504883,
+      "learning_rate": 4.351954254619156e-06,
+      "loss": 1.687,
+      "step": 22615
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 0.8127630949020386,
+      "learning_rate": 4.344030642100133e-06,
+      "loss": 1.8386,
+      "step": 22620
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 0.8806772828102112,
+      "learning_rate": 4.336113563044339e-06,
+      "loss": 1.6666,
+      "step": 22625
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 1.0629003047943115,
+      "learning_rate": 4.328203019955951e-06,
+      "loss": 1.8621,
+      "step": 22630
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 0.7076244354248047,
+      "learning_rate": 4.320299015337065e-06,
+      "loss": 1.7092,
+      "step": 22635
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 1.0784873962402344,
+      "learning_rate": 4.312401551687714e-06,
+      "loss": 1.6767,
+      "step": 22640
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 1.1845802068710327,
+      "learning_rate": 4.304510631505848e-06,
+      "loss": 1.7707,
+      "step": 22645
+    },
+    {
+      "epoch": 4.05,
+      "grad_norm": 0.7437544465065002,
+      "learning_rate": 4.296626257287367e-06,
+      "loss": 1.8259,
+      "step": 22650
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 1.0520071983337402,
+      "learning_rate": 4.288748431526082e-06,
+      "loss": 1.9545,
+      "step": 22655
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 0.893604040145874,
+      "learning_rate": 4.280877156713748e-06,
+      "loss": 1.7078,
+      "step": 22660
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 3.4858219623565674,
+      "learning_rate": 4.273012435340038e-06,
+      "loss": 2.1413,
+      "step": 22665
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 0.628693163394928,
+      "learning_rate": 4.265154269892557e-06,
+      "loss": 1.5482,
+      "step": 22670
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 1.399335265159607,
+      "learning_rate": 4.257302662856838e-06,
+      "loss": 1.751,
+      "step": 22675
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 0.8125795722007751,
+      "learning_rate": 4.24945761671634e-06,
+      "loss": 1.9325,
+      "step": 22680
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 1.0702699422836304,
+      "learning_rate": 4.241619133952432e-06,
+      "loss": 1.8058,
+      "step": 22685
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 0.5640363097190857,
+      "learning_rate": 4.23378721704443e-06,
+      "loss": 1.8514,
+      "step": 22690
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 0.6807628273963928,
+      "learning_rate": 4.225961868469553e-06,
+      "loss": 1.6916,
+      "step": 22695
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 0.7918212413787842,
+      "learning_rate": 4.218143090702953e-06,
+      "loss": 1.8782,
+      "step": 22700
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 0.6629428863525391,
+      "learning_rate": 4.210330886217706e-06,
+      "loss": 2.0677,
+      "step": 22705
+    },
+    {
+      "epoch": 4.06,
+      "grad_norm": 0.6261160373687744,
+      "learning_rate": 4.202525257484807e-06,
+      "loss": 1.8922,
+      "step": 22710
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 0.9577205181121826,
+      "learning_rate": 4.194726206973157e-06,
+      "loss": 1.6897,
+      "step": 22715
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 1.4098031520843506,
+      "learning_rate": 4.186933737149598e-06,
+      "loss": 1.9442,
+      "step": 22720
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 1.430624008178711,
+      "learning_rate": 4.179147850478876e-06,
+      "loss": 1.9234,
+      "step": 22725
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 1.2811557054519653,
+      "learning_rate": 4.171368549423665e-06,
+      "loss": 1.9003,
+      "step": 22730
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 0.9562197923660278,
+      "learning_rate": 4.163595836444551e-06,
+      "loss": 1.8844,
+      "step": 22735
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 0.8576022982597351,
+      "learning_rate": 4.1558297140000314e-06,
+      "loss": 1.6477,
+      "step": 22740
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 2.603815793991089,
+      "learning_rate": 4.1480701845465205e-06,
+      "loss": 1.988,
+      "step": 22745
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 0.5836585760116577,
+      "learning_rate": 4.14031725053835e-06,
+      "loss": 2.0429,
+      "step": 22750
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 1.4115853309631348,
+      "learning_rate": 4.132570914427772e-06,
+      "loss": 2.0125,
+      "step": 22755
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 0.5206339359283447,
+      "learning_rate": 4.12483117866494e-06,
+      "loss": 1.9237,
+      "step": 22760
+    },
+    {
+      "epoch": 4.07,
+      "grad_norm": 1.0422269105911255,
+      "learning_rate": 4.117098045697931e-06,
+      "loss": 1.6165,
+      "step": 22765
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 1.0876985788345337,
+      "learning_rate": 4.109371517972721e-06,
+      "loss": 1.6538,
+      "step": 22770
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 0.7427853345870972,
+      "learning_rate": 4.101651597933204e-06,
+      "loss": 2.3134,
+      "step": 22775
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 1.1198269128799438,
+      "learning_rate": 4.09393828802119e-06,
+      "loss": 1.8009,
+      "step": 22780
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 0.47877076268196106,
+      "learning_rate": 4.086231590676382e-06,
+      "loss": 1.9703,
+      "step": 22785
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 0.6987916231155396,
+      "learning_rate": 4.078531508336403e-06,
+      "loss": 1.4546,
+      "step": 22790
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 0.5703626275062561,
+      "learning_rate": 4.070838043436786e-06,
+      "loss": 1.911,
+      "step": 22795
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 1.0218112468719482,
+      "learning_rate": 4.063151198410969e-06,
+      "loss": 1.8741,
+      "step": 22800
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 0.6620162725448608,
+      "learning_rate": 4.055470975690284e-06,
+      "loss": 1.8318,
+      "step": 22805
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 2.1791741847991943,
+      "learning_rate": 4.047797377703985e-06,
+      "loss": 1.6375,
+      "step": 22810
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 0.8899803757667542,
+      "learning_rate": 4.040130406879219e-06,
+      "loss": 1.7384,
+      "step": 22815
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 0.5828153491020203,
+      "learning_rate": 4.032470065641047e-06,
+      "loss": 1.7185,
+      "step": 22820
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 0.5616529583930969,
+      "learning_rate": 4.024816356412431e-06,
+      "loss": 1.7911,
+      "step": 22825
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 0.8453607559204102,
+      "learning_rate": 4.017169281614225e-06,
+      "loss": 1.6574,
+      "step": 22830
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 0.6679532527923584,
+      "learning_rate": 4.009528843665186e-06,
+      "loss": 1.7073,
+      "step": 22835
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 1.1486124992370605,
+      "learning_rate": 4.001895044981982e-06,
+      "loss": 1.7303,
+      "step": 22840
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 0.7910205721855164,
+      "learning_rate": 3.99426788797918e-06,
+      "loss": 1.9838,
+      "step": 22845
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 1.2929234504699707,
+      "learning_rate": 3.986647375069241e-06,
+      "loss": 1.8347,
+      "step": 22850
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 0.7869713306427002,
+      "learning_rate": 3.9790335086625275e-06,
+      "loss": 1.8313,
+      "step": 22855
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 1.0830061435699463,
+      "learning_rate": 3.971426291167294e-06,
+      "loss": 1.9421,
+      "step": 22860
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 1.3566019535064697,
+      "learning_rate": 3.963825724989695e-06,
+      "loss": 1.8847,
+      "step": 22865
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 0.9219987988471985,
+      "learning_rate": 3.956231812533784e-06,
+      "loss": 2.0614,
+      "step": 22870
+    },
+    {
+      "epoch": 4.09,
+      "grad_norm": 1.1585674285888672,
+      "learning_rate": 3.948644556201517e-06,
+      "loss": 2.011,
+      "step": 22875
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 1.1043351888656616,
+      "learning_rate": 3.941063958392721e-06,
+      "loss": 2.111,
+      "step": 22880
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 0.9520683884620667,
+      "learning_rate": 3.93349002150514e-06,
+      "loss": 1.7933,
+      "step": 22885
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 1.2550685405731201,
+      "learning_rate": 3.925922747934407e-06,
+      "loss": 1.7792,
+      "step": 22890
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 0.6732292175292969,
+      "learning_rate": 3.918362140074031e-06,
+      "loss": 1.9497,
+      "step": 22895
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 1.17646324634552,
+      "learning_rate": 3.9108082003154325e-06,
+      "loss": 1.7155,
+      "step": 22900
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 0.677506148815155,
+      "learning_rate": 3.903260931047917e-06,
+      "loss": 1.7703,
+      "step": 22905
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 1.1620838642120361,
+      "learning_rate": 3.895720334658676e-06,
+      "loss": 1.6102,
+      "step": 22910
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 0.9558453559875488,
+      "learning_rate": 3.888186413532799e-06,
+      "loss": 1.9861,
+      "step": 22915
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 0.9614786505699158,
+      "learning_rate": 3.880659170053253e-06,
+      "loss": 1.8458,
+      "step": 22920
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 0.40708935260772705,
+      "learning_rate": 3.87313860660089e-06,
+      "loss": 1.845,
+      "step": 22925
+    },
+    {
+      "epoch": 4.1,
+      "grad_norm": 0.8563295006752014,
+      "learning_rate": 3.865624725554465e-06,
+      "loss": 1.598,
+      "step": 22930
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 2.2305939197540283,
+      "learning_rate": 3.8581175292906084e-06,
+      "loss": 1.9656,
+      "step": 22935
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 0.7936108112335205,
+      "learning_rate": 3.850617020183845e-06,
+      "loss": 1.9327,
+      "step": 22940
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 1.0209077596664429,
+      "learning_rate": 3.843123200606577e-06,
+      "loss": 2.0603,
+      "step": 22945
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 1.030806541442871,
+      "learning_rate": 3.835636072929088e-06,
+      "loss": 1.7855,
+      "step": 22950
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 0.9423354864120483,
+      "learning_rate": 3.828155639519551e-06,
+      "loss": 1.9576,
+      "step": 22955
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 1.0258960723876953,
+      "learning_rate": 3.82068190274402e-06,
+      "loss": 1.9906,
+      "step": 22960
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 0.5299312472343445,
+      "learning_rate": 3.81321486496643e-06,
+      "loss": 1.8934,
+      "step": 22965
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 1.3776746988296509,
+      "learning_rate": 3.8057545285486053e-06,
+      "loss": 1.9732,
+      "step": 22970
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 0.8569226264953613,
+      "learning_rate": 3.798300895850232e-06,
+      "loss": 1.8981,
+      "step": 22975
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 0.8730267882347107,
+      "learning_rate": 3.790853969228897e-06,
+      "loss": 2.07,
+      "step": 22980
+    },
+    {
+      "epoch": 4.11,
+      "grad_norm": 0.7692422270774841,
+      "learning_rate": 3.7834137510400465e-06,
+      "loss": 1.5263,
+      "step": 22985
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 0.8507816195487976,
+      "learning_rate": 3.7759802436370174e-06,
+      "loss": 1.905,
+      "step": 22990
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 0.851302444934845,
+      "learning_rate": 3.7685534493710235e-06,
+      "loss": 1.7006,
+      "step": 22995
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 1.0782297849655151,
+      "learning_rate": 3.7611333705911523e-06,
+      "loss": 1.9236,
+      "step": 23000
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 0.5770434141159058,
+      "learning_rate": 3.753720009644371e-06,
+      "loss": 1.9372,
+      "step": 23005
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 0.7802814841270447,
+      "learning_rate": 3.7463133688755066e-06,
+      "loss": 1.8211,
+      "step": 23010
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 1.3625346422195435,
+      "learning_rate": 3.7389134506272903e-06,
+      "loss": 1.8262,
+      "step": 23015
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 0.8293248414993286,
+      "learning_rate": 3.7315202572402896e-06,
+      "loss": 1.7531,
+      "step": 23020
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 0.6201703548431396,
+      "learning_rate": 3.724133791052975e-06,
+      "loss": 1.9715,
+      "step": 23025
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 1.0974987745285034,
+      "learning_rate": 3.7167540544016775e-06,
+      "loss": 2.0178,
+      "step": 23030
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 0.7133528590202332,
+      "learning_rate": 3.709381049620608e-06,
+      "loss": 1.6524,
+      "step": 23035
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 0.9783563017845154,
+      "learning_rate": 3.7020147790418263e-06,
+      "loss": 2.1077,
+      "step": 23040
+    },
+    {
+      "epoch": 4.12,
+      "grad_norm": 0.7139909863471985,
+      "learning_rate": 3.6946552449952847e-06,
+      "loss": 1.8681,
+      "step": 23045
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 2.149613380432129,
+      "learning_rate": 3.6873024498087938e-06,
+      "loss": 1.9458,
+      "step": 23050
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 1.5502854585647583,
+      "learning_rate": 3.67995639580804e-06,
+      "loss": 2.0566,
+      "step": 23055
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 0.7524001598358154,
+      "learning_rate": 3.6726170853165763e-06,
+      "loss": 1.4629,
+      "step": 23060
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 0.4955444633960724,
+      "learning_rate": 3.665284520655807e-06,
+      "loss": 1.5796,
+      "step": 23065
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 0.7171782851219177,
+      "learning_rate": 3.6579587041450286e-06,
+      "loss": 2.1074,
+      "step": 23070
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 0.5848670601844788,
+      "learning_rate": 3.6506396381013804e-06,
+      "loss": 1.8271,
+      "step": 23075
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 1.1107635498046875,
+      "learning_rate": 3.6433273248398763e-06,
+      "loss": 1.6248,
+      "step": 23080
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 1.556101679801941,
+      "learning_rate": 3.6360217666733975e-06,
+      "loss": 1.8316,
+      "step": 23085
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 1.174280047416687,
+      "learning_rate": 3.628722965912687e-06,
+      "loss": 1.6224,
+      "step": 23090
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 0.5544099807739258,
+      "learning_rate": 3.621430924866348e-06,
+      "loss": 2.1943,
+      "step": 23095
+    },
+    {
+      "epoch": 4.13,
+      "grad_norm": 0.6023959517478943,
+      "learning_rate": 3.6141456458408383e-06,
+      "loss": 1.9006,
+      "step": 23100
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 0.9502812027931213,
+      "learning_rate": 3.6068671311404927e-06,
+      "loss": 1.6577,
+      "step": 23105
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 0.8921631574630737,
+      "learning_rate": 3.5995953830675e-06,
+      "loss": 1.1959,
+      "step": 23110
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 0.8794203996658325,
+      "learning_rate": 3.592330403921898e-06,
+      "loss": 1.6737,
+      "step": 23115
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 0.5600360035896301,
+      "learning_rate": 3.585072196001599e-06,
+      "loss": 2.1309,
+      "step": 23120
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 1.8456734418869019,
+      "learning_rate": 3.5778207616023684e-06,
+      "loss": 2.0503,
+      "step": 23125
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 0.7940518856048584,
+      "learning_rate": 3.5705761030178224e-06,
+      "loss": 1.8708,
+      "step": 23130
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 1.4641777276992798,
+      "learning_rate": 3.563338222539442e-06,
+      "loss": 1.7284,
+      "step": 23135
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 0.6386775374412537,
+      "learning_rate": 3.5561071224565617e-06,
+      "loss": 1.8958,
+      "step": 23140
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 0.9825034737586975,
+      "learning_rate": 3.548882805056372e-06,
+      "loss": 1.6577,
+      "step": 23145
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 0.915550172328949,
+      "learning_rate": 3.541665272623923e-06,
+      "loss": 1.6564,
+      "step": 23150
+    },
+    {
+      "epoch": 4.14,
+      "grad_norm": 0.926216185092926,
+      "learning_rate": 3.5344545274421077e-06,
+      "loss": 1.9108,
+      "step": 23155
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 1.014319896697998,
+      "learning_rate": 3.527250571791674e-06,
+      "loss": 1.7595,
+      "step": 23160
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 0.9334921836853027,
+      "learning_rate": 3.5200534079512295e-06,
+      "loss": 1.7663,
+      "step": 23165
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 0.9318943619728088,
+      "learning_rate": 3.512863038197231e-06,
+      "loss": 2.109,
+      "step": 23170
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 0.7537411451339722,
+      "learning_rate": 3.505679464803985e-06,
+      "loss": 1.7133,
+      "step": 23175
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 0.6408186554908752,
+      "learning_rate": 3.498502690043651e-06,
+      "loss": 1.7261,
+      "step": 23180
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 0.8619161248207092,
+      "learning_rate": 3.491332716186238e-06,
+      "loss": 1.7955,
+      "step": 23185
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 0.9069615006446838,
+      "learning_rate": 3.4841695454995936e-06,
+      "loss": 1.9615,
+      "step": 23190
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 0.5599504113197327,
+      "learning_rate": 3.4770131802494278e-06,
+      "loss": 1.8248,
+      "step": 23195
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 1.420801043510437,
+      "learning_rate": 3.469863622699293e-06,
+      "loss": 1.7244,
+      "step": 23200
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 0.7983185052871704,
+      "learning_rate": 3.462720875110584e-06,
+      "loss": 2.053,
+      "step": 23205
+    },
+    {
+      "epoch": 4.15,
+      "grad_norm": 0.9097558856010437,
+      "learning_rate": 3.45558493974254e-06,
+      "loss": 1.8446,
+      "step": 23210
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.5858725905418396,
+      "learning_rate": 3.448455818852267e-06,
+      "loss": 2.001,
+      "step": 23215
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 1.1941709518432617,
+      "learning_rate": 3.4413335146946807e-06,
+      "loss": 1.8848,
+      "step": 23220
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.5961235761642456,
+      "learning_rate": 3.434218029522568e-06,
+      "loss": 1.7481,
+      "step": 23225
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 1.2655341625213623,
+      "learning_rate": 3.4271093655865476e-06,
+      "loss": 1.5217,
+      "step": 23230
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.8132046461105347,
+      "learning_rate": 3.4200075251350826e-06,
+      "loss": 1.5498,
+      "step": 23235
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.48109278082847595,
+      "learning_rate": 3.41291251041449e-06,
+      "loss": 1.7054,
+      "step": 23240
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.6113245487213135,
+      "learning_rate": 3.4058243236688993e-06,
+      "loss": 1.7679,
+      "step": 23245
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 1.200701117515564,
+      "learning_rate": 3.3987429671403088e-06,
+      "loss": 1.8985,
+      "step": 23250
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.7015065550804138,
+      "learning_rate": 3.391668443068535e-06,
+      "loss": 1.9091,
+      "step": 23255
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 1.0896352529525757,
+      "learning_rate": 3.3846007536912473e-06,
+      "loss": 1.5577,
+      "step": 23260
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 1.282090425491333,
+      "learning_rate": 3.3775399012439506e-06,
+      "loss": 1.7736,
+      "step": 23265
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 0.48688602447509766,
+      "learning_rate": 3.3704858879599843e-06,
+      "loss": 1.6715,
+      "step": 23270
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 0.6023693084716797,
+      "learning_rate": 3.3634387160705325e-06,
+      "loss": 1.7387,
+      "step": 23275
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 1.431891679763794,
+      "learning_rate": 3.356398387804596e-06,
+      "loss": 1.6708,
