diff --git "a/UNIVLA_LIBERO_IMG_BS192_8K/trainer_state.json" "b/UNIVLA_LIBERO_IMG_BS192_8K/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/UNIVLA_LIBERO_IMG_BS192_8K/trainer_state.json"
@@ -0,0 +1,5642 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 444.44444444444446,
+  "eval_steps": 500,
+  "global_step": 8000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.5555555555555556,
+      "grad_norm": 9.624756252620806,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 18.1578,
+      "step": 10
+    },
+    {
+      "epoch": 1.1111111111111112,
+      "grad_norm": 22.971985050664838,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 9.8958,
+      "step": 20
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 8.887738609061802,
+      "learning_rate": 4.8e-05,
+      "loss": 6.2347,
+      "step": 30
+    },
+    {
+      "epoch": 2.2222222222222223,
+      "grad_norm": 10.45842103554072,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 5.6558,
+      "step": 40
+    },
+    {
+      "epoch": 2.7777777777777777,
+      "grad_norm": 5.5562623443294745,
+      "learning_rate": 8e-05,
+      "loss": 5.3792,
+      "step": 50
+    },
+    {
+      "epoch": 3.3333333333333335,
+      "grad_norm": 7.087782177025818,
+      "learning_rate": 7.999970720328831e-05,
+      "loss": 5.2697,
+      "step": 60
+    },
+    {
+      "epoch": 3.888888888888889,
+      "grad_norm": 5.2287321277434256,
+      "learning_rate": 7.999882881772551e-05,
+      "loss": 5.0329,
+      "step": 70
+    },
+    {
+      "epoch": 4.444444444444445,
+      "grad_norm": 7.695007875871005,
+      "learning_rate": 7.999736485702831e-05,
+      "loss": 4.929,
+      "step": 80
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 6.380963995163142,
+      "learning_rate": 7.999531534405768e-05,
+      "loss": 4.7967,
+      "step": 90
+    },
+    {
+      "epoch": 5.555555555555555,
+      "grad_norm": 23.681790672849388,
+      "learning_rate": 7.999268031081841e-05,
+      "loss": 4.6923,
+      "step": 100
+    },
+    {
+      "epoch": 6.111111111111111,
+      "grad_norm": 11.998572582543087,
+      "learning_rate": 7.998945979845876e-05,
+      "loss": 4.787,
+      "step": 110
+    },
+    {
+      "epoch": 6.666666666666667,
+      "grad_norm": 6.894315185619389,
+      "learning_rate": 7.998565385726968e-05,
+      "loss": 4.6014,
+      "step": 120
+    },
+    {
+      "epoch": 7.222222222222222,
+      "grad_norm": 11.41476966709072,
+      "learning_rate": 7.998126254668407e-05,
+      "loss": 4.5322,
+      "step": 130
+    },
+    {
+      "epoch": 7.777777777777778,
+      "grad_norm": 12.932268789729228,
+      "learning_rate": 7.997628593527586e-05,
+      "loss": 4.4906,
+      "step": 140
+    },
+    {
+      "epoch": 8.333333333333334,
+      "grad_norm": 16.308505332844156,
+      "learning_rate": 7.997072410075895e-05,
+      "loss": 4.3894,
+      "step": 150
+    },
+    {
+      "epoch": 8.88888888888889,
+      "grad_norm": 6.109462415455175,
+      "learning_rate": 7.996457712998597e-05,
+      "loss": 4.2926,
+      "step": 160
+    },
+    {
+      "epoch": 9.444444444444445,
+      "grad_norm": 6.422037758354163,
+      "learning_rate": 7.995784511894694e-05,
+      "loss": 4.3249,
+      "step": 170
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 5.549363128962184,
+      "learning_rate": 7.995052817276779e-05,
+      "loss": 4.2333,
+      "step": 180
+    },
+    {
+      "epoch": 10.555555555555555,
+      "grad_norm": 5.5940405061610265,
+      "learning_rate": 7.994262640570861e-05,
+      "loss": 4.1332,
+      "step": 190
+    },
+    {
+      "epoch": 11.11111111111111,
+      "grad_norm": 6.715748387015782,
+      "learning_rate": 7.993413994116206e-05,
+      "loss": 4.1768,
+      "step": 200
+    },
+    {
+      "epoch": 11.666666666666666,
+      "grad_norm": 3.933021177955018,
+      "learning_rate": 7.992506891165125e-05,
+      "loss": 4.083,
+      "step": 210
+    },
+    {
+      "epoch": 12.222222222222221,
+      "grad_norm": 5.482671837294648,
+      "learning_rate": 7.991541345882781e-05,
+      "loss": 4.0218,
+      "step": 220
+    },
+    {
+      "epoch": 12.777777777777779,
+      "grad_norm": 7.350567708829696,
+      "learning_rate": 7.990517373346957e-05,
+      "loss": 3.9837,
+      "step": 230
+    },
+    {
+      "epoch": 13.333333333333334,
+      "grad_norm": 7.659932150278813,
+      "learning_rate": 7.989434989547832e-05,
+      "loss": 3.9255,
+      "step": 240
+    },
+    {
+      "epoch": 13.88888888888889,
+      "grad_norm": 7.540779890610956,
+      "learning_rate": 7.988294211387719e-05,
+      "loss": 3.9092,
+      "step": 250
+    },
+    {
+      "epoch": 14.444444444444445,
+      "grad_norm": 5.26091086706904,
+      "learning_rate": 7.987095056680811e-05,
+      "loss": 3.8718,
+      "step": 260
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 8.929974568876512,
+      "learning_rate": 7.985837544152896e-05,
+      "loss": 3.8535,
+      "step": 270
+    },
+    {
+      "epoch": 15.555555555555555,
+      "grad_norm": 4.304275409891799,
+      "learning_rate": 7.984521693441071e-05,
+      "loss": 3.821,
+      "step": 280
+    },
+    {
+      "epoch": 16.11111111111111,
+      "grad_norm": 12.663064210151889,
+      "learning_rate": 7.983147525093428e-05,
+      "loss": 3.7561,
+      "step": 290
+    },
+    {
+      "epoch": 16.666666666666668,
+      "grad_norm": 5.468595666070034,
+      "learning_rate": 7.981715060568741e-05,
+      "loss": 3.7577,
+      "step": 300
+    },
+    {
+      "epoch": 17.22222222222222,
+      "grad_norm": 6.019803743426774,
+      "learning_rate": 7.980224322236123e-05,
+      "loss": 3.6914,
+      "step": 310
+    },
+    {
+      "epoch": 17.77777777777778,
+      "grad_norm": 5.361918477003201,
+      "learning_rate": 7.978675333374685e-05,
+      "loss": 3.6932,
+      "step": 320
+    },
+    {
+      "epoch": 18.333333333333332,
+      "grad_norm": 7.1639847579615,
+      "learning_rate": 7.977068118173162e-05,
+      "loss": 3.6518,
+      "step": 330
+    },
+    {
+      "epoch": 18.88888888888889,
+      "grad_norm": 7.261448469872607,
+      "learning_rate": 7.975402701729546e-05,
+      "loss": 3.5769,
+      "step": 340
+    },
+    {
+      "epoch": 19.444444444444443,
+      "grad_norm": 8.0019257135087,
+      "learning_rate": 7.97367911005069e-05,
+      "loss": 3.5828,
+      "step": 350
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 4.234907166095184,
+      "learning_rate": 7.971897370051896e-05,
+      "loss": 3.5309,
+      "step": 360
+    },
+    {
+      "epoch": 20.555555555555557,
+      "grad_norm": 4.987259272452073,
+      