+      "step": 23280
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 0.6251328587532043,
+      "learning_rate": 3.3493649053890326e-06,
+      "loss": 1.8878,
+      "step": 23285
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 1.2962218523025513,
+      "learning_rate": 3.342338271048526e-06,
+      "loss": 1.6701,
+      "step": 23290
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 0.9591766595840454,
+      "learning_rate": 3.3353184870055954e-06,
+      "loss": 1.78,
+      "step": 23295
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 1.259443759918213,
+      "learning_rate": 3.3283055554805848e-06,
+      "loss": 1.6922,
+      "step": 23300
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 1.0876199007034302,
+      "learning_rate": 3.3212994786916836e-06,
+      "loss": 1.9445,
+      "step": 23305
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 2.2998528480529785,
+      "learning_rate": 3.314300258854902e-06,
+      "loss": 1.7383,
+      "step": 23310
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 1.3563592433929443,
+      "learning_rate": 3.307307898184089e-06,
+      "loss": 1.8597,
+      "step": 23315
+    },
+    {
+      "epoch": 4.17,
+      "grad_norm": 0.7580876350402832,
+      "learning_rate": 3.3003223988909234e-06,
+      "loss": 1.7143,
+      "step": 23320
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 0.7381443381309509,
+      "learning_rate": 3.2933437631849085e-06,
+      "loss": 1.8132,
+      "step": 23325
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 0.7825798988342285,
+      "learning_rate": 3.2863719932733904e-06,
+      "loss": 1.8894,
+      "step": 23330
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 0.96683269739151,
+      "learning_rate": 3.2794070913615165e-06,
+      "loss": 1.8698,
+      "step": 23335
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 0.9194793701171875,
+      "learning_rate": 3.2724490596522937e-06,
+      "loss": 1.8079,
+      "step": 23340
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 0.9263105988502502,
+      "learning_rate": 3.2654979003465313e-06,
+      "loss": 1.6324,
+      "step": 23345
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 0.6331471800804138,
+      "learning_rate": 3.258553615642873e-06,
+      "loss": 1.9431,
+      "step": 23350
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 0.6416884064674377,
+      "learning_rate": 3.2516162077377953e-06,
+      "loss": 1.865,
+      "step": 23355
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 0.7806147933006287,
+      "learning_rate": 3.2446856788255958e-06,
+      "loss": 1.8503,
+      "step": 23360
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 0.5219438076019287,
+      "learning_rate": 3.2377620310983873e-06,
+      "loss": 1.8821,
+      "step": 23365
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 0.8223024606704712,
+      "learning_rate": 3.230845266746113e-06,
+      "loss": 1.6572,
+      "step": 23370
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 0.9462054967880249,
+      "learning_rate": 3.223935387956545e-06,
+      "loss": 1.7214,
+      "step": 23375
+    },
+    {
+      "epoch": 4.18,
+      "grad_norm": 0.538049042224884,
+      "learning_rate": 3.217032396915265e-06,
+      "loss": 1.751,
+      "step": 23380
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 0.767585813999176,
+      "learning_rate": 3.210136295805685e-06,
+      "loss": 1.6549,
+      "step": 23385
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 0.9927555322647095,
+      "learning_rate": 3.2032470868090414e-06,
+      "loss": 1.7279,
+      "step": 23390
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 0.8647646903991699,
+      "learning_rate": 3.196364772104379e-06,
+      "loss": 1.6213,
+      "step": 23395
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 0.8548038005828857,
+      "learning_rate": 3.189489353868563e-06,
+      "loss": 2.1221,
+      "step": 23400
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 0.7027987241744995,
+      "learning_rate": 3.182620834276284e-06,
+      "loss": 2.0151,
+      "step": 23405
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 0.8726295828819275,
+      "learning_rate": 3.1757592155000486e-06,
+      "loss": 1.8991,
+      "step": 23410
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 0.5377846360206604,
+      "learning_rate": 3.168904499710182e-06,
+      "loss": 2.0556,
+      "step": 23415
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 0.492355078458786,
+      "learning_rate": 3.162056689074827e-06,
+      "loss": 1.737,
+      "step": 23420
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 0.7007409930229187,
+      "learning_rate": 3.1552157857599324e-06,
+      "loss": 1.9326,
+      "step": 23425
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 0.9527152180671692,
+      "learning_rate": 3.1483817919292726e-06,
+      "loss": 2.0368,
+      "step": 23430
+    },
+    {
+      "epoch": 4.19,
+      "grad_norm": 2.5235965251922607,
+      "learning_rate": 3.1415547097444374e-06,
+      "loss": 1.7703,
+      "step": 23435
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 0.6163821816444397,
+      "learning_rate": 3.134734541364817e-06,
+      "loss": 1.7116,
+      "step": 23440
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 0.6082310080528259,
+      "learning_rate": 3.12792128894763e-06,
+      "loss": 1.7943,
+      "step": 23445
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 0.7631220817565918,
+      "learning_rate": 3.121114954647908e-06,
+      "loss": 2.0513,
+      "step": 23450
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 0.9659873843193054,
+      "learning_rate": 3.114315540618473e-06,
+      "loss": 1.911,
+      "step": 23455
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 0.7039878368377686,
+      "learning_rate": 3.107523049009983e-06,
+      "loss": 1.9678,
+      "step": 23460
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 0.5036177039146423,
+      "learning_rate": 3.1007374819708947e-06,
+      "loss": 1.941,
+      "step": 23465
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 0.522300124168396,
+      "learning_rate": 3.0939588416474762e-06,
+      "loss": 1.882,
+      "step": 23470
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 1.4440438747406006,
+      "learning_rate": 3.0871871301838053e-06,
+      "loss": 1.4968,
+      "step": 23475
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 0.7592721581459045,
+      "learning_rate": 3.0804223497217706e-06,
+      "loss": 1.8058,
+      "step": 23480
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 1.0022876262664795,
+      "learning_rate": 3.0736645024010664e-06,
+      "loss": 1.7875,
+      "step": 23485
+    },
+    {
+      "epoch": 4.2,
+      "grad_norm": 0.6901012063026428,
+      "learning_rate": 3.066913590359183e-06,
+      "loss": 1.7997,
+      "step": 23490
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 0.9948673844337463,
+      "learning_rate": 3.0601696157314315e-06,
+      "loss": 1.9634,
+      "step": 23495
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 0.49334949254989624,
+      "learning_rate": 3.0534325806509307e-06,
+      "loss": 1.8824,
+      "step": 23500
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 0.7687468528747559,
+      "learning_rate": 3.046702487248593e-06,
+      "loss": 1.7471,
+      "step": 23505
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 0.5290055274963379,
+      "learning_rate": 3.0399793376531484e-06,
+      "loss": 1.7987,
+      "step": 23510
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 0.5497667789459229,
+      "learning_rate": 3.03326313399111e-06,
+      "loss": 1.8059,
+      "step": 23515
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 1.2975267171859741,
+      "learning_rate": 3.0265538783868117e-06,
+      "loss": 1.8914,
+      "step": 23520
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 1.4567995071411133,
+      "learning_rate": 3.01985157296239e-06,
+      "loss": 1.7871,
+      "step": 23525
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 1.1147876977920532,
+      "learning_rate": 3.013156219837776e-06,
+      "loss": 1.9195,
+      "step": 23530
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 0.48365309834480286,
+      "learning_rate": 3.006467821130696e-06,
+      "loss": 1.8373,
+      "step": 23535
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 0.6062785983085632,
+      "learning_rate": 2.9997863789566953e-06,
+      "loss": 1.726,
+      "step": 23540
+    },
+    {
+      "epoch": 4.21,
+      "grad_norm": 1.1633416414260864,
+      "learning_rate": 2.993111895429093e-06,
+      "loss": 1.7636,
+      "step": 23545
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 0.9300481677055359,
+      "learning_rate": 2.9864443726590335e-06,
+      "loss": 1.6565,
+      "step": 23550
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 1.1455543041229248,
+      "learning_rate": 2.9797838127554433e-06,
+      "loss": 2.0796,
+      "step": 23555
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 0.33199450373649597,
+      "learning_rate": 2.9731302178250543e-06,
+      "loss": 2.0225,
+      "step": 23560
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 0.6419559717178345,
+      "learning_rate": 2.966483589972394e-06,
+      "loss": 1.9895,
+      "step": 23565
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 0.8773286938667297,
+      "learning_rate": 2.9598439312997745e-06,
+      "loss": 1.6463,
+      "step": 23570
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 1.4193726778030396,
+      "learning_rate": 2.953211243907325e-06,
+      "loss": 1.9775,
+      "step": 23575
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 0.9343559741973877,
+      "learning_rate": 2.946585529892951e-06,
+      "loss": 1.7366,
+      "step": 23580
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 0.9293035268783569,
+      "learning_rate": 2.9399667913523582e-06,
+      "loss": 1.8817,
+      "step": 23585
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 0.7766603827476501,
+      "learning_rate": 2.933355030379048e-06,
+      "loss": 1.904,
+      "step": 23590
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 0.7661356329917908,
+      "learning_rate": 2.9267502490643194e-06,
+      "loss": 1.5144,
+      "step": 23595
+    },
+    {
+      "epoch": 4.22,
+      "grad_norm": 1.3593940734863281,
+      "learning_rate": 2.92015244949726e-06,
+      "loss": 1.9262,
+      "step": 23600
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 1.014451265335083,
+      "learning_rate": 2.9135616337647386e-06,
+      "loss": 1.6614,
+      "step": 23605
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 1.7844657897949219,
+      "learning_rate": 2.906977803951427e-06,
+      "loss": 1.8037,
+      "step": 23610
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 1.5160473585128784,
+      "learning_rate": 2.9004009621397847e-06,
+      "loss": 1.6912,
+      "step": 23615
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 1.6145546436309814,
+      "learning_rate": 2.893831110410067e-06,
+      "loss": 1.9369,
+      "step": 23620
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 1.6145838499069214,
+      "learning_rate": 2.8872682508403e-06,
+      "loss": 1.9497,
+      "step": 23625
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 2.104137659072876,
+      "learning_rate": 2.880712385506318e-06,
+      "loss": 1.916,
+      "step": 23630
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 0.9931949973106384,
+      "learning_rate": 2.8741635164817315e-06,
+      "loss": 1.6568,
+      "step": 23635
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 1.3967357873916626,
+      "learning_rate": 2.8676216458379396e-06,
+      "loss": 1.5115,
+      "step": 23640
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 0.5755307078361511,
+      "learning_rate": 2.861086775644134e-06,
+      "loss": 1.8759,
+      "step": 23645
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 0.9810478091239929,
+      "learning_rate": 2.8545589079672865e-06,
+      "loss": 1.6617,
+      "step": 23650
+    },
+    {
+      "epoch": 4.23,
+      "grad_norm": 0.7798149585723877,
+      "learning_rate": 2.848038044872159e-06,
+      "loss": 1.6842,
+      "step": 23655
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 1.8288719654083252,
+      "learning_rate": 2.8415241884212853e-06,
+      "loss": 1.7013,
+      "step": 23660
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 1.104956030845642,
+      "learning_rate": 2.8350173406749973e-06,
+      "loss": 1.8724,
+      "step": 23665
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 1.367289423942566,
+      "learning_rate": 2.8285175036914107e-06,
+      "loss": 1.7069,
+      "step": 23670
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 1.1730290651321411,
+      "learning_rate": 2.8220246795264057e-06,
+      "loss": 1.851,
+      "step": 23675
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 1.0620871782302856,
+      "learning_rate": 2.815538870233661e-06,
+      "loss": 1.7259,
+      "step": 23680
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 1.1758960485458374,
+      "learning_rate": 2.8090600778646362e-06,
+      "loss": 1.7401,
+      "step": 23685
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 1.1776374578475952,
+      "learning_rate": 2.8025883044685646e-06,
+      "loss": 1.9266,
+      "step": 23690
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 0.8739143013954163,
+      "learning_rate": 2.7961235520924584e-06,
+      "loss": 1.9129,
+      "step": 23695
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 0.8561976552009583,
+      "learning_rate": 2.789665822781115e-06,
+      "loss": 2.0287,
+      "step": 23700
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 0.6630309224128723,
+      "learning_rate": 2.7832151185771093e-06,
+      "loss": 1.8712,
+      "step": 23705
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 0.8686838150024414,
+      "learning_rate": 2.7767714415207896e-06,
+      "loss": 1.9666,
+      "step": 23710
+    },
+    {
+      "epoch": 4.24,
+      "grad_norm": 1.0738399028778076,
+      "learning_rate": 2.770334793650292e-06,
+      "loss": 1.7293,
+      "step": 23715
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 0.6418025493621826,
+      "learning_rate": 2.763905177001519e-06,
+      "loss": 1.5567,
+      "step": 23720
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 0.8317776322364807,
+      "learning_rate": 2.757482593608143e-06,
+      "loss": 2.0659,
+      "step": 23725
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 1.3214927911758423,
+      "learning_rate": 2.751067045501629e-06,
+      "loss": 1.9451,
+      "step": 23730
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 1.1789897680282593,
+      "learning_rate": 2.7446585347112086e-06,
+      "loss": 1.7314,
+      "step": 23735
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 0.8994229435920715,
+      "learning_rate": 2.7382570632638854e-06,
+      "loss": 1.8044,
+      "step": 23740
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 1.0152734518051147,
+      "learning_rate": 2.7318626331844456e-06,
+      "loss": 1.7334,
+      "step": 23745
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 1.6840885877609253,
+      "learning_rate": 2.725475246495432e-06,
+      "loss": 1.9423,
+      "step": 23750
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 1.69658362865448,
+      "learning_rate": 2.7190949052171703e-06,
+      "loss": 1.6838,
+      "step": 23755
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 0.9725009799003601,
+      "learning_rate": 2.7127216113677635e-06,
+      "loss": 1.9143,
+      "step": 23760
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 0.49594831466674805,
+      "learning_rate": 2.7063553669630702e-06,
+      "loss": 1.8105,
+      "step": 23765
+    },
+    {
+      "epoch": 4.25,
+      "grad_norm": 1.0006866455078125,
+      "learning_rate": 2.6999961740167305e-06,
+      "loss": 1.8241,
+      "step": 23770
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 0.8839683532714844,
+      "learning_rate": 2.6936440345401493e-06,
+      "loss": 1.7172,
+      "step": 23775
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 0.9161496162414551,
+      "learning_rate": 2.6872989505425105e-06,
+      "loss": 1.8713,
+      "step": 23780
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 0.7400037050247192,
+      "learning_rate": 2.6809609240307456e-06,
+      "loss": 1.6108,
+      "step": 23785
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 1.0765010118484497,
+      "learning_rate": 2.6746299570095722e-06,
+      "loss": 2.0275,
+      "step": 23790
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 1.344102144241333,
+      "learning_rate": 2.66830605148147e-06,
+      "loss": 1.6618,
+      "step": 23795
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 2.3063645362854004,
+      "learning_rate": 2.661989209446683e-06,
+      "loss": 1.5832,
+      "step": 23800
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 0.9692860245704651,
+      "learning_rate": 2.655679432903227e-06,
+      "loss": 1.9494,
+      "step": 23805
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 1.3169384002685547,
+      "learning_rate": 2.649376723846875e-06,
+      "loss": 1.7188,
+      "step": 23810
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 0.763640820980072,
+      "learning_rate": 2.6430810842711595e-06,
+      "loss": 1.8531,
+      "step": 23815
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 0.6789270639419556,
+      "learning_rate": 2.636792516167394e-06,
+      "loss": 1.6805,
+      "step": 23820
+    },
+    {
+      "epoch": 4.26,
+      "grad_norm": 2.9606266021728516,
+      "learning_rate": 2.6305110215246455e-06,
+      "loss": 1.6875,
+      "step": 23825
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 0.8681931495666504,
+      "learning_rate": 2.624236602329744e-06,
+      "loss": 1.8095,
+      "step": 23830
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 0.6462576985359192,
+      "learning_rate": 2.6179692605672877e-06,
+      "loss": 1.8016,
+      "step": 23835
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 0.38575172424316406,
+      "learning_rate": 2.6117089982196197e-06,
+      "loss": 1.9387,
+      "step": 23840
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 0.7060482501983643,
+      "learning_rate": 2.6054558172668607e-06,
+      "loss": 1.691,
+      "step": 23845
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 0.779434859752655,
+      "learning_rate": 2.5992097196868847e-06,
+      "loss": 1.6565,
+      "step": 23850
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 4.394678115844727,
+      "learning_rate": 2.5929707074553363e-06,
+      "loss": 1.7052,
+      "step": 23855
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 0.5779292583465576,
+      "learning_rate": 2.5867387825455917e-06,
+      "loss": 1.861,
+      "step": 23860
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 0.7862660884857178,
+      "learning_rate": 2.580513946928817e-06,
+      "loss": 1.8878,
+      "step": 23865
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 1.2498642206192017,
+      "learning_rate": 2.5742962025739115e-06,
+      "loss": 1.8267,
+      "step": 23870
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 0.6954331398010254,
+      "learning_rate": 2.5680855514475466e-06,
+      "loss": 1.8246,
+      "step": 23875
+    },
+    {
+      "epoch": 4.27,
+      "grad_norm": 0.8826925754547119,
+      "learning_rate": 2.5618819955141456e-06,
+      "loss": 1.7513,
+      "step": 23880
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 1.086248517036438,
+      "learning_rate": 2.5556855367358857e-06,
+      "loss": 1.6789,
+      "step": 23885
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 0.739668071269989,
+      "learning_rate": 2.549496177072702e-06,
+      "loss": 1.6101,
+      "step": 23890
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 1.1162097454071045,
+      "learning_rate": 2.5433139184822876e-06,
+      "loss": 1.8692,
+      "step": 23895
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 0.627552330493927,
+      "learning_rate": 2.53713876292008e-06,
+      "loss": 1.9747,
+      "step": 23900
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 1.044128179550171,
+      "learning_rate": 2.5309707123392727e-06,
+      "loss": 1.7328,
+      "step": 23905
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 1.0667206048965454,
+      "learning_rate": 2.524809768690814e-06,
+      "loss": 1.8482,
+      "step": 23910
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 0.584099292755127,
+      "learning_rate": 2.5186559339234085e-06,
+      "loss": 1.7584,
+      "step": 23915
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 1.1192121505737305,
+      "learning_rate": 2.5125092099835047e-06,
+      "loss": 1.5993,
+      "step": 23920
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 0.9891035556793213,
+      "learning_rate": 2.506369598815314e-06,
+      "loss": 1.6044,
+      "step": 23925
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 0.9022369384765625,
+      "learning_rate": 2.500237102360778e-06,
+      "loss": 1.5102,
+      "step": 23930
+    },
+    {
+      "epoch": 4.28,
+      "grad_norm": 1.2190988063812256,
+      "learning_rate": 2.4941117225596054e-06,
+      "loss": 1.8696,
+      "step": 23935
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 1.111484408378601,
+      "learning_rate": 2.4879934613492444e-06,
+      "loss": 1.9025,
+      "step": 23940
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 1.1540158987045288,
+      "learning_rate": 2.4818823206649024e-06,
+      "loss": 1.9573,
+      "step": 23945
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 0.7060664892196655,
+      "learning_rate": 2.475778302439524e-06,
+      "loss": 1.7662,
+      "step": 23950
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 1.312042474746704,
+      "learning_rate": 2.469681408603802e-06,
+      "loss": 1.7862,
+      "step": 23955
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 0.7066253423690796,
+      "learning_rate": 2.463591641086177e-06,
+      "loss": 1.8912,
+      "step": 23960
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 0.7877931594848633,
+      "learning_rate": 2.4575090018128383e-06,
+      "loss": 1.8976,
+      "step": 23965
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 1.185813307762146,
+      "learning_rate": 2.4514334927077167e-06,
+      "loss": 1.9366,
+      "step": 23970
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 1.511576533317566,
+      "learning_rate": 2.445365115692494e-06,
+      "loss": 1.7834,
+      "step": 23975
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 0.49797123670578003,
+      "learning_rate": 2.4393038726865907e-06,
+      "loss": 1.9477,
+      "step": 23980
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 0.5764663815498352,
+      "learning_rate": 2.4332497656071764e-06,
+      "loss": 1.9526,
+      "step": 23985
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 0.518380880355835,
+      "learning_rate": 2.427202796369152e-06,
+      "loss": 1.6421,
+      "step": 23990
+    },
+    {
+      "epoch": 4.29,
+      "grad_norm": 0.8762568831443787,
+      "learning_rate": 2.421162966885171e-06,
+      "loss": 1.8422,
+      "step": 23995
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 0.7528052926063538,
+      "learning_rate": 2.415130279065625e-06,
+      "loss": 1.9284,
+      "step": 24000
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 1.1197984218597412,
+      "learning_rate": 2.409104734818646e-06,
+      "loss": 1.8608,
+      "step": 24005
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 0.6748790144920349,
+      "learning_rate": 2.4030863360501087e-06,
+      "loss": 1.6257,
+      "step": 24010
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 0.5878778100013733,
+      "learning_rate": 2.397075084663636e-06,
+      "loss": 1.884,
+      "step": 24015
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 0.6087291240692139,
+      "learning_rate": 2.391070982560564e-06,
+      "loss": 1.8711,
+      "step": 24020
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 0.7970364689826965,
+      "learning_rate": 2.3850740316399946e-06,
+      "loss": 2.0835,
+      "step": 24025
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 1.3959071636199951,
+      "learning_rate": 2.3790842337987564e-06,
+      "loss": 1.6191,
+      "step": 24030
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 0.6767655611038208,
+      "learning_rate": 2.3731015909314176e-06,
+      "loss": 1.6902,
+      "step": 24035
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 0.8441483378410339,
+      "learning_rate": 2.3671261049302865e-06,
+      "loss": 2.0472,
+      "step": 24040
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 1.3324369192123413,
+      "learning_rate": 2.3611577776853966e-06,
+      "loss": 1.7657,
+      "step": 24045
+    },
+    {
+      "epoch": 4.3,
+      "grad_norm": 1.7804863452911377,
+      "learning_rate": 2.3551966110845203e-06,
+      "loss": 1.7996,
+      "step": 24050
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 1.5548887252807617,
+      "learning_rate": 2.3492426070131747e-06,
+      "loss": 1.9328,
+      "step": 24055
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 0.956676185131073,
+      "learning_rate": 2.3432957673546087e-06,
+      "loss": 1.9565,
+      "step": 24060
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 0.9713985919952393,
+      "learning_rate": 2.3373560939897954e-06,
+      "loss": 1.65,
+      "step": 24065
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 2.308432102203369,
+      "learning_rate": 2.3314235887974547e-06,
+      "loss": 1.9528,
+      "step": 24070
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 1.1004524230957031,
+      "learning_rate": 2.3254982536540267e-06,
+      "loss": 1.617,
+      "step": 24075
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 0.641258180141449,
+      "learning_rate": 2.3195800904336874e-06,
+      "loss": 1.6639,
+      "step": 24080
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 1.2109898328781128,
+      "learning_rate": 2.3136691010083535e-06,
+      "loss": 1.8781,
+      "step": 24085
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 0.9537338614463806,
+      "learning_rate": 2.3077652872476624e-06,
+      "loss": 1.925,
+      "step": 24090
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 0.7373508214950562,
+      "learning_rate": 2.301868651018982e-06,
+      "loss": 1.8247,
+      "step": 24095
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 1.0468422174453735,
+      "learning_rate": 2.2959791941874143e-06,
+      "loss": 1.6905,
+      "step": 24100
+    },
+    {
+      "epoch": 4.31,
+      "grad_norm": 0.8555766940116882,
+      "learning_rate": 2.2900969186157916e-06,
+      "loss": 2.035,
+      "step": 24105
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 1.271920084953308,
+      "learning_rate": 2.2842218261646677e-06,
+      "loss": 1.887,
+      "step": 24110
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 1.4928789138793945,
+      "learning_rate": 2.2783539186923274e-06,
+      "loss": 1.9741,
+      "step": 24115
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 1.0468617677688599,
+      "learning_rate": 2.272493198054787e-06,
+      "loss": 1.6921,
+      "step": 24120
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.6943525075912476,
+      "learning_rate": 2.2666396661057907e-06,
+      "loss": 2.0084,
+      "step": 24125
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.5706409811973572,
+      "learning_rate": 2.2607933246968027e-06,
+      "loss": 2.0996,
+      "step": 24130
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 1.1674983501434326,
+      "learning_rate": 2.2549541756770156e-06,
+      "loss": 1.8062,
+      "step": 24135
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 1.248380422592163,
+      "learning_rate": 2.2491222208933376e-06,
+      "loss": 1.8164,
+      "step": 24140
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 1.2721612453460693,
+      "learning_rate": 2.243297462190419e-06,
+      "loss": 1.6068,
+      "step": 24145
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 1.1479939222335815,
+      "learning_rate": 2.2374799014106246e-06,
+      "loss": 1.8664,
+      "step": 24150
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.7115007638931274,
+      "learning_rate": 2.2316695403940433e-06,
+      "loss": 2.029,
+      "step": 24155
+    },
+    {
+      "epoch": 4.32,
+      "grad_norm": 0.615996778011322,
+      "learning_rate": 2.2258663809784892e-06,
+      "loss": 1.8326,
+      "step": 24160
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 0.7172175049781799,
+      "learning_rate": 2.2200704249994874e-06,
+      "loss": 1.791,
+      "step": 24165
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 0.530174732208252,
+      "learning_rate": 2.2142816742903e-06,
+      "loss": 1.8528,
+      "step": 24170
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 0.49278607964515686,
+      "learning_rate": 2.208500130681901e-06,
+      "loss": 1.9395,
+      "step": 24175
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 0.946496307849884,
+      "learning_rate": 2.2027257960029914e-06,
+      "loss": 1.7902,
+      "step": 24180
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 0.7843234539031982,
+      "learning_rate": 2.1969586720799852e-06,
+      "loss": 1.9402,
+      "step": 24185
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 1.0197824239730835,
+      "learning_rate": 2.1911987607370123e-06,
+      "loss": 1.7697,
+      "step": 24190
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 0.8750340342521667,
+      "learning_rate": 2.18544606379594e-06,
+      "loss": 1.9873,
+      "step": 24195
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 1.0324968099594116,
+      "learning_rate": 2.1797005830763246e-06,
+      "loss": 1.8219,
+      "step": 24200
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 1.028013825416565,
+      "learning_rate": 2.1739623203954674e-06,
+      "loss": 1.7537,
+      "step": 24205
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 1.2253060340881348,
+      "learning_rate": 2.1682312775683723e-06,
+      "loss": 1.8392,
+      "step": 24210
+    },
+    {
+      "epoch": 4.33,
+      "grad_norm": 3.8781888484954834,
+      "learning_rate": 2.1625074564077612e-06,
+      "loss": 1.7971,
+      "step": 24215
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 0.5331137776374817,
+      "learning_rate": 2.156790858724078e-06,
+      "loss": 1.9863,
+      "step": 24220
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 0.9256309270858765,
+      "learning_rate": 2.1510814863254737e-06,
+      "loss": 2.0353,
+      "step": 24225
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 0.932483971118927,
+      "learning_rate": 2.1453793410178196e-06,
+      "loss": 1.7526,
+      "step": 24230
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 0.6007688641548157,
+      "learning_rate": 2.1396844246046903e-06,
+      "loss": 1.6332,
+      "step": 24235
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 0.9822659492492676,
+      "learning_rate": 2.1339967388873895e-06,
+      "loss": 1.3974,
+      "step": 24240
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 0.8610334992408752,
+      "learning_rate": 2.1283162856649273e-06,
+      "loss": 2.0177,
+      "step": 24245
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 0.5514267086982727,
+      "learning_rate": 2.1226430667340285e-06,
+      "loss": 1.8909,
+      "step": 24250
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 0.7719696164131165,
+      "learning_rate": 2.1169770838891155e-06,
+      "loss": 1.6996,
+      "step": 24255
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 0.8033381700515747,
+      "learning_rate": 2.111318338922344e-06,
+      "loss": 1.7499,
+      "step": 24260
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 1.272058367729187,
+      "learning_rate": 2.1056668336235622e-06,
+      "loss": 1.4928,
+      "step": 24265
+    },
+    {
+      "epoch": 4.34,
+      "grad_norm": 1.0982578992843628,
+      "learning_rate": 2.100022569780341e-06,
+      "loss": 1.8814,
+      "step": 24270
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 0.8227397799491882,
+      "learning_rate": 2.09438554917796e-06,
+      "loss": 1.7183,
+      "step": 24275
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 1.012026071548462,
+      "learning_rate": 2.0887557735993908e-06,
+      "loss": 1.7103,
+      "step": 24280
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 1.1738848686218262,
+      "learning_rate": 2.0831332448253386e-06,
+      "loss": 1.6326,
+      "step": 24285
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 0.9276083707809448,
+      "learning_rate": 2.077517964634193e-06,
+      "loss": 1.7646,
+      "step": 24290
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 3.0199804306030273,
+      "learning_rate": 2.071909934802066e-06,
+      "loss": 1.9786,
+      "step": 24295
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 0.6301084756851196,
+      "learning_rate": 2.066309157102772e-06,
+      "loss": 1.6944,
+      "step": 24300
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 1.1437808275222778,
+      "learning_rate": 2.0607156333078332e-06,
+      "loss": 1.9109,
+      "step": 24305
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 0.41809070110321045,
+      "learning_rate": 2.0551293651864783e-06,
+      "loss": 2.0125,
+      "step": 24310
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 3.662910223007202,
+      "learning_rate": 2.0495503545056303e-06,
+      "loss": 1.8321,
+      "step": 24315
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 1.1465100049972534,
+      "learning_rate": 2.0439786030299306e-06,
+      "loss": 1.7392,
+      "step": 24320
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 1.2465860843658447,
+      "learning_rate": 2.038414112521722e-06,
+      "loss": 1.7251,
+      "step": 24325
+    },
+    {
+      "epoch": 4.35,
+      "grad_norm": 0.9465045928955078,
+      "learning_rate": 2.032856884741041e-06,
+      "loss": 1.8722,
+      "step": 24330
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 0.7800765633583069,
+      "learning_rate": 2.027306921445632e-06,
+      "loss": 1.719,
+      "step": 24335
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 0.5892726182937622,
+      "learning_rate": 2.021764224390954e-06,
+      "loss": 1.4895,
+      "step": 24340
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 1.069069504737854,
+      "learning_rate": 2.016228795330144e-06,
+      "loss": 1.7287,
+      "step": 24345
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 0.602766215801239,
+      "learning_rate": 2.0107006360140597e-06,
+      "loss": 1.8655,
+      "step": 24350
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 1.3404914140701294,
+      "learning_rate": 2.005179748191249e-06,
+      "loss": 1.8222,
+      "step": 24355
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 0.8859615325927734,
+      "learning_rate": 1.9996661336079693e-06,
+      "loss": 2.0234,
+      "step": 24360
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 0.8665515780448914,
+      "learning_rate": 1.9941597940081687e-06,
+      "loss": 1.6752,
+      "step": 24365
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 0.8328173160552979,
+      "learning_rate": 1.988660731133499e-06,
+      "loss": 1.7841,
+      "step": 24370
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 0.7565486431121826,
+      "learning_rate": 1.9831689467233015e-06,
+      "loss": 2.0957,
+      "step": 24375
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 1.5388751029968262,
+      "learning_rate": 1.977684442514624e-06,
+      "loss": 2.0002,
+      "step": 24380
+    },
+    {
+      "epoch": 4.36,
+      "grad_norm": 1.155758261680603,
+      "learning_rate": 1.9722072202422144e-06,
+      "loss": 2.0851,
+      "step": 24385
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 0.5179358720779419,
+      "learning_rate": 1.9667372816385114e-06,
+      "loss": 2.1737,
+      "step": 24390
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 0.9456248879432678,
+      "learning_rate": 1.9612746284336513e-06,
+      "loss": 1.7682,
+      "step": 24395
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 1.9018527269363403,
+      "learning_rate": 1.955819262355471e-06,
+      "loss": 1.6936,
+      "step": 24400
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 0.984753429889679,
+      "learning_rate": 1.9503711851294846e-06,
+      "loss": 1.6365,
+      "step": 24405
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 0.8661744594573975,
+      "learning_rate": 1.9449303984789253e-06,
+      "loss": 1.7751,
+      "step": 24410
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 2.2209432125091553,
+      "learning_rate": 1.9394969041247074e-06,
+      "loss": 1.8097,
+      "step": 24415
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 0.5390074253082275,
+      "learning_rate": 1.9340707037854316e-06,
+      "loss": 1.8734,
+      "step": 24420
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 0.6612688899040222,
+      "learning_rate": 1.9286517991774085e-06,
+      "loss": 1.8623,
+      "step": 24425
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 1.4000182151794434,
+      "learning_rate": 1.9232401920146306e-06,
+      "loss": 1.5901,
+      "step": 24430
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 0.7332963943481445,
+      "learning_rate": 1.9178358840087814e-06,
+      "loss": 1.6914,
+      "step": 24435
+    },
+    {
+      "epoch": 4.37,
+      "grad_norm": 1.1628941297531128,
+      "learning_rate": 1.912438876869238e-06,
+      "loss": 1.7781,
+      "step": 24440
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 2.224209785461426,
+      "learning_rate": 1.9070491723030715e-06,
+      "loss": 1.4185,
+      "step": 24445
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 0.7662902474403381,
+      "learning_rate": 1.90166677201504e-06,
+      "loss": 1.845,
+      "step": 24450
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 1.3199970722198486,
+      "learning_rate": 1.8962916777075928e-06,
+      "loss": 1.8531,
+      "step": 24455
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 0.7640072107315063,
+      "learning_rate": 1.890923891080862e-06,
+      "loss": 2.048,
+      "step": 24460
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 0.6150866746902466,
+      "learning_rate": 1.8855634138326806e-06,
+      "loss": 1.8992,
+      "step": 24465
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 1.9289946556091309,
+      "learning_rate": 1.8802102476585537e-06,
+      "loss": 1.9587,
+      "step": 24470
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 0.9716728925704956,
+      "learning_rate": 1.874864394251688e-06,
+      "loss": 1.6301,
+      "step": 24475
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 0.83192378282547,
+      "learning_rate": 1.8695258553029699e-06,
+      "loss": 1.9311,
+      "step": 24480
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 1.113206148147583,
+      "learning_rate": 1.8641946325009735e-06,
+      "loss": 1.419,
+      "step": 24485
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 0.6721107363700867,
+      "learning_rate": 1.8588707275319667e-06,
+      "loss": 1.9316,
+      "step": 24490
+    },
+    {
+      "epoch": 4.38,
+      "grad_norm": 0.9275158643722534,
+      "learning_rate": 1.8535541420798858e-06,
+      "loss": 2.1295,
+      "step": 24495
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 0.5885508060455322,
+      "learning_rate": 1.8482448778263662e-06,
+      "loss": 1.882,
+      "step": 24500
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 0.7526519894599915,
+      "learning_rate": 1.8429429364507228e-06,
+      "loss": 1.9953,
+      "step": 24505
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 0.7567302584648132,
+      "learning_rate": 1.837648319629956e-06,
+      "loss": 1.769,
+      "step": 24510
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 0.6800612211227417,
+      "learning_rate": 1.8323610290387454e-06,
+      "loss": 1.8711,
+      "step": 24515
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 0.9701638221740723,
+      "learning_rate": 1.827081066349459e-06,
+      "loss": 1.8497,
+      "step": 24520
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 1.5889774560928345,
+      "learning_rate": 1.821808433232139e-06,
+      "loss": 2.0364,
+      "step": 24525
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 0.7154751420021057,
+      "learning_rate": 1.8165431313545144e-06,
+      "loss": 1.7296,
+      "step": 24530
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 1.3648035526275635,
+      "learning_rate": 1.811285162382001e-06,
+      "loss": 1.804,
+      "step": 24535
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 0.6310946941375732,
+      "learning_rate": 1.8060345279776848e-06,
+      "loss": 2.1313,
+      "step": 24540
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 0.9399888515472412,
+      "learning_rate": 1.80079122980234e-06,
+      "loss": 1.5419,
+      "step": 24545
+    },
+    {
+      "epoch": 4.39,
+      "grad_norm": 0.7775284647941589,
+      "learning_rate": 1.7955552695144123e-06,
+      "loss": 2.019,
+      "step": 24550
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 2.3950979709625244,
+      "learning_rate": 1.7903266487700382e-06,
+      "loss": 2.064,
+      "step": 24555
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.8698521852493286,
+      "learning_rate": 1.785105369223014e-06,
+      "loss": 1.8614,
+      "step": 24560
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.7981832027435303,
+      "learning_rate": 1.7798914325248328e-06,
+      "loss": 1.7595,
+      "step": 24565
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 1.5428972244262695,
+      "learning_rate": 1.7746848403246559e-06,
+      "loss": 1.869,
+      "step": 24570
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.7430126667022705,
+      "learning_rate": 1.7694855942693239e-06,
+      "loss": 1.8618,
+      "step": 24575
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.7668565511703491,
+      "learning_rate": 1.7642936960033578e-06,
+      "loss": 1.753,
+      "step": 24580
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 2.907975196838379,
+      "learning_rate": 1.7591091471689414e-06,
+      "loss": 1.9419,
+      "step": 24585
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 1.0696461200714111,
+      "learning_rate": 1.7539319494059458e-06,
+      "loss": 1.8617,
+      "step": 24590
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.8900358080863953,
+      "learning_rate": 1.748762104351917e-06,
+      "loss": 1.8093,
+      "step": 24595
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.6597603559494019,
+      "learning_rate": 1.743599613642069e-06,
+      "loss": 1.7068,
+      "step": 24600
+    },
+    {
+      "epoch": 4.4,
+      "grad_norm": 0.9883655309677124,
+      "learning_rate": 1.7384444789092957e-06,
+      "loss": 1.6025,
+      "step": 24605
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 4.697280406951904,
+      "learning_rate": 1.7332967017841623e-06,
+      "loss": 1.8796,
+      "step": 24610
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 1.5211069583892822,
+      "learning_rate": 1.7281562838948966e-06,
+      "loss": 1.7911,
+      "step": 24615
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 0.9767752885818481,
+      "learning_rate": 1.7230232268674124e-06,
+      "loss": 1.7474,
+      "step": 24620
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 1.0923303365707397,
+      "learning_rate": 1.7178975323252937e-06,
+      "loss": 1.6909,
+      "step": 24625
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 0.7364457845687866,
+      "learning_rate": 1.7127792018897914e-06,
+      "loss": 1.9563,
+      "step": 24630
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 0.9624817371368408,
+      "learning_rate": 1.7076682371798298e-06,
+      "loss": 1.9367,
+      "step": 24635
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 0.6353481411933899,
+      "learning_rate": 1.7025646398119988e-06,
+      "loss": 1.8782,
+      "step": 24640
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 1.050934910774231,
+      "learning_rate": 1.6974684114005601e-06,
+      "loss": 1.6196,
+      "step": 24645
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 1.2270129919052124,
+      "learning_rate": 1.692379553557455e-06,
+      "loss": 2.0497,
+      "step": 24650
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 0.5485097765922546,
+      "learning_rate": 1.6872980678922734e-06,
+      "loss": 1.781,
+      "step": 24655
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 0.6211299300193787,
+      "learning_rate": 1.6822239560122882e-06,
+      "loss": 1.8401,
+      "step": 24660
+    },
+    {
+      "epoch": 4.41,
+      "grad_norm": 4.1435017585754395,
+      "learning_rate": 1.6771572195224433e-06,
+      "loss": 1.8793,
+      "step": 24665
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 0.7795052528381348,
+      "learning_rate": 1.672097860025329e-06,
+      "loss": 1.629,
+      "step": 24670
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 0.4488808512687683,
+      "learning_rate": 1.6670458791212263e-06,
+      "loss": 2.0304,
+      "step": 24675
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 0.94510817527771,
+      "learning_rate": 1.6620012784080652e-06,
+      "loss": 1.6173,
+      "step": 24680
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 0.4074093997478485,
+      "learning_rate": 1.6569640594814528e-06,
+      "loss": 2.0128,
+      "step": 24685
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 0.7733169794082642,
+      "learning_rate": 1.6519342239346564e-06,
+      "loss": 1.9417,
+      "step": 24690
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 0.7011056542396545,
+      "learning_rate": 1.6469117733586087e-06,
+      "loss": 2.0944,
+      "step": 24695
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 0.8170765042304993,
+      "learning_rate": 1.6418967093419058e-06,
+      "loss": 1.9794,
+      "step": 24700
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 0.5875701308250427,
+      "learning_rate": 1.6368890334708037e-06,
+      "loss": 1.9456,
+      "step": 24705
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 0.7417488694190979,
+      "learning_rate": 1.6318887473292243e-06,
+      "loss": 1.6503,
+      "step": 24710
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 1.339301347732544,
+      "learning_rate": 1.626895852498761e-06,
+      "loss": 1.7944,
+      "step": 24715
+    },
+    {
+      "epoch": 4.42,
+      "grad_norm": 0.6533433198928833,
+      "learning_rate": 1.6219103505586531e-06,
+      "loss": 1.764,
+      "step": 24720
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 0.9360514283180237,
+      "learning_rate": 1.6169322430858198e-06,
+      "loss": 1.7368,
+      "step": 24725
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 0.906342625617981,
+      "learning_rate": 1.6119615316548237e-06,
+      "loss": 1.7317,
+      "step": 24730
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 2.1452555656433105,
+      "learning_rate": 1.606998217837899e-06,
+      "loss": 1.6536,
+      "step": 24735
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 0.49390387535095215,
+      "learning_rate": 1.6020423032049342e-06,
+      "loss": 1.8927,
+      "step": 24740
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 0.8980386257171631,
+      "learning_rate": 1.5970937893234894e-06,
+      "loss": 1.5554,
+      "step": 24745
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 0.8987436890602112,
+      "learning_rate": 1.5921526777587625e-06,
+      "loss": 1.7345,
+      "step": 24750
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 0.7175293564796448,
+      "learning_rate": 1.5872189700736339e-06,
+      "loss": 2.1268,
+      "step": 24755
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 1.050519585609436,
+      "learning_rate": 1.5822926678286194e-06,
+      "loss": 1.9497,
+      "step": 24760
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 0.6734400987625122,
+      "learning_rate": 1.577373772581911e-06,
+      "loss": 1.5229,
+      "step": 24765
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 1.3259496688842773,
+      "learning_rate": 1.572462285889345e-06,
+      "loss": 1.8825,
+      "step": 24770
+    },
+    {
+      "epoch": 4.43,
+      "grad_norm": 1.473742127418518,
+      "learning_rate": 1.5675582093044261e-06,
+      "loss": 1.9713,
+      "step": 24775
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 1.0151005983352661,
+      "learning_rate": 1.5626615443783077e-06,
+      "loss": 1.524,
+      "step": 24780
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 0.7851746678352356,
+      "learning_rate": 1.557772292659801e-06,
+      "loss": 1.8783,
+      "step": 24785
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 3.6031758785247803,
+      "learning_rate": 1.552890455695369e-06,
+      "loss": 1.9658,
+      "step": 24790
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 1.0066990852355957,
+      "learning_rate": 1.5480160350291317e-06,
+      "loss": 1.7486,
+      "step": 24795
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 0.8470297455787659,
+      "learning_rate": 1.543149032202862e-06,
+      "loss": 1.7935,
+      "step": 24800
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 3.9266157150268555,
+      "learning_rate": 1.538289448755989e-06,
+      "loss": 1.6526,
+      "step": 24805
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 1.250511884689331,
+      "learning_rate": 1.5334372862256002e-06,
+      "loss": 1.6387,
+      "step": 24810
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 0.848745584487915,
+      "learning_rate": 1.5285925461464263e-06,
+      "loss": 2.0486,
+      "step": 24815
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 0.8530756831169128,
+      "learning_rate": 1.5237552300508524e-06,
+      "loss": 1.4405,
+      "step": 24820
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 0.5230215787887573,
+      "learning_rate": 1.5189253394689156e-06,
+      "loss": 1.7863,
+      "step": 24825
+    },
+    {
+      "epoch": 4.44,
+      "grad_norm": 1.091163992881775,
+      "learning_rate": 1.514102875928311e-06,
+      "loss": 1.5996,
+      "step": 24830
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 1.1136764287948608,
+      "learning_rate": 1.5092878409543764e-06,
+      "loss": 1.8101,
+      "step": 24835
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 1.0806092023849487,
+      "learning_rate": 1.5044802360700994e-06,
+      "loss": 1.949,
+      "step": 24840
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 0.9689674973487854,
+      "learning_rate": 1.49968006279613e-06,
+      "loss": 1.7404,
+      "step": 24845
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 1.1383538246154785,
+      "learning_rate": 1.4948873226507481e-06,
+      "loss": 1.6446,
+      "step": 24850
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 0.7477939128875732,
+      "learning_rate": 1.490102017149894e-06,
+      "loss": 1.7101,
+      "step": 24855
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 0.6716863512992859,
+      "learning_rate": 1.48532414780716e-06,
+      "loss": 1.7832,
+      "step": 24860
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 1.1172704696655273,
+      "learning_rate": 1.4805537161337813e-06,
+      "loss": 1.6232,
+      "step": 24865
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 0.48017409443855286,
+      "learning_rate": 1.4757907236386426e-06,
+      "loss": 1.7984,
+      "step": 24870
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 0.6822175979614258,
+      "learning_rate": 1.4710351718282667e-06,
+      "loss": 1.745,
+      "step": 24875
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 0.6264699697494507,
+      "learning_rate": 1.466287062206831e-06,
+      "loss": 1.778,
+      "step": 24880
+    },
+    {
+      "epoch": 4.45,
+      "grad_norm": 0.8009613752365112,
+      "learning_rate": 1.4615463962761671e-06,
+      "loss": 1.8883,
+      "step": 24885
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 26.099721908569336,
+      "learning_rate": 1.4568131755357346e-06,
+      "loss": 1.8559,
+      "step": 24890
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 1.0402318239212036,
+      "learning_rate": 1.4520874014826463e-06,
+      "loss": 1.7209,
+      "step": 24895
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 0.6136242151260376,
+      "learning_rate": 1.4473690756116654e-06,
+      "loss": 1.8066,
+      "step": 24900
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 1.0361982583999634,
+      "learning_rate": 1.442658199415192e-06,
+      "loss": 1.5658,
+      "step": 24905
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 0.9705806374549866,
+      "learning_rate": 1.4379547743832683e-06,
+      "loss": 1.9806,
+      "step": 24910
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 0.6670936346054077,
+      "learning_rate": 1.433258802003587e-06,
+      "loss": 1.9256,
+      "step": 24915
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 0.9053754806518555,
+      "learning_rate": 1.4285702837614768e-06,
+      "loss": 1.8113,
+      "step": 24920
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 0.8029524087905884,
+      "learning_rate": 1.4238892211399152e-06,
+      "loss": 1.9991,
+      "step": 24925
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 0.7751874327659607,
+      "learning_rate": 1.4192156156195151e-06,
+      "loss": 1.6467,
+      "step": 24930
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 0.59059077501297,
+      "learning_rate": 1.4145494686785387e-06,
+      "loss": 2.0298,
+      "step": 24935
+    },
+    {
+      "epoch": 4.46,
+      "grad_norm": 0.6298719048500061,
+      "learning_rate": 1.409890781792872e-06,
+      "loss": 1.7394,
+      "step": 24940
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 1.1347723007202148,
+      "learning_rate": 1.405239556436061e-06,
+      "loss": 1.8038,
+      "step": 24945
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 1.0450183153152466,
+      "learning_rate": 1.4005957940792846e-06,
+      "loss": 1.9163,
+      "step": 24950
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 0.9296015501022339,
+      "learning_rate": 1.395959496191357e-06,
+      "loss": 1.8794,
+      "step": 24955
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 0.8542814254760742,
+      "learning_rate": 1.3913306642387409e-06,
+      "loss": 1.7646,
+      "step": 24960
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 1.308875560760498,
+      "learning_rate": 1.3867092996855236e-06,
+      "loss": 1.9903,
+      "step": 24965
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 0.5955923795700073,
+      "learning_rate": 1.3820954039934413e-06,
+      "loss": 1.9698,
+      "step": 24970
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 0.6191591024398804,
+      "learning_rate": 1.3774889786218687e-06,
+      "loss": 1.7368,
+      "step": 24975
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 0.9830487966537476,
+      "learning_rate": 1.3728900250278065e-06,
+      "loss": 1.802,
+      "step": 24980
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 1.0741496086120605,
+      "learning_rate": 1.3682985446659025e-06,
+      "loss": 1.7966,
+      "step": 24985
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 0.583555281162262,
+      "learning_rate": 1.3637145389884394e-06,
+      "loss": 1.7261,
+      "step": 24990
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 0.9671819806098938,
+      "learning_rate": 1.359138009445335e-06,
+      "loss": 1.8888,
+      "step": 24995
+    },
+    {
+      "epoch": 4.47,
+      "grad_norm": 0.8640641570091248,
+      "learning_rate": 1.3545689574841342e-06,
+      "loss": 1.8415,
+      "step": 25000
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.8222974538803101,
+      "learning_rate": 1.3500073845500284e-06,
+      "loss": 1.9049,
+      "step": 25005
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 1.0911484956741333,
+      "learning_rate": 1.345453292085841e-06,
+      "loss": 1.9566,
+      "step": 25010
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.6996002793312073,
+      "learning_rate": 1.3409066815320226e-06,
+      "loss": 1.9323,
+      "step": 25015
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.5919438600540161,
+      "learning_rate": 1.3363675543266674e-06,
+      "loss": 1.8602,
+      "step": 25020
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.7225549221038818,
+      "learning_rate": 1.3318359119054935e-06,
+      "loss": 2.2012,
+      "step": 25025
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 1.0485141277313232,
+      "learning_rate": 1.327311755701849e-06,
+      "loss": 1.6628,
+      "step": 25030
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.6466124057769775,
+      "learning_rate": 1.3227950871467304e-06,
+      "loss": 1.8366,
+      "step": 25035
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 1.215564489364624,
+      "learning_rate": 1.3182859076687481e-06,
+      "loss": 1.8886,
+      "step": 25040
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 1.6709582805633545,
+      "learning_rate": 1.313784218694153e-06,
+      "loss": 1.6098,
+      "step": 25045
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 1.1631311178207397,
+      "learning_rate": 1.3092900216468336e-06,
+      "loss": 1.9456,
+      "step": 25050
+    },
+    {
+      "epoch": 4.48,
+      "grad_norm": 0.9077101349830627,
+      "learning_rate": 1.304803317948286e-06,
+      "loss": 1.9834,
+      "step": 25055
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 1.2171112298965454,
+      "learning_rate": 1.3003241090176587e-06,
+      "loss": 1.7559,
+      "step": 25060
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 1.7720481157302856,
+      "learning_rate": 1.2958523962717185e-06,
+      "loss": 1.7203,
+      "step": 25065
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 0.6833968162536621,
+      "learning_rate": 1.2913881811248696e-06,
+      "loss": 1.8261,
+      "step": 25070
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 0.8987765312194824,
+      "learning_rate": 1.28693146498913e-06,
+      "loss": 1.977,
+      "step": 25075
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 0.8224658966064453,
+      "learning_rate": 1.2824822492741584e-06,
+      "loss": 1.7903,
+      "step": 25080
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 0.8352959752082825,
+      "learning_rate": 1.2780405353872426e-06,
+      "loss": 1.6669,
+      "step": 25085
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 1.408292293548584,
+      "learning_rate": 1.273606324733284e-06,
+      "loss": 1.5876,
+      "step": 25090
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 1.146064281463623,
+      "learning_rate": 1.2691796187148247e-06,
+      "loss": 1.7544,
+      "step": 25095
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 1.177251935005188,
+      "learning_rate": 1.2647604187320223e-06,
+      "loss": 1.918,
+      "step": 25100
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 0.5201444029808044,
+      "learning_rate": 1.2603487261826724e-06,
+      "loss": 2.0014,
+      "step": 25105
+    },
+    {
+      "epoch": 4.49,
+      "grad_norm": 2.2395122051239014,
+      "learning_rate": 1.2559445424621868e-06,
+      "loss": 1.7281,
+      "step": 25110
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.5539356470108032,
+      "learning_rate": 1.251547868963604e-06,
+      "loss": 1.9028,
+      "step": 25115
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.8390756249427795,
+      "learning_rate": 1.2471587070775837e-06,
+      "loss": 1.9719,
+      "step": 25120
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.5405310392379761,
+      "learning_rate": 1.2427770581924181e-06,
+      "loss": 1.8543,
+      "step": 25125
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 2.3233208656311035,
+      "learning_rate": 1.2384029236940181e-06,
+      "loss": 1.5512,
+      "step": 25130
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.636073648929596,
+      "learning_rate": 1.2340363049659159e-06,
+      "loss": 1.9004,
+      "step": 25135
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.8926206827163696,
+      "learning_rate": 1.2296772033892733e-06,
+      "loss": 1.7828,
+      "step": 25140
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.6021214127540588,
+      "learning_rate": 1.2253256203428653e-06,
+      "loss": 2.0191,
+      "step": 25145
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.6334590911865234,
+      "learning_rate": 1.2209815572030937e-06,
+      "loss": 2.1132,
+      "step": 25150
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 7.690194606781006,
+      "learning_rate": 1.216645015343984e-06,
+      "loss": 1.5651,
+      "step": 25155
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.8695514798164368,
+      "learning_rate": 1.212315996137181e-06,
+      "loss": 1.8521,
+      "step": 25160
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 1.237356185913086,
+      "learning_rate": 1.2079945009519478e-06,
+      "loss": 1.638,
+      "step": 25165
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 0.9352409243583679,
+      "learning_rate": 1.2036805311551714e-06,
+      "loss": 1.7746,
+      "step": 25170
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 0.6192563772201538,
+      "learning_rate": 1.1993740881113492e-06,
+      "loss": 1.8845,
+      "step": 25175
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 0.8423218131065369,
+      "learning_rate": 1.1950751731826083e-06,
+      "loss": 1.7119,
+      "step": 25180
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 0.6161177158355713,
+      "learning_rate": 1.1907837877286942e-06,
+      "loss": 2.1457,
+      "step": 25185
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 0.7731286287307739,
+      "learning_rate": 1.1864999331069687e-06,
+      "loss": 1.849,
+      "step": 25190
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 1.3494420051574707,
+      "learning_rate": 1.182223610672406e-06,
+      "loss": 1.5065,
+      "step": 25195
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 0.9249375462532043,
+      "learning_rate": 1.1779548217776077e-06,
+      "loss": 1.7186,
+      "step": 25200
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 1.1438442468643188,
+      "learning_rate": 1.1736935677727852e-06,
+      "loss": 1.7934,
+      "step": 25205
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 1.2738542556762695,
+      "learning_rate": 1.1694398500057714e-06,
+      "loss": 1.6784,
+      "step": 25210
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 0.7526887059211731,
+      "learning_rate": 1.1651936698220068e-06,
+      "loss": 2.0072,
+      "step": 25215
+    },
+    {
+      "epoch": 4.51,
+      "grad_norm": 3.0581088066101074,
+      "learning_rate": 1.1609550285645588e-06,
+      "loss": 1.598,
+      "step": 25220
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 1.558251976966858,
+      "learning_rate": 1.156723927574105e-06,
+      "loss": 1.9926,
+      "step": 25225
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 1.8919259309768677,
+      "learning_rate": 1.1525003681889413e-06,
+      "loss": 1.9012,
+      "step": 25230
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 0.8434092998504639,
+      "learning_rate": 1.1482843517449688e-06,
+      "loss": 2.1936,
+      "step": 25235
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 1.0325753688812256,
+      "learning_rate": 1.144075879575715e-06,
+      "loss": 1.9423,
+      "step": 25240
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 0.6491246223449707,
+      "learning_rate": 1.1398749530123127e-06,
+      "loss": 1.7421,
+      "step": 25245
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 1.0722302198410034,
+      "learning_rate": 1.1356815733835153e-06,
+      "loss": 1.8059,
+      "step": 25250
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 0.7224903702735901,
+      "learning_rate": 1.1314957420156812e-06,
+      "loss": 1.8328,
+      "step": 25255
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 0.8088222742080688,
+      "learning_rate": 1.127317460232788e-06,
+      "loss": 1.9416,
+      "step": 25260
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 0.5477873682975769,
+      "learning_rate": 1.123146729356414e-06,
+      "loss": 1.818,
+      "step": 25265
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 2.747405767440796,
+      "learning_rate": 1.1189835507057655e-06,
+      "loss": 1.9102,
+      "step": 25270
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 0.7023490071296692,
+      "learning_rate": 1.1148279255976475e-06,
+      "loss": 1.7424,
+      "step": 25275
+    },
+    {
+      "epoch": 4.52,
+      "grad_norm": 0.6943578720092773,
+      "learning_rate": 1.1106798553464804e-06,
+      "loss": 1.7916,
+      "step": 25280
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 1.317313313484192,
+      "learning_rate": 1.1065393412642982e-06,
+      "loss": 1.8638,
+      "step": 25285
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 1.2347160577774048,
+      "learning_rate": 1.102406384660745e-06,
+      "loss": 1.8903,
+      "step": 25290
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 0.872135579586029,
+      "learning_rate": 1.0982809868430589e-06,
+      "loss": 1.7901,
+      "step": 25295
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 0.9331486225128174,
+      "learning_rate": 1.0941631491161093e-06,
+      "loss": 1.8265,
+      "step": 25300
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 0.7161145806312561,
+      "learning_rate": 1.0900528727823627e-06,
+      "loss": 1.7577,
+      "step": 25305
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 0.8512054085731506,
+      "learning_rate": 1.0859501591418907e-06,
+      "loss": 1.6612,
+      "step": 25310
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 0.9483695030212402,
+      "learning_rate": 1.081855009492383e-06,
+      "loss": 1.9453,
+      "step": 25315
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 1.1600202322006226,
+      "learning_rate": 1.0777674251291308e-06,
+      "loss": 2.1087,
+      "step": 25320
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 0.7305920124053955,
+      "learning_rate": 1.0736874073450282e-06,
+      "loss": 1.8974,
+      "step": 25325
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 0.8582291603088379,
+      "learning_rate": 1.0696149574305842e-06,
+      "loss": 1.9827,
+      "step": 25330
+    },
+    {
+      "epoch": 4.53,
+      "grad_norm": 0.6721082329750061,
+      "learning_rate": 1.065550076673913e-06,
+      "loss": 1.8574,
+      "step": 25335
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 0.8562425971031189,
+      "learning_rate": 1.0614927663607278e-06,
+      "loss": 2.1481,
+      "step": 25340
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 0.5766037702560425,
+      "learning_rate": 1.0574430277743602e-06,
+      "loss": 1.8303,
+      "step": 25345
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 1.4017845392227173,
+      "learning_rate": 1.0534008621957331e-06,
+      "loss": 1.7077,
+      "step": 25350
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 1.1896699666976929,
+      "learning_rate": 1.0493662709033764e-06,
+      "loss": 1.6934,
+      "step": 25355
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 1.20075261592865,
+      "learning_rate": 1.0453392551734276e-06,
+      "loss": 1.705,
+      "step": 25360
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 0.5349302887916565,
+      "learning_rate": 1.0413198162796346e-06,
+      "loss": 2.1699,
+      "step": 25365
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 0.656079113483429,
+      "learning_rate": 1.0373079554933357e-06,
+      "loss": 1.6925,
+      "step": 25370
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 0.7854536771774292,
+      "learning_rate": 1.0333036740834856e-06,
+      "loss": 2.0432,
+      "step": 25375
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 1.6201163530349731,
+      "learning_rate": 1.0293069733166294e-06,
+      "loss": 1.7864,
+      "step": 25380
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 0.7943071722984314,
+      "learning_rate": 1.025317854456917e-06,
+      "loss": 1.573,
+      "step": 25385
+    },
+    {
+      "epoch": 4.54,
+      "grad_norm": 1.6613608598709106,
+      "learning_rate": 1.0213363187661084e-06,
+      "loss": 1.4672,
+      "step": 25390
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 1.1010265350341797,
+      "learning_rate": 1.0173623675035604e-06,
+      "loss": 1.9815,
+      "step": 25395
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 0.885063648223877,
+      "learning_rate": 1.0133960019262256e-06,
+      "loss": 1.6383,
+      "step": 25400
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 0.5052914619445801,
+      "learning_rate": 1.0094372232886617e-06,
+      "loss": 1.8147,
+      "step": 25405
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 0.9033421277999878,
+      "learning_rate": 1.005486032843031e-06,
+      "loss": 1.7289,
+      "step": 25410
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 0.7714688181877136,
+      "learning_rate": 1.0015424318390837e-06,
+      "loss": 1.5649,
+      "step": 25415
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 0.6860387921333313,
+      "learning_rate": 9.97606421524186e-07,
+      "loss": 1.8035,
+      "step": 25420
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 0.9091027975082397,
+      "learning_rate": 9.936780031432863e-07,
+      "loss": 1.7979,
+      "step": 25425
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 1.081512212753296,
+      "learning_rate": 9.897571779389437e-07,
+      "loss": 1.6822,
+      "step": 25430
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 0.972420871257782,
+      "learning_rate": 9.858439471513131e-07,
+      "loss": 1.8356,
+      "step": 25435
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 0.838660478591919,
+      "learning_rate": 9.819383120181436e-07,
+      "loss": 1.553,
+      "step": 25440
+    },
+    {
+      "epoch": 4.55,
+      "grad_norm": 0.8015768527984619,
+      "learning_rate": 9.780402737747828e-07,
+      "loss": 1.5111,
+      "step": 25445
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 0.8660455942153931,
+      "learning_rate": 9.741498336541754e-07,
+      "loss": 1.9797,
+      "step": 25450
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 0.7163037061691284,
+      "learning_rate": 9.702669928868674e-07,
+      "loss": 1.6109,
+      "step": 25455
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 1.0026417970657349,
+      "learning_rate": 9.66391752700993e-07,
+      "loss": 1.8042,
+      "step": 25460
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 0.9547916650772095,
+      "learning_rate": 9.62524114322294e-07,
+      "loss": 1.7232,
+      "step": 25465
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 0.43152350187301636,
+      "learning_rate": 9.586640789740946e-07,
+      "loss": 2.017,
+      "step": 25470
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 0.8599415421485901,
+      "learning_rate": 9.54811647877321e-07,
+      "loss": 1.6867,
+      "step": 25475
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 1.4471591711044312,
+      "learning_rate": 9.509668222504958e-07,
+      "loss": 1.8693,
+      "step": 25480
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 0.8084843158721924,
+      "learning_rate": 9.471296033097321e-07,
+      "loss": 1.8386,
+      "step": 25485
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 0.5984373688697815,
+      "learning_rate": 9.432999922687396e-07,
+      "loss": 1.7504,
+      "step": 25490
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 0.80012047290802,
+      "learning_rate": 9.39477990338819e-07,
+      "loss": 2.0373,
+      "step": 25495
+    },
+    {
+      "epoch": 4.56,
+      "grad_norm": 1.0529673099517822,
+      "learning_rate": 9.356635987288698e-07,
+      "loss": 1.5368,
+      "step": 25500
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 0.46101412177085876,
+      "learning_rate": 9.318568186453741e-07,
+      "loss": 1.7577,
+      "step": 25505
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 0.7907940149307251,
+      "learning_rate": 9.280576512924188e-07,
+      "loss": 1.8975,
+      "step": 25510
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 1.587792992591858,
+      "learning_rate": 9.242660978716705e-07,
+      "loss": 2.0775,
+      "step": 25515
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 0.5791236758232117,
+      "learning_rate": 9.204821595823976e-07,
+      "loss": 2.0065,
+      "step": 25520
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 0.8675762414932251,
+      "learning_rate": 9.167058376214621e-07,
+      "loss": 1.8338,
+      "step": 25525
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 0.7436877489089966,
+      "learning_rate": 9.129371331832975e-07,
+      "loss": 1.7142,
+      "step": 25530
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 0.7845025658607483,
+      "learning_rate": 9.091760474599559e-07,
+      "loss": 1.9165,
+      "step": 25535
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 0.5198835730552673,
+      "learning_rate": 9.054225816410522e-07,
+      "loss": 1.7308,
+      "step": 25540
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 3.1437716484069824,
+      "learning_rate": 9.016767369138118e-07,
+      "loss": 1.8479,
+      "step": 25545
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 1.3514385223388672,
+      "learning_rate": 8.979385144630397e-07,
+      "loss": 2.1328,
+      "step": 25550
+    },
+    {
+      "epoch": 4.57,
+      "grad_norm": 0.615734875202179,
+      "learning_rate": 8.942079154711347e-07,
+      "loss": 1.9266,
+      "step": 25555
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 0.8276461958885193,
+      "learning_rate": 8.904849411180749e-07,
+      "loss": 1.844,
+      "step": 25560
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 0.8791263103485107,
+      "learning_rate": 8.867695925814406e-07,
+      "loss": 2.0249,
+      "step": 25565
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 1.3986464738845825,
+      "learning_rate": 8.830618710363892e-07,
+      "loss": 1.6506,
+      "step": 25570
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 1.0997428894042969,
+      "learning_rate": 8.793617776556685e-07,
+      "loss": 1.9504,
+      "step": 25575
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 0.9867760539054871,
+      "learning_rate": 8.7566931360962e-07,
+      "loss": 1.8399,
+      "step": 25580
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 0.6265989542007446,
+      "learning_rate": 8.719844800661625e-07,
+      "loss": 2.0406,
+      "step": 25585
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 1.4594182968139648,
+      "learning_rate": 8.68307278190808e-07,
+      "loss": 2.0839,
+      "step": 25590
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 1.0297359228134155,
+      "learning_rate": 8.646377091466457e-07,
+      "loss": 1.5879,
+      "step": 25595
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 0.9616431593894958,
+      "learning_rate": 8.609757740943608e-07,
+      "loss": 1.7454,
+      "step": 25600
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 0.7618218064308167,
+      "learning_rate": 8.5805172330512e-07,
+      "loss": 1.7113,
+      "step": 25605
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 0.9779794216156006,
+      "learning_rate": 8.544035323554217e-07,
+      "loss": 1.6699,
+      "step": 25610
+    },
+    {
+      "epoch": 4.58,
+      "grad_norm": 0.6378960013389587,
+      "learning_rate": 8.507629786346671e-07,
+      "loss": 1.5722,
+      "step": 25615
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 1.0023730993270874,
+      "learning_rate": 8.47130063294349e-07,
+      "loss": 1.7534,
+      "step": 25620
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 0.8808668255805969,
+      "learning_rate": 8.435047874835594e-07,
+      "loss": 1.5374,
+      "step": 25625
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 6.013002395629883,
+      "learning_rate": 8.398871523489727e-07,
+      "loss": 1.658,
+      "step": 25630
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 1.1340384483337402,
+      "learning_rate": 8.362771590348373e-07,
+      "loss": 1.7486,
+      "step": 25635
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 0.4705217480659485,
+      "learning_rate": 8.326748086829956e-07,
+      "loss": 1.6366,
+      "step": 25640
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 0.8444132804870605,
+      "learning_rate": 8.290801024328693e-07,
+      "loss": 1.8246,
+      "step": 25645
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 1.68918776512146,
+      "learning_rate": 8.254930414214518e-07,
+      "loss": 1.8512,
+      "step": 25650
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 0.906139075756073,
+      "learning_rate": 8.219136267833355e-07,
+      "loss": 1.8157,
+      "step": 25655
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 0.8343780040740967,
+      "learning_rate": 8.183418596506786e-07,
+      "loss": 2.0715,
+      "step": 25660
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 0.7080813050270081,
+      "learning_rate": 8.147777411532359e-07,
+      "loss": 2.1484,
+      "step": 25665
+    },
+    {
+      "epoch": 4.59,
+      "grad_norm": 0.721260130405426,
+      "learning_rate": 8.112212724183277e-07,
+      "loss": 1.6142,
+      "step": 25670
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 1.5041923522949219,
+      "learning_rate": 8.076724545708681e-07,
+      "loss": 2.0871,
+      "step": 25675
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 3.3682503700256348,
+      "learning_rate": 8.041312887333397e-07,
+      "loss": 1.5679,
+      "step": 25680
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 0.7922682166099548,
+      "learning_rate": 8.005977760258105e-07,
+      "loss": 1.696,
+      "step": 25685
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 1.6431972980499268,
+      "learning_rate": 7.970719175659281e-07,
+      "loss": 1.7266,
+      "step": 25690
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 0.8077892065048218,
+      "learning_rate": 7.935537144689142e-07,
+      "loss": 1.7022,
+      "step": 25695
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 1.4477850198745728,
+      "learning_rate": 7.900431678475817e-07,
+      "loss": 1.9362,
+      "step": 25700
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 1.0458213090896606,
+      "learning_rate": 7.865402788123088e-07,
+      "loss": 1.6532,
+      "step": 25705
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 1.3129156827926636,
+      "learning_rate": 7.830450484710511e-07,
+      "loss": 1.86,
+      "step": 25710
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 0.689514696598053,
+      "learning_rate": 7.795574779293519e-07,
+      "loss": 1.9175,
+      "step": 25715
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 1.1813409328460693,
+      "learning_rate": 7.76077568290326e-07,
+      "loss": 1.8346,
+      "step": 25720
+    },
+    {
+      "epoch": 4.6,
+      "grad_norm": 0.8686524033546448,
+      "learning_rate": 7.72605320654668e-07,
+      "loss": 1.9223,
+      "step": 25725
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 0.858902096748352,
+      "learning_rate": 7.691407361206382e-07,
+      "loss": 1.9596,
+      "step": 25730
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 0.6625041961669922,
+      "learning_rate": 7.656838157840906e-07,
+      "loss": 1.9814,
+      "step": 25735
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 0.6935396790504456,
+      "learning_rate": 7.622345607384396e-07,
+      "loss": 1.8825,
+      "step": 25740
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 1.0851209163665771,
+      "learning_rate": 7.587929720746817e-07,
+      "loss": 1.6372,
+      "step": 25745
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 0.629657506942749,
+      "learning_rate": 7.553590508813907e-07,
+      "loss": 2.0008,
+      "step": 25750
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 2.5433287620544434,
+      "learning_rate": 7.519327982447117e-07,
+      "loss": 1.7939,
+      "step": 25755
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 0.5418505668640137,
+      "learning_rate": 7.485142152483665e-07,
+      "loss": 1.8458,
+      "step": 25760
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 1.0671522617340088,
+      "learning_rate": 7.451033029736459e-07,
+      "loss": 1.7501,
+      "step": 25765
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 0.7587215304374695,
+      "learning_rate": 7.417000624994231e-07,
+      "loss": 1.713,
+      "step": 25770
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 0.7475984692573547,
+      "learning_rate": 7.383044949021339e-07,
+      "loss": 1.6557,
+      "step": 25775
+    },
+    {
+      "epoch": 4.61,
+      "grad_norm": 3.0722339153289795,
+      "learning_rate": 7.349166012557973e-07,
+      "loss": 1.6571,
+      "step": 25780
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 0.7572489976882935,
+      "learning_rate": 7.315363826320005e-07,
+      "loss": 1.9279,
+      "step": 25785
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 1.7964684963226318,
+      "learning_rate": 7.281638400998992e-07,
+      "loss": 1.8172,
+      "step": 25790
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 1.0750715732574463,
+      "learning_rate": 7.247989747262346e-07,
+      "loss": 1.7758,
+      "step": 25795
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 1.502988338470459,
+      "learning_rate": 7.214417875752999e-07,
+      "loss": 1.9997,
+      "step": 25800
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 1.200313687324524,
+      "learning_rate": 7.180922797089762e-07,
+      "loss": 1.8816,
+      "step": 25805
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 1.1731305122375488,
+      "learning_rate": 7.147504521867076e-07,
+      "loss": 2.0038,
+      "step": 25810
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 0.778863787651062,
+      "learning_rate": 7.114163060655126e-07,
+      "loss": 1.4799,
+      "step": 25815
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 1.777086853981018,
+      "learning_rate": 7.080898423999782e-07,
+      "loss": 1.2548,
+      "step": 25820
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 0.545483410358429,
+      "learning_rate": 7.047710622422599e-07,
+      "loss": 1.5645,
+      "step": 25825
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 1.450973629951477,
+      "learning_rate": 7.014599666420846e-07,
+      "loss": 1.5726,
+      "step": 25830
+    },
+    {
+      "epoch": 4.62,
+      "grad_norm": 0.8481055498123169,
+      "learning_rate": 6.981565566467507e-07,
+      "loss": 1.7464,
+      "step": 25835
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 0.9643117785453796,
+      "learning_rate": 6.948608333011197e-07,
+      "loss": 1.5418,
+      "step": 25840
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 0.6587722301483154,
+      "learning_rate": 6.915727976476299e-07,
+      "loss": 1.7643,
+      "step": 25845
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 0.37052008509635925,
+      "learning_rate": 6.882924507262855e-07,
+      "loss": 1.6896,
+      "step": 25850
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 0.8936975002288818,
+      "learning_rate": 6.850197935746511e-07,
+      "loss": 1.7418,
+      "step": 25855
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 0.858160674571991,
+      "learning_rate": 6.817548272278706e-07,
+      "loss": 1.986,
+      "step": 25860
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 0.920091450214386,
+      "learning_rate": 6.78497552718646e-07,
+      "loss": 1.8721,
+      "step": 25865
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 0.7904292941093445,
+      "learning_rate": 6.752479710772502e-07,
+      "loss": 1.7012,
+      "step": 25870
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 0.7170063853263855,
+      "learning_rate": 6.720060833315223e-07,
+      "loss": 1.9776,
+      "step": 25875
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 0.6264550685882568,
+      "learning_rate": 6.687718905068697e-07,
+      "loss": 1.953,
+      "step": 25880
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 1.133542537689209,
+      "learning_rate": 6.655453936262685e-07,
+      "loss": 1.8424,
+      "step": 25885
+    },
+    {
+      "epoch": 4.63,
+      "grad_norm": 1.201373815536499,
+      "learning_rate": 6.623265937102496e-07,
+      "loss": 1.7449,
+      "step": 25890
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 1.5119297504425049,
+      "learning_rate": 6.591154917769183e-07,
+      "loss": 2.0277,
+      "step": 25895
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 1.423034906387329,
+      "learning_rate": 6.559120888419451e-07,
+      "loss": 1.6196,
+      "step": 25900
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.8784555196762085,
+      "learning_rate": 6.527163859185614e-07,
+      "loss": 1.806,
+      "step": 25905
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.8068978190422058,
+      "learning_rate": 6.495283840175697e-07,
+      "loss": 1.8265,
+      "step": 25910
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.9359546899795532,
+      "learning_rate": 6.463480841473274e-07,
+      "loss": 1.5987,
+      "step": 25915
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 1.0493196249008179,
+      "learning_rate": 6.431754873137575e-07,
+      "loss": 1.8015,
+      "step": 25920
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.918982207775116,
+      "learning_rate": 6.400105945203544e-07,
+      "loss": 1.9332,
+      "step": 25925
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.6983169913291931,
+      "learning_rate": 6.368534067681675e-07,
+      "loss": 1.8346,
+      "step": 25930
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.8050747513771057,
+      "learning_rate": 6.337039250558119e-07,
+      "loss": 2.021,
+      "step": 25935
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 2.0071403980255127,
+      "learning_rate": 6.305621503794712e-07,
+      "loss": 1.8157,
+      "step": 25940
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 1.5265626907348633,
+      "learning_rate": 6.274280837328811e-07,
+      "loss": 1.7936,
+      "step": 25945
+    },
+    {
+      "epoch": 4.64,
+      "grad_norm": 0.8027945160865784,
+      "learning_rate": 6.24301726107343e-07,
+      "loss": 1.8368,
+      "step": 25950
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 0.6967876553535461,
+      "learning_rate": 6.211830784917244e-07,
+      "loss": 1.8268,
+      "step": 25955
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 0.5953835844993591,
+      "learning_rate": 6.1807214187245e-07,
+      "loss": 1.7233,
+      "step": 25960
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 1.5460771322250366,
+      "learning_rate": 6.149689172334994e-07,
+      "loss": 1.4829,
+      "step": 25965
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 0.8539479374885559,
+      "learning_rate": 6.118734055564291e-07,
+      "loss": 1.9009,
+      "step": 25970
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 0.4527439475059509,
+      "learning_rate": 6.087856078203419e-07,
+      "loss": 1.8119,
+      "step": 25975
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 0.5889864563941956,
+      "learning_rate": 6.057055250019039e-07,
+      "loss": 1.9578,
+      "step": 25980
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 1.3849375247955322,
+      "learning_rate": 6.026331580753442e-07,
+      "loss": 1.8432,
+      "step": 25985
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 0.8188489079475403,
+      "learning_rate": 5.995685080124491e-07,
+      "loss": 1.8869,
+      "step": 25990
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 0.6627269983291626,
+      "learning_rate": 5.965115757825657e-07,
+      "loss": 1.8587,
+      "step": 25995
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 1.0890119075775146,
+      "learning_rate": 5.934623623526009e-07,
+      "loss": 1.7339,
+      "step": 26000
+    },
+    {
+      "epoch": 4.65,
+      "grad_norm": 1.010817050933838,
+      "learning_rate": 5.90420868687011e-07,
+      "loss": 1.9268,
+      "step": 26005
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 1.0825525522232056,
+      "learning_rate": 5.873870957478211e-07,
+      "loss": 1.7062,
+      "step": 26010
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 0.8610982894897461,
+      "learning_rate": 5.843610444946135e-07,
+      "loss": 1.726,
+      "step": 26015
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 1.058080792427063,
+      "learning_rate": 5.813427158845198e-07,
+      "loss": 1.7532,
+      "step": 26020
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 2.5200893878936768,
+      "learning_rate": 5.783321108722373e-07,
+      "loss": 1.719,
+      "step": 26025
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 0.7497760057449341,
+      "learning_rate": 5.753292304100183e-07,
+      "loss": 1.9987,
+      "step": 26030
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 0.674341082572937,
+      "learning_rate": 5.723340754476697e-07,
+      "loss": 1.7749,
+      "step": 26035
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 1.0011730194091797,
+      "learning_rate": 5.693466469325587e-07,
+      "loss": 1.8461,
+      "step": 26040
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 2.4020795822143555,
+      "learning_rate": 5.663669458096015e-07,
+      "loss": 1.8198,
+      "step": 26045
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 0.7626039981842041,
+      "learning_rate": 5.633949730212806e-07,
+      "loss": 1.7917,
+      "step": 26050
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 1.1910889148712158,
+      "learning_rate": 5.604307295076188e-07,
+      "loss": 1.9962,
+      "step": 26055
+    },
+    {
+      "epoch": 4.66,
+      "grad_norm": 0.8416028618812561,
+      "learning_rate": 5.574742162062163e-07,
+      "loss": 1.8465,
+      "step": 26060
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 1.4080479145050049,
+      "learning_rate": 5.545254340522027e-07,
+      "loss": 1.948,
+      "step": 26065
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 0.9326726794242859,
+      "learning_rate": 5.515843839782848e-07,
+      "loss": 1.9587,
+      "step": 26070
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 0.7399839162826538,
+      "learning_rate": 5.486510669147072e-07,
+      "loss": 1.7994,
+      "step": 26075
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 0.7813273072242737,
+      "learning_rate": 5.457254837892778e-07,
+      "loss": 1.7135,
+      "step": 26080
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 0.8069972991943359,
+      "learning_rate": 5.428076355273564e-07,
+      "loss": 1.7997,
+      "step": 26085
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 0.7533904910087585,
+      "learning_rate": 5.398975230518577e-07,
+      "loss": 1.745,
+      "step": 26090
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 0.5249228477478027,
+      "learning_rate": 5.369951472832424e-07,
+      "loss": 1.8554,
+      "step": 26095
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 0.8555483818054199,
+      "learning_rate": 5.341005091395318e-07,
+      "loss": 1.7167,
+      "step": 26100
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 2.859463691711426,
+      "learning_rate": 5.312136095362963e-07,
+      "loss": 1.8371,
+      "step": 26105
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 0.6293110847473145,
+      "learning_rate": 5.28334449386661e-07,
+      "loss": 1.8762,
+      "step": 26110
+    },
+    {
+      "epoch": 4.67,
+      "grad_norm": 0.6581356525421143,
+      "learning_rate": 5.254630296013002e-07,
+      "loss": 1.9024,
+      "step": 26115
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 0.9610373377799988,
+      "learning_rate": 5.22599351088443e-07,
+      "loss": 1.9915,
+      "step": 26120
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 0.6756641864776611,
+      "learning_rate": 5.19743414753865e-07,
+      "loss": 1.9754,
+      "step": 26125
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 0.6355516910552979,
+      "learning_rate": 5.168952215008988e-07,
+      "loss": 1.6428,
+      "step": 26130
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 1.207905888557434,
+      "learning_rate": 5.140547722304212e-07,
+      "loss": 1.999,
+      "step": 26135
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 0.9611450433731079,
+      "learning_rate": 5.112220678408692e-07,
+      "loss": 1.9363,
+      "step": 26140
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 1.3729864358901978,
+      "learning_rate": 5.083971092282258e-07,
+      "loss": 1.5812,
+      "step": 26145
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 1.105804204940796,
+      "learning_rate": 5.055798972860154e-07,
+      "loss": 2.1176,
+      "step": 26150
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 0.9496538043022156,
+      "learning_rate": 5.027704329053223e-07,
+      "loss": 1.7854,
+      "step": 26155
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 0.8394901156425476,
+      "learning_rate": 4.999687169747775e-07,
+      "loss": 1.6806,
+      "step": 26160
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 0.6391941905021667,
+      "learning_rate": 4.971747503805608e-07,
+      "loss": 1.415,
+      "step": 26165
+    },
+    {
+      "epoch": 4.68,
+      "grad_norm": 1.3702791929244995,
+      "learning_rate": 4.943885340064042e-07,
+      "loss": 1.4161,
+      "step": 26170
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 2.06182599067688,
+      "learning_rate": 4.916100687335806e-07,
+      "loss": 2.0121,
+      "step": 26175
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 0.6842718124389648,
+      "learning_rate": 4.888393554409204e-07,
+      "loss": 1.5254,
+      "step": 26180
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 1.0825421810150146,
+      "learning_rate": 4.860763950047948e-07,
+      "loss": 1.6063,
+      "step": 26185
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 1.3437987565994263,
+      "learning_rate": 4.833211882991268e-07,
+      "loss": 1.8736,
+      "step": 26190
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 0.6650128960609436,
+      "learning_rate": 4.805737361953833e-07,
+      "loss": 1.7054,
+      "step": 26195
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 1.0552135705947876,
+      "learning_rate": 4.77834039562583e-07,
+      "loss": 1.6637,
+      "step": 26200
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 1.4072233438491821,
+      "learning_rate": 4.7510209926728834e-07,
+      "loss": 1.7958,
+      "step": 26205
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 0.7076230049133301,
+      "learning_rate": 4.7237791617361085e-07,
+      "loss": 1.9844,
+      "step": 26210
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 1.0958219766616821,
+      "learning_rate": 4.696614911432057e-07,
+      "loss": 1.8564,
+      "step": 26215
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 1.130049467086792,
+      "learning_rate": 4.669528250352745e-07,
+      "loss": 1.9523,
+      "step": 26220
+    },
+    {
+      "epoch": 4.69,
+      "grad_norm": 0.9261044263839722,
+      "learning_rate": 4.6425191870656525e-07,
+      "loss": 1.6797,
+      "step": 26225
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 0.8678873181343079,
+      "learning_rate": 4.615587730113752e-07,
+      "loss": 1.8165,
+      "step": 26230
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 0.6471136212348938,
+      "learning_rate": 4.5887338880154504e-07,
+      "loss": 1.8645,
+      "step": 26235
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 0.8412356972694397,
+      "learning_rate": 4.561957669264566e-07,
+      "loss": 1.6603,
+      "step": 26240
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 0.5464197993278503,
+      "learning_rate": 4.535259082330351e-07,
+      "loss": 1.9706,
+      "step": 26245
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 0.6575042009353638,
+      "learning_rate": 4.508638135657606e-07,
+      "loss": 1.9023,
+      "step": 26250
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 1.1826757192611694,
+      "learning_rate": 4.482094837666456e-07,
+      "loss": 1.766,
+      "step": 26255
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 0.831210196018219,
+      "learning_rate": 4.4556291967525755e-07,
+      "loss": 1.8469,
+      "step": 26260
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 1.009545922279358,
+      "learning_rate": 4.4292412212869895e-07,
+      "loss": 1.803,
+      "step": 26265
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 0.8731971979141235,
+      "learning_rate": 4.402930919616216e-07,
+      "loss": 1.9441,
+      "step": 26270
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 0.8651910424232483,
+      "learning_rate": 4.3766983000621266e-07,
+      "loss": 1.6736,
+      "step": 26275
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 1.4278265237808228,
+      "learning_rate": 4.350543370922111e-07,
+      "loss": 1.6739,
+      "step": 26280
+    },
+    {
+      "epoch": 4.7,
+      "grad_norm": 0.8426223993301392,
+      "learning_rate": 4.324466140468969e-07,
+      "loss": 1.9826,
+      "step": 26285
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 1.1639264822006226,
+      "learning_rate": 4.2984666169508525e-07,
+      "loss": 1.5842,
+      "step": 26290
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 1.0826095342636108,
+      "learning_rate": 4.2725448085914056e-07,
+      "loss": 1.6018,
+      "step": 26295
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 0.7595674991607666,
+      "learning_rate": 4.246700723589708e-07,
+      "loss": 2.0188,
+      "step": 26300
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 1.1753851175308228,
+      "learning_rate": 4.2209343701201666e-07,
+      "loss": 1.6287,
+      "step": 26305
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 1.429380178451538,
+      "learning_rate": 4.1952457563326773e-07,
+      "loss": 1.9633,
+      "step": 26310
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 0.7897043228149414,
+      "learning_rate": 4.16963489035252e-07,
+      "loss": 1.6705,
+      "step": 26315
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 0.7799373865127563,
+      "learning_rate": 4.1441017802803793e-07,
+      "loss": 2.1176,
+      "step": 26320
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 0.7283823490142822,
+      "learning_rate": 4.1186464341923805e-07,
+      "loss": 1.5486,
+      "step": 26325
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 1.3124147653579712,
+      "learning_rate": 4.093268860139998e-07,
+      "loss": 1.9634,
+      "step": 26330
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 1.9731324911117554,
+      "learning_rate": 4.067969066150146e-07,
+      "loss": 1.4846,
+      "step": 26335
+    },
+    {
+      "epoch": 4.71,
+      "grad_norm": 0.9217849969863892,
+      "learning_rate": 4.042747060225116e-07,
+      "loss": 2.1779,
+      "step": 26340
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 2.033876419067383,
+      "learning_rate": 4.0176028503425835e-07,
+      "loss": 1.6705,
+      "step": 26345
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 0.941889226436615,
+      "learning_rate": 3.9925364444556857e-07,
+      "loss": 1.6641,
+      "step": 26350
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 1.0339444875717163,
+      "learning_rate": 3.9675478504929144e-07,
+      "loss": 1.7571,
+      "step": 26355
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 0.7928982973098755,
+      "learning_rate": 3.942637076358058e-07,
+      "loss": 1.8514,
+      "step": 26360
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 1.492556095123291,
+      "learning_rate": 3.917804129930452e-07,
+      "loss": 1.8306,
+      "step": 26365
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 0.8601492643356323,
+      "learning_rate": 3.893049019064676e-07,
+      "loss": 1.8331,
+      "step": 26370
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 1.3913602828979492,
+      "learning_rate": 3.8683717515907714e-07,
+      "loss": 1.7354,
+      "step": 26375
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 1.4564555883407593,
+      "learning_rate": 3.843772335314133e-07,
+      "loss": 1.5457,
+      "step": 26380
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 0.7062227725982666,
+      "learning_rate": 3.8192507780155375e-07,
+      "loss": 1.6495,
+      "step": 26385
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 0.7881594300270081,
+      "learning_rate": 3.794807087451141e-07,
+      "loss": 1.963,
+      "step": 26390
+    },
+    {
+      "epoch": 4.72,
+      "grad_norm": 1.6842689514160156,
+      "learning_rate": 3.770441271352426e-07,
+      "loss": 1.5499,
+      "step": 26395
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 0.4173828065395355,
+      "learning_rate": 3.7461533374262837e-07,
+      "loss": 2.0362,
+      "step": 26400
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 1.6171375513076782,
+      "learning_rate": 3.7219432933549845e-07,
+      "loss": 1.6515,
+      "step": 26405
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 0.5490692853927612,
+      "learning_rate": 3.6978111467961263e-07,
+      "loss": 1.8352,
+      "step": 26410
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 0.5645069479942322,
+      "learning_rate": 3.6737569053827137e-07,
+      "loss": 1.7449,
+      "step": 26415
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 0.6575131416320801,
+      "learning_rate": 3.6497805767230495e-07,
+      "loss": 1.8291,
+      "step": 26420
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 2.23494815826416,
+      "learning_rate": 3.625882168400846e-07,
+      "loss": 1.6557,
+      "step": 26425
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 2.7113113403320312,
+      "learning_rate": 3.6020616879750835e-07,
+      "loss": 1.8655,
+      "step": 26430
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 0.5787123441696167,
+      "learning_rate": 3.5783191429802354e-07,
+      "loss": 1.7335,
+      "step": 26435
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 1.4365309476852417,
+      "learning_rate": 3.554654540925989e-07,
+      "loss": 1.8369,
+      "step": 26440
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 0.9646137356758118,
+      "learning_rate": 3.531067889297496e-07,
+      "loss": 1.6218,
+      "step": 26445
+    },
+    {
+      "epoch": 4.73,
+      "grad_norm": 1.048437237739563,
+      "learning_rate": 3.507559195555149e-07,
+      "loss": 1.8628,
+      "step": 26450
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 1.032910943031311,
+      "learning_rate": 3.484128467134723e-07,
+      "loss": 1.6514,
+      "step": 26455
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 1.4005285501480103,
+      "learning_rate": 3.4607757114473174e-07,
+      "loss": 1.5883,
+      "step": 26460
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 0.918569803237915,
+      "learning_rate": 3.4375009358794395e-07,
+      "loss": 1.9173,
+      "step": 26465
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 1.5146937370300293,
+      "learning_rate": 3.4143041477928686e-07,
+      "loss": 1.736,
+      "step": 26470
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 1.1808782815933228,
+      "learning_rate": 3.391185354524706e-07,
+      "loss": 1.8146,
+      "step": 26475
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 1.120993971824646,
+      "learning_rate": 3.36814456338741e-07,
+      "loss": 1.7978,
+      "step": 26480
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 0.922327995300293,
+      "learning_rate": 3.3451817816687323e-07,
+      "loss": 1.7257,
+      "step": 26485
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 2.25972056388855,
+      "learning_rate": 3.322297016631809e-07,
+      "loss": 1.6793,
+      "step": 26490
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 0.4978795349597931,
+      "learning_rate": 3.2994902755150715e-07,
+      "loss": 1.9777,
+      "step": 26495
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 0.6013870239257812,
+      "learning_rate": 3.276761565532277e-07,
+      "loss": 1.9321,
+      "step": 26500
+    },
+    {
+      "epoch": 4.74,
+      "grad_norm": 0.6871728897094727,
+      "learning_rate": 3.254110893872481e-07,
+      "loss": 1.8226,
+      "step": 26505
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 1.049520492553711,
+      "learning_rate": 3.231538267700063e-07,
+      "loss": 1.6252,
+      "step": 26510
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 1.513724684715271,
+      "learning_rate": 3.209043694154729e-07,
+      "loss": 1.7386,
+      "step": 26515
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 3.1947197914123535,
+      "learning_rate": 3.186627180351509e-07,
+      "loss": 1.9751,
+      "step": 26520
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 0.9903146624565125,
+      "learning_rate": 3.164288733380677e-07,
+      "loss": 1.5443,
+      "step": 26525
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 0.7547667622566223,
+      "learning_rate": 3.1420283603079135e-07,
+      "loss": 1.9813,
+      "step": 26530
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 1.0118529796600342,
+      "learning_rate": 3.1198460681741427e-07,
+      "loss": 1.9654,
+      "step": 26535
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 0.8024758100509644,
+      "learning_rate": 3.097741863995557e-07,
+      "loss": 1.6795,
+      "step": 26540
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 0.8070296049118042,
+      "learning_rate": 3.0757157547637307e-07,
+      "loss": 1.9996,
+      "step": 26545
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 0.769829273223877,
+      "learning_rate": 3.053767747445535e-07,
+      "loss": 1.8527,
+      "step": 26550
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 0.8586844801902771,
+      "learning_rate": 3.0318978489830286e-07,
+      "loss": 1.6287,
+      "step": 26555
+    },
+    {
+      "epoch": 4.75,
+      "grad_norm": 0.8367322683334351,
+      "learning_rate": 3.010106066293733e-07,
+      "loss": 2.0245,
+      "step": 26560
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 0.4964026212692261,
+      "learning_rate": 2.988392406270274e-07,
+      "loss": 1.4126,
+      "step": 26565
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 0.8849166035652161,
+      "learning_rate": 2.966756875780713e-07,
+      "loss": 1.9657,
+      "step": 26570
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 0.6733694076538086,
+      "learning_rate": 2.945199481668326e-07,
+      "loss": 1.8609,
+      "step": 26575
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 2.2621817588806152,
+      "learning_rate": 2.923720230751714e-07,
+      "loss": 1.9039,
+      "step": 26580
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 0.6737426519393921,
+      "learning_rate": 2.9023191298247487e-07,
+      "loss": 1.6034,
+      "step": 26585
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 0.8793543577194214,
+      "learning_rate": 2.8809961856565425e-07,
+      "loss": 2.0761,
+      "step": 26590
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 0.6130470633506775,
+      "learning_rate": 2.8597514049915617e-07,
+      "loss": 2.0985,
+      "step": 26595
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 0.6239492297172546,
+      "learning_rate": 2.838584794549487e-07,
+      "loss": 1.7084,
+      "step": 26600
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 1.3211489915847778,
+      "learning_rate": 2.817496361025296e-07,
+      "loss": 1.6366,
+      "step": 26605
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 0.7192025184631348,
+      "learning_rate": 2.7964861110892636e-07,
+      "loss": 1.8488,
+      "step": 26610
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 1.1748217344284058,
+      "learning_rate": 2.7755540513868805e-07,
+      "loss": 1.9784,
+      "step": 26615
+    },
+    {
+      "epoch": 4.76,
+      "grad_norm": 0.6899864673614502,
+      "learning_rate": 2.754700188538961e-07,
+      "loss": 1.8471,
+      "step": 26620
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 1.0577079057693481,
+      "learning_rate": 2.733924529141535e-07,
+      "loss": 1.7249,
+      "step": 26625
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 0.9071574211120605,
+      "learning_rate": 2.7132270797659563e-07,
+      "loss": 1.6626,
+      "step": 26630
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 0.7166086435317993,
+      "learning_rate": 2.6926078469587944e-07,
+      "loss": 2.038,
+      "step": 26635
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 0.6789818406105042,
+      "learning_rate": 2.67206683724186e-07,
+      "loss": 1.8971,
+      "step": 26640
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 1.1377719640731812,
+      "learning_rate": 2.6516040571123166e-07,
+      "loss": 1.5134,
+      "step": 26645
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 1.127935767173767,
+      "learning_rate": 2.6312195130424875e-07,
+      "loss": 2.0224,
+      "step": 26650
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 1.656747817993164,
+      "learning_rate": 2.610913211479993e-07,
+      "loss": 1.9184,
+      "step": 26655
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 1.3809555768966675,
+      "learning_rate": 2.5906851588476945e-07,
+      "loss": 1.5658,
+      "step": 26660
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 0.8653160333633423,
+      "learning_rate": 2.570535361543669e-07,
+      "loss": 1.7374,
+      "step": 26665
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 1.040223479270935,
+      "learning_rate": 2.550463825941346e-07,
+      "loss": 1.9036,
+      "step": 26670
+    },
+    {
+      "epoch": 4.77,
+      "grad_norm": 1.1421611309051514,
+      "learning_rate": 2.530470558389286e-07,
+      "loss": 1.9153,
+      "step": 26675
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 1.0162934064865112,
+      "learning_rate": 2.510555565211348e-07,
+      "loss": 1.6926,
+      "step": 26680
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 0.722922682762146,
+      "learning_rate": 2.4907188527066303e-07,
+      "loss": 1.6469,
+      "step": 26685
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 0.6988521218299866,
+      "learning_rate": 2.4709604271494766e-07,
+      "loss": 1.4591,
+      "step": 26690
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 0.9216463565826416,
+      "learning_rate": 2.451280294789443e-07,
+      "loss": 1.6103,
+      "step": 26695
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 0.6147362589836121,
+      "learning_rate": 2.4316784618513276e-07,
+      "loss": 1.788,
+      "step": 26700
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 0.583010196685791,
+      "learning_rate": 2.412154934535199e-07,
+      "loss": 1.7635,
+      "step": 26705
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 1.2794963121414185,
+      "learning_rate": 2.392709719016312e-07,
+      "loss": 1.8667,
+      "step": 26710
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 1.555069923400879,
+      "learning_rate": 2.3733428214451903e-07,
+      "loss": 1.7455,
+      "step": 26715
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 0.6522120833396912,
+      "learning_rate": 2.354054247947518e-07,
+      "loss": 1.8311,
+      "step": 26720
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 0.864811360836029,
+      "learning_rate": 2.3348440046243035e-07,
+      "loss": 1.8641,
+      "step": 26725
+    },
+    {
+      "epoch": 4.78,
+      "grad_norm": 0.828033983707428,
+      "learning_rate": 2.3157120975516867e-07,
+      "loss": 1.8355,
+      "step": 26730
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 0.6203212141990662,
+      "learning_rate": 2.2966585327811041e-07,
+      "loss": 2.0341,
+      "step": 26735
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 0.5540245771408081,
+      "learning_rate": 2.2776833163391796e-07,
+      "loss": 2.0733,
+      "step": 26740
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 0.819915235042572,
+      "learning_rate": 2.2587864542277237e-07,
+      "loss": 1.8464,
+      "step": 26745
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 3.0156285762786865,
+      "learning_rate": 2.239967952423816e-07,
+      "loss": 1.5715,
+      "step": 26750
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 0.48900169134140015,
+      "learning_rate": 2.2212278168797507e-07,
+      "loss": 1.6596,
+      "step": 26755
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 0.8913301825523376,
+      "learning_rate": 2.202566053522953e-07,
+      "loss": 1.7543,
+      "step": 26760
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 0.641745924949646,
+      "learning_rate": 2.1839826682562015e-07,
+      "loss": 2.1231,
+      "step": 26765
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 0.7494519352912903,
+      "learning_rate": 2.1654776669573496e-07,
+      "loss": 1.89,
+      "step": 26770
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 1.3078620433807373,
+      "learning_rate": 2.147051055479521e-07,
+      "loss": 1.4582,
+      "step": 26775
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 3.589418411254883,
+      "learning_rate": 2.1287028396510534e-07,
+      "loss": 1.5499,
+      "step": 26780
+    },
+    {
+      "epoch": 4.79,
+      "grad_norm": 1.803573489189148,
+      "learning_rate": 2.1104330252754435e-07,
+      "loss": 1.9718,
+      "step": 26785
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.8699598908424377,
+      "learning_rate": 2.092241618131402e-07,
+      "loss": 1.8986,
+      "step": 26790
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.7145684361457825,
+      "learning_rate": 2.0741286239729097e-07,
+      "loss": 1.6495,
+      "step": 26795
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 1.2166204452514648,
+      "learning_rate": 2.0560940485290782e-07,
+      "loss": 1.5914,
+      "step": 26800
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.681698739528656,
+      "learning_rate": 2.0381378975041775e-07,
+      "loss": 1.5633,
+      "step": 26805
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.5658136606216431,
+      "learning_rate": 2.0202601765777762e-07,
+      "loss": 1.8271,
+      "step": 26810
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.5894467234611511,
+      "learning_rate": 2.0024608914045173e-07,
+      "loss": 1.8671,
+      "step": 26815
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 1.0941160917282104,
+      "learning_rate": 1.9847400476143695e-07,
+      "loss": 1.679,
+      "step": 26820
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 1.179622769355774,
+      "learning_rate": 1.9670976508123495e-07,
+      "loss": 2.0309,
+      "step": 26825
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 1.4097144603729248,
+      "learning_rate": 1.9495337065787988e-07,
+      "loss": 2.1893,
+      "step": 26830
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.7634019255638123,
+      "learning_rate": 1.9320482204691348e-07,
+      "loss": 1.7503,
+      "step": 26835
+    },
+    {
+      "epoch": 4.8,
+      "grad_norm": 0.8397713303565979,
+      "learning_rate": 1.914641198013989e-07,
+      "loss": 2.0055,
+      "step": 26840
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 0.6048543453216553,
+      "learning_rate": 1.8973126447192345e-07,
+      "loss": 2.0813,
+      "step": 26845
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 0.7331050634384155,
+      "learning_rate": 1.880062566065821e-07,
+      "loss": 1.8715,
+      "step": 26850
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 0.8746078014373779,
+      "learning_rate": 1.862890967509967e-07,
+      "loss": 1.719,
+      "step": 26855
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 1.2344975471496582,
+      "learning_rate": 1.8457978544830224e-07,
+      "loss": 2.062,
+      "step": 26860
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 0.6543428301811218,
+      "learning_rate": 1.8287832323915233e-07,
+      "loss": 1.6365,
+      "step": 26865
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 0.6258736252784729,
+      "learning_rate": 1.8118471066171648e-07,
+      "loss": 1.6869,
+      "step": 26870
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 0.3230958580970764,
+      "learning_rate": 1.7949894825168568e-07,
+      "loss": 1.9677,
+      "step": 26875
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 0.5818564891815186,
+      "learning_rate": 1.778210365422611e-07,
+      "loss": 1.7614,
+      "step": 26880
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 0.5392193794250488,
+      "learning_rate": 1.761509760641683e-07,
+      "loss": 1.8393,
+      "step": 26885
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 0.8611071109771729,
+      "learning_rate": 1.744887673456458e-07,
+      "loss": 1.7183,
+      "step": 26890
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 1.2561523914337158,
+      "learning_rate": 1.7283441091244523e-07,
+      "loss": 1.9883,
+      "step": 26895
+    },
+    {
+      "epoch": 4.81,
+      "grad_norm": 0.6996719837188721,
+      "learning_rate": 1.7118790728783973e-07,
+      "loss": 2.3327,
+      "step": 26900
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 0.7118642330169678,
+      "learning_rate": 1.6954925699261825e-07,
+      "loss": 1.9309,
+      "step": 26905
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 1.1299420595169067,
+      "learning_rate": 1.6791846054508008e-07,
+      "loss": 1.8533,
+      "step": 26910
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 1.176224946975708,
+      "learning_rate": 1.6629551846104874e-07,
+      "loss": 1.7047,
+      "step": 26915
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 0.4716276228427887,
+      "learning_rate": 1.646804312538608e-07,
+      "loss": 2.2831,
+      "step": 26920
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 0.8160402774810791,
+      "learning_rate": 1.630731994343604e-07,
+      "loss": 1.8131,
+      "step": 26925
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 1.1298454999923706,
+      "learning_rate": 1.6147382351091865e-07,
+      "loss": 1.829,
+      "step": 26930
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 1.2337285280227661,
+      "learning_rate": 1.5988230398941428e-07,
+      "loss": 1.9875,
+      "step": 26935
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 0.5422369837760925,
+      "learning_rate": 1.5829864137324457e-07,
+      "loss": 2.1579,
+      "step": 26940
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 0.5247811079025269,
+      "learning_rate": 1.5672283616331717e-07,
+      "loss": 1.7073,
+      "step": 26945
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 0.9248299598693848,
+      "learning_rate": 1.5515488885806394e-07,
+      "loss": 1.8775,
+      "step": 26950
+    },
+    {
+      "epoch": 4.82,
+      "grad_norm": 0.7368882298469543,
+      "learning_rate": 1.535947999534215e-07,
+      "loss": 1.7339,
+      "step": 26955
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 1.1814441680908203,
+      "learning_rate": 1.5204256994284238e-07,
+      "loss": 1.7054,
+      "step": 26960
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 0.6178357005119324,
+      "learning_rate": 1.504981993172977e-07,
+      "loss": 2.0362,
+      "step": 26965
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 0.6023995280265808,
+      "learning_rate": 1.489616885652717e-07,
+      "loss": 2.0766,
+      "step": 26970
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 0.8028101921081543,
+      "learning_rate": 1.4743303817275899e-07,
+      "loss": 2.2482,
+      "step": 26975
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 0.9648069143295288,
+      "learning_rate": 1.459122486232728e-07,
+      "loss": 1.8054,
+      "step": 26980
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 1.3614518642425537,
+      "learning_rate": 1.4439932039783665e-07,
+      "loss": 1.6154,
+      "step": 26985
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 1.080051302909851,
+      "learning_rate": 1.4289425397498445e-07,
+      "loss": 1.6912,
+      "step": 26990
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 0.6755841374397278,
+      "learning_rate": 1.413970498307715e-07,
+      "loss": 1.7686,
+      "step": 26995
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 12.691487312316895,
+      "learning_rate": 1.3990770843876067e-07,
+      "loss": 1.681,
+      "step": 27000
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 0.9654897451400757,
+      "learning_rate": 1.3842623027002787e-07,
+      "loss": 1.8103,
+      "step": 27005
+    },
+    {
+      "epoch": 4.83,
+      "grad_norm": 1.236765742301941,
+      "learning_rate": 1.3695261579316777e-07,
+      "loss": 1.5977,
+      "step": 27010
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 2.983935594558716,
+      "learning_rate": 1.3548686547427692e-07,
+      "loss": 1.7417,
+      "step": 27015
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 0.7265644669532776,
+      "learning_rate": 1.3402897977697614e-07,
+      "loss": 1.9217,
+      "step": 27020
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 1.0526750087738037,
+      "learning_rate": 1.3257895916238828e-07,
+      "loss": 1.8388,
+      "step": 27025
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 0.9290766716003418,
+      "learning_rate": 1.3113680408915752e-07,
+      "loss": 1.7867,
+      "step": 27030
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 0.6184346079826355,
+      "learning_rate": 1.2970251501343844e-07,
+      "loss": 1.7741,
+      "step": 27035
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 1.0642505884170532,
+      "learning_rate": 1.2827609238888762e-07,
+      "loss": 1.7068,
+      "step": 27040
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 0.8582693338394165,
+      "learning_rate": 1.2685753666668576e-07,
+      "loss": 1.4969,
+      "step": 27045
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 0.6961191892623901,
+      "learning_rate": 1.2544684829552122e-07,
+      "loss": 1.6328,
+      "step": 27050
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 1.6947689056396484,
+      "learning_rate": 1.2404402772159262e-07,
+      "loss": 1.6862,
+      "step": 27055
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 0.803805947303772,
+      "learning_rate": 1.2264907538860893e-07,
+      "loss": 1.5166,
+      "step": 27060
+    },
+    {
+      "epoch": 4.84,
+      "grad_norm": 1.7080804109573364,
+      "learning_rate": 1.2126199173779496e-07,
+      "loss": 1.632,
+      "step": 27065
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 0.6481025815010071,
+      "learning_rate": 1.1988277720788034e-07,
+      "loss": 1.9797,
+      "step": 27070
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 1.0819138288497925,
+      "learning_rate": 1.1851143223511329e-07,
+      "loss": 1.5857,
+      "step": 27075
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 0.7787542343139648,
+      "learning_rate": 1.1714795725324967e-07,
+      "loss": 1.5872,
+      "step": 27080
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 2.253755807876587,
+      "learning_rate": 1.1579235269355005e-07,
+      "loss": 1.7227,
+      "step": 27085
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 0.5482271313667297,
+      "learning_rate": 1.1444461898479364e-07,
+      "loss": 1.7388,
+      "step": 27090
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 1.8450013399124146,
+      "learning_rate": 1.1310475655326724e-07,
+      "loss": 1.908,
+      "step": 27095
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 0.5819000601768494,
+      "learning_rate": 1.1177276582276797e-07,
+      "loss": 1.613,
+      "step": 27100
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 0.8016330003738403,
+      "learning_rate": 1.1044864721460602e-07,
+      "loss": 1.6827,
+      "step": 27105
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 1.288928508758545,
+      "learning_rate": 1.091324011475936e-07,
+      "loss": 1.846,
+      "step": 27110
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 1.4904510974884033,
+      "learning_rate": 1.0782402803806158e-07,
+      "loss": 2.0203,
+      "step": 27115
+    },
+    {
+      "epoch": 4.85,
+      "grad_norm": 2.6243813037872314,
+      "learning_rate": 1.0652352829984558e-07,
+      "loss": 1.6311,
+      "step": 27120
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 2.045247793197632,
+      "learning_rate": 1.0523090234429434e-07,
+      "loss": 2.0991,
+      "step": 27125
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 0.736707329750061,
+      "learning_rate": 1.0394615058026414e-07,
+      "loss": 1.7908,
+      "step": 27130
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 1.2909306287765503,
+      "learning_rate": 1.0266927341411602e-07,
+      "loss": 2.2229,
+      "step": 27135
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 0.4913961887359619,
+      "learning_rate": 1.0140027124973251e-07,
+      "loss": 1.8207,
+      "step": 27140
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 0.5628072023391724,
+      "learning_rate": 1.0013914448848971e-07,
+      "loss": 2.0149,
+      "step": 27145
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 0.8567386865615845,
+      "learning_rate": 9.8885893529288e-08,
+      "loss": 2.0128,
+      "step": 27150
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 0.6592267155647278,
+      "learning_rate": 9.764051876852421e-08,
+      "loss": 2.0507,
+      "step": 27155
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 1.0533310174942017,
+      "learning_rate": 9.640302060011375e-08,
+      "loss": 1.8298,
+      "step": 27160
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 1.0910136699676514,
+      "learning_rate": 9.517339941547132e-08,
+      "loss": 1.7826,
+      "step": 27165
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 1.4886125326156616,
+      "learning_rate": 9.395165560352748e-08,
+      "loss": 1.424,
+      "step": 27170
+    },
+    {
+      "epoch": 4.86,
+      "grad_norm": 0.8163747191429138,
+      "learning_rate": 9.273778955072032e-08,
+      "loss": 1.8783,
+      "step": 27175
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 0.813804566860199,
+      "learning_rate": 9.153180164098995e-08,
+      "loss": 1.9613,
+      "step": 27180
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 1.107713222503662,
+      "learning_rate": 9.033369225579514e-08,
+      "loss": 1.783,
+      "step": 27185
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 1.2432942390441895,
+      "learning_rate": 8.914346177409105e-08,
+      "loss": 1.7609,
+      "step": 27190
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 0.697809636592865,
+      "learning_rate": 8.79611105723488e-08,
+      "loss": 1.8995,
+      "step": 27195
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 0.8270529508590698,
+      "learning_rate": 8.67866390245442e-08,
+      "loss": 1.8255,
+      "step": 27200
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 0.9013446569442749,
+      "learning_rate": 8.562004750216346e-08,
+      "loss": 1.7947,
+      "step": 27205
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 1.4320096969604492,
+      "learning_rate": 8.446133637419751e-08,
+      "loss": 1.6338,
+      "step": 27210
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 0.9843838810920715,
+      "learning_rate": 8.331050600714485e-08,
+      "loss": 1.8977,
+      "step": 27215
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 1.381026029586792,
+      "learning_rate": 8.216755676501431e-08,
+      "loss": 1.7343,
+      "step": 27220
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 1.0219390392303467,
+      "learning_rate": 8.103248900931393e-08,
+      "loss": 1.7915,
+      "step": 27225
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 1.4810404777526855,
+      "learning_rate": 7.99053030990704e-08,
+      "loss": 1.9256,
+      "step": 27230
+    },
+    {
+      "epoch": 4.87,
+      "grad_norm": 1.307970643043518,
+      "learning_rate": 7.878599939080689e-08,
+      "loss": 1.9488,
+      "step": 27235
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 0.675598680973053,
+      "learning_rate": 7.767457823856239e-08,
+      "loss": 1.6861,
+      "step": 27240
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 1.2670272588729858,
+      "learning_rate": 7.657103999387794e-08,
+      "loss": 1.7473,
+      "step": 27245
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 1.0107417106628418,
+      "learning_rate": 7.547538500579932e-08,
+      "loss": 1.5708,
+      "step": 27250
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 0.8755956888198853,
+      "learning_rate": 7.438761362087987e-08,
+      "loss": 1.728,
+      "step": 27255
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 1.5213618278503418,
+      "learning_rate": 7.330772618318604e-08,
+      "loss": 1.6688,
+      "step": 27260
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 0.8350248336791992,
+      "learning_rate": 7.223572303428072e-08,
+      "loss": 1.5138,
+      "step": 27265
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 0.5836665630340576,
+      "learning_rate": 7.11716045132399e-08,
+      "loss": 1.6876,
+      "step": 27270
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 0.3720304071903229,
+      "learning_rate": 7.011537095663878e-08,
+      "loss": 2.2579,
+      "step": 27275
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 0.8098456859588623,
+      "learning_rate": 6.906702269856846e-08,
+      "loss": 1.8464,
+      "step": 27280
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 1.2636229991912842,
+      "learning_rate": 6.802656007061925e-08,
+      "loss": 1.9191,
+      "step": 27285
+    },
+    {
+      "epoch": 4.88,
+      "grad_norm": 0.7597413063049316,
+      "learning_rate": 6.699398340188623e-08,
+      "loss": 1.9295,
+      "step": 27290
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 0.806469738483429,
+      "learning_rate": 6.596929301897481e-08,
+      "loss": 1.8776,
+      "step": 27295
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 0.9175061583518982,
+      "learning_rate": 6.49524892459924e-08,
+      "loss": 1.7537,
+      "step": 27300
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 1.0410821437835693,
+      "learning_rate": 6.394357240455118e-08,
+      "loss": 1.7922,
+      "step": 27305
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 0.9964597821235657,
+      "learning_rate": 6.294254281377366e-08,
+      "loss": 1.9924,
+      "step": 27310
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 0.8122162818908691,
+      "learning_rate": 6.194940079028432e-08,
+      "loss": 1.8831,
+      "step": 27315
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 2.2000720500946045,
+      "learning_rate": 6.096414664821248e-08,
+      "loss": 1.7875,
+      "step": 27320
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 0.8936071395874023,
+      "learning_rate": 5.998678069919216e-08,
+      "loss": 1.8224,
+      "step": 27325
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 0.8874375224113464,
+      "learning_rate": 5.901730325236221e-08,
+      "loss": 2.0273,
+      "step": 27330
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 1.6214640140533447,
+      "learning_rate": 5.805571461437176e-08,
+      "loss": 1.7428,
+      "step": 27335
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 1.3129786252975464,
+      "learning_rate": 5.7102015089366415e-08,
+      "loss": 1.6939,
+      "step": 27340
+    },
+    {
+      "epoch": 4.89,
+      "grad_norm": 1.0049517154693604,
+      "learning_rate": 5.615620497900209e-08,
+      "loss": 1.6544,
+      "step": 27345
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 0.76762455701828,
+      "learning_rate": 5.521828458243672e-08,
+      "loss": 1.9076,
+      "step": 27350
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 0.8442772030830383,
+      "learning_rate": 5.428825419633299e-08,
+      "loss": 1.7669,
+      "step": 27355
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 1.0416758060455322,
+      "learning_rate": 5.336611411486114e-08,
+      "loss": 2.1945,
+      "step": 27360
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 0.9710814356803894,
+      "learning_rate": 5.245186462969065e-08,
+      "loss": 1.6822,
+      "step": 27365
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 0.7552046775817871,
+      "learning_rate": 5.15455060300013e-08,
+      "loss": 1.9038,
+      "step": 27370
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 0.8535296320915222,
+      "learning_rate": 5.064703860246933e-08,
+      "loss": 1.8449,
+      "step": 27375
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 5.469326019287109,
+      "learning_rate": 4.975646263128131e-08,
+      "loss": 1.8349,
+      "step": 27380
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 0.8397666215896606,
+      "learning_rate": 4.8873778398123015e-08,
+      "loss": 1.7978,
+      "step": 27385
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 0.7340749502182007,
+      "learning_rate": 4.799898618219057e-08,
+      "loss": 1.7184,
+      "step": 27390
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 0.6980647444725037,
+      "learning_rate": 4.713208626017929e-08,
+      "loss": 1.8878,
+      "step": 27395
+    },
+    {
+      "epoch": 4.9,
+      "grad_norm": 0.6763983964920044,
+      "learning_rate": 4.627307890628374e-08,
+      "loss": 2.0292,
+      "step": 27400
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 2.8043484687805176,
+      "learning_rate": 4.5421964392214354e-08,
+      "loss": 1.76,
+      "step": 27405
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 1.528351068496704,
+      "learning_rate": 4.457874298717246e-08,
+      "loss": 1.8167,
+      "step": 27410
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 1.0853112936019897,
+      "learning_rate": 4.3743414957872485e-08,
+      "loss": 2.0553,
+      "step": 27415
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 0.819485068321228,
+      "learning_rate": 4.291598056852253e-08,
+      "loss": 1.9577,
+      "step": 27420
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 0.8028863072395325,
+      "learning_rate": 4.209644008084379e-08,
+      "loss": 1.927,
+      "step": 27425
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 0.6567749977111816,
+      "learning_rate": 4.128479375405392e-08,
+      "loss": 1.9202,
+      "step": 27430
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 1.1371631622314453,
+      "learning_rate": 4.048104184487811e-08,
+      "loss": 2.0074,
+      "step": 27435
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 0.6734791994094849,
+      "learning_rate": 3.968518460753801e-08,
+      "loss": 1.7147,
+      "step": 27440
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 1.4970152378082275,
+      "learning_rate": 3.8897222293768356e-08,
+      "loss": 1.8979,
+      "step": 27445
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 0.4336738586425781,
+      "learning_rate": 3.8117155152797567e-08,
+      "loss": 1.8246,
+      "step": 27450
+    },
+    {
+      "epoch": 4.91,
+      "grad_norm": 1.3205974102020264,
+      "learning_rate": 3.734498343135884e-08,
+      "loss": 1.837,
+      "step": 27455
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 1.5103232860565186,
+      "learning_rate": 3.658070737369013e-08,
+      "loss": 1.7212,
+      "step": 27460
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 0.7287768721580505,
+      "learning_rate": 3.5824327221534174e-08,
+      "loss": 1.9573,
+      "step": 27465
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 1.077132225036621,
+      "learning_rate": 3.507584321412738e-08,
+      "loss": 1.9364,
+      "step": 27470
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 0.7396588921546936,
+      "learning_rate": 3.433525558822204e-08,
+      "loss": 1.8377,
+      "step": 27475
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 0.9306398034095764,
+      "learning_rate": 3.360256457805855e-08,
+      "loss": 1.8822,
+      "step": 27480
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 1.1197668313980103,
+      "learning_rate": 3.287777041539042e-08,
+      "loss": 1.9135,
+      "step": 27485
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 1.111188292503357,
+      "learning_rate": 3.2160873329470374e-08,
+      "loss": 1.8155,
+      "step": 27490
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 0.7195913791656494,
+      "learning_rate": 3.145187354704759e-08,
+      "loss": 1.7282,
+      "step": 27495
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 1.0142110586166382,
+      "learning_rate": 3.075077129238158e-08,
+      "loss": 1.9826,
+      "step": 27500
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 0.8807904720306396,
+      "learning_rate": 3.005756678722826e-08,
+      "loss": 1.9266,
+      "step": 27505
+    },
+    {
+      "epoch": 4.92,
+      "grad_norm": 0.3830811381340027,
+      "learning_rate": 2.9372260250848384e-08,
+      "loss": 1.7985,
+      "step": 27510
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 1.0831133127212524,
+      "learning_rate": 2.869485190000465e-08,
+      "loss": 1.6724,
+      "step": 27515
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 1.3438478708267212,
+      "learning_rate": 2.8025341948959007e-08,
+      "loss": 1.7674,
+      "step": 27520
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 0.8136791586875916,
+      "learning_rate": 2.7363730609478166e-08,
+      "loss": 1.8693,
+      "step": 27525
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 0.6372392177581787,
+      "learning_rate": 2.6710018090828072e-08,
+      "loss": 1.7738,
+      "step": 27530
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 0.6655257344245911,
+      "learning_rate": 2.606420459977943e-08,
+      "loss": 1.7049,
+      "step": 27535
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 0.8369601964950562,
+      "learning_rate": 2.555324124416669e-08,
+      "loss": 1.6822,
+      "step": 27540
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 0.9216079711914062,
+      "learning_rate": 2.4921646515879692e-08,
+      "loss": 1.6634,
+      "step": 27545
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 0.874138593673706,
+      "learning_rate": 2.429795138085278e-08,
+      "loss": 1.948,
+      "step": 27550
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 0.4770451784133911,
+      "learning_rate": 2.3682156036358704e-08,
+      "loss": 1.7761,
+      "step": 27555
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 0.8585416078567505,
+      "learning_rate": 2.3074260677177773e-08,
+      "loss": 1.7923,
+      "step": 27560
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 0.9223566055297852,
+      "learning_rate": 2.247426549558118e-08,
+      "loss": 1.7544,
+      "step": 27565
+    },
+    {
+      "epoch": 4.93,
+      "grad_norm": 3.3756611347198486,
+      "learning_rate": 2.188217068134768e-08,
+      "loss": 1.961,
+      "step": 27570
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 0.6999983191490173,
+      "learning_rate": 2.129797642176079e-08,
+      "loss": 1.8334,
+      "step": 27575
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 0.6343751549720764,
+      "learning_rate": 2.0721682901594928e-08,
+      "loss": 2.1177,
+      "step": 27580
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 1.4471246004104614,
+      "learning_rate": 2.0153290303134843e-08,
+      "loss": 1.4078,
+      "step": 27585
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 0.7915092706680298,
+      "learning_rate": 1.9592798806161716e-08,
+      "loss": 1.6325,
+      "step": 27590
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 0.42571014165878296,
+      "learning_rate": 1.9040208587958742e-08,
+      "loss": 2.0373,
+      "step": 27595
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 1.225578784942627,
+      "learning_rate": 1.8495519823308327e-08,
+      "loss": 1.7639,
+      "step": 27600
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 3.2549374103546143,
+      "learning_rate": 1.7958732684497657e-08,
+      "loss": 1.7882,
+      "step": 27605
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 0.732199490070343,
+      "learning_rate": 1.7429847341307594e-08,
+      "loss": 1.9269,
+      "step": 27610
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 0.9305698871612549,
+      "learning_rate": 1.6908863961026534e-08,
+      "loss": 1.9614,
+      "step": 27615
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 1.5303813219070435,
+      "learning_rate": 1.6395782708442108e-08,
+      "loss": 1.7238,
+      "step": 27620
+    },
+    {
+      "epoch": 4.94,
+      "grad_norm": 1.0078951120376587,
+      "learning_rate": 1.589060374584117e-08,
+      "loss": 1.7562,
+      "step": 27625
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 0.5599997639656067,
+      "learning_rate": 1.539332723300979e-08,
+      "loss": 1.6895,
+      "step": 27630
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 1.057673692703247,
+      "learning_rate": 1.4903953327238818e-08,
+      "loss": 1.6365,
+      "step": 27635
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 1.3415483236312866,
+      "learning_rate": 1.442248218331277e-08,
+      "loss": 1.8398,
+      "step": 27640
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 2.2374091148376465,
+      "learning_rate": 1.3948913953523713e-08,
+      "loss": 1.6291,
+      "step": 27645
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 0.6912126541137695,
+      "learning_rate": 1.3483248787657387e-08,
+      "loss": 1.5426,
+      "step": 27650
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 0.8157421350479126,
+      "learning_rate": 1.3025486833007083e-08,
+      "loss": 1.8275,
+      "step": 27655
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 0.5781735181808472,
+      "learning_rate": 1.257562823435976e-08,
+      "loss": 1.7979,
+      "step": 27660
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 0.9726621508598328,
+      "learning_rate": 1.2133673134007152e-08,
+      "loss": 1.719,
+      "step": 27665
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 0.7362151145935059,
+      "learning_rate": 1.1699621671740212e-08,
+      "loss": 1.5477,
+      "step": 27670
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 0.6761454343795776,
+      "learning_rate": 1.1273473984843575e-08,
+      "loss": 2.0092,
+      "step": 27675
+    },
+    {
+      "epoch": 4.95,
+      "grad_norm": 0.835422158241272,
+      "learning_rate": 1.0855230208114964e-08,
+      "loss": 1.7663,
+      "step": 27680
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.9953777194023132,
+      "learning_rate": 1.0444890473837454e-08,
+      "loss": 1.5584,
+      "step": 27685
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.6336172223091125,
+      "learning_rate": 1.0042454911804444e-08,
+      "loss": 1.968,
+      "step": 27690
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 1.8674088716506958,
+      "learning_rate": 9.647923649308554e-09,
+      "loss": 1.7018,
+      "step": 27695
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.79061359167099,
+      "learning_rate": 9.261296811133301e-09,
+      "loss": 2.1185,
+      "step": 27700
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.5622634887695312,
+      "learning_rate": 8.882574519572529e-09,
+      "loss": 1.7226,
+      "step": 27705
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.5330947637557983,
+      "learning_rate": 8.511756894416523e-09,
+      "loss": 1.8869,
+      "step": 27710
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.9711994528770447,
+      "learning_rate": 8.14884405295202e-09,
+      "loss": 1.9226,
+      "step": 27715
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.642650842666626,
+      "learning_rate": 7.793836109970532e-09,
+      "loss": 1.8734,
+      "step": 27720
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.5877659916877747,
+      "learning_rate": 7.446733177757237e-09,
+      "loss": 1.7147,
+      "step": 27725
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.6836099028587341,
+      "learning_rate": 7.107535366102092e-09,
+      "loss": 1.6646,
+      "step": 27730
+    },
+    {
+      "epoch": 4.96,
+      "grad_norm": 0.694749653339386,
+      "learning_rate": 6.776242782297049e-09,
+      "loss": 1.6795,
+      "step": 27735
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 1.7434329986572266,
+      "learning_rate": 6.452855531122182e-09,
+      "loss": 1.8301,
+      "step": 27740
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 1.0804805755615234,
+      "learning_rate": 6.137373714870664e-09,
+      "loss": 1.6298,
+      "step": 27745
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 0.8923995494842529,
+      "learning_rate": 5.829797433323791e-09,
+      "loss": 1.7,
+      "step": 27750
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 1.1635453701019287,
+      "learning_rate": 5.530126783773182e-09,
+      "loss": 1.6506,
+      "step": 27755
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 2.5559163093566895,
+      "learning_rate": 5.238361861001351e-09,
+      "loss": 1.4222,
+      "step": 27760
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 0.6483443975448608,
+      "learning_rate": 4.954502757295587e-09,
+      "loss": 2.0705,
+      "step": 27765
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 0.3663721978664398,
+      "learning_rate": 4.678549562434076e-09,
+      "loss": 1.8228,
+      "step": 27770
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 0.8673094511032104,
+      "learning_rate": 4.410502363708102e-09,
+      "loss": 1.8538,
+      "step": 27775
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 1.0372596979141235,
+      "learning_rate": 4.150361245897072e-09,
+      "loss": 1.7484,
+      "step": 27780
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 0.5771445035934448,
+      "learning_rate": 3.898126291285164e-09,
+      "loss": 1.9486,
+      "step": 27785
+    },
+    {
+      "epoch": 4.97,
+      "grad_norm": 0.7575253844261169,
+      "learning_rate": 3.653797579650231e-09,
+      "loss": 1.8841,
+      "step": 27790
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 0.6591800451278687,
+      "learning_rate": 3.417375188274896e-09,
+      "loss": 2.1256,
+      "step": 27795
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 0.8006610870361328,
+      "learning_rate": 3.1888591919437826e-09,
+      "loss": 2.0,
+      "step": 27800
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 1.2859355211257935,
+      "learning_rate": 2.968249662929634e-09,
+      "loss": 1.8523,
+      "step": 27805
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 0.5861554145812988,
+      "learning_rate": 2.755546671015519e-09,
+      "loss": 1.7343,
+      "step": 27810
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 0.8748874068260193,
+      "learning_rate": 2.550750283478176e-09,
+      "loss": 1.8467,
+      "step": 27815
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 0.9922271966934204,
+      "learning_rate": 2.353860565090793e-09,
+      "loss": 1.5163,
+      "step": 27820
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 1.116941213607788,
+      "learning_rate": 2.164877578136881e-09,
+      "loss": 1.9498,
+      "step": 27825
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 1.195860743522644,
+      "learning_rate": 1.9838013823852974e-09,
+      "loss": 1.593,
+      "step": 27830
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 0.7228025794029236,
+      "learning_rate": 1.8106320351124472e-09,
+      "loss": 2.0734,
+      "step": 27835
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 0.9624055624008179,
+      "learning_rate": 1.6453695910911837e-09,
+      "loss": 1.6158,
+      "step": 27840
+    },
+    {
+      "epoch": 4.98,
+      "grad_norm": 1.733687162399292,
+      "learning_rate": 1.4880141025935824e-09,
+      "loss": 2.0339,
+      "step": 27845
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 0.7384397983551025,
+      "learning_rate": 1.3385656193909413e-09,
+      "loss": 1.9362,
+      "step": 27850
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 1.2739673852920532,
+      "learning_rate": 1.197024188756557e-09,
+      "loss": 1.6635,
+      "step": 27855
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 0.9207239747047424,
+      "learning_rate": 1.0633898554573972e-09,
+      "loss": 1.7917,
+      "step": 27860
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 2.776045560836792,
+      "learning_rate": 9.376626617624286e-10,
+      "loss": 1.9558,
+      "step": 27865
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 0.9185320734977722,
+      "learning_rate": 8.198426474370635e-10,
+      "loss": 1.8871,
+      "step": 27870
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 0.9698821306228638,
+      "learning_rate": 7.099298497487139e-10,
+      "loss": 1.5789,
+      "step": 27875
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 1.133810043334961,
+      "learning_rate": 6.079243034667891e-10,
+      "loss": 1.7469,
+      "step": 27880
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 0.8469708561897278,
+      "learning_rate": 5.138260408488194e-10,
+      "loss": 2.05,
+      "step": 27885
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 1.0808436870574951,
+      "learning_rate": 4.2763509166265927e-10,
+      "loss": 1.9934,
+      "step": 27890
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 1.0401432514190674,
+      "learning_rate": 3.4935148316705967e-10,
+      "loss": 1.8931,
+      "step": 27895
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 0.6832406520843506,
+      "learning_rate": 2.789752401283208e-10,
+      "loss": 2.1373,
+      "step": 27900
+    },
+    {
+      "epoch": 4.99,
+      "grad_norm": 0.7210304141044617,
+      "learning_rate": 2.1650638480086305e-10,
+      "loss": 1.9971,
+      "step": 27905
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 1.0963140726089478,
+      "learning_rate": 1.6194493694665635e-10,
+      "loss": 1.7998,
+      "step": 27910
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.7948571443557739,
+      "learning_rate": 1.1529091382134205e-10,
+      "loss": 1.9675,
+      "step": 27915
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.8925894498825073,
+      "learning_rate": 7.654433018255969e-11,
+      "loss": 1.7117,
+      "step": 27920
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.588504433631897,
+      "learning_rate": 4.5705198287171456e-11,
+      "loss": 1.6579,
+      "step": 27925
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 2.567326068878174,
+      "learning_rate": 2.2773527888486635e-11,
+      "loss": 1.817,
+      "step": 27930
+    },
+    {
+      "epoch": 5.0,
+      "step": 27930,
+      "total_flos": 5.219882421458043e+18,
+      "train_loss": 1.860821307947505,
+      "train_runtime": 129586.2862,
+      "train_samples_per_second": 1.724,
+      "train_steps_per_second": 0.216
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 27930,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 100,
+  "total_flos": 5.219882421458043e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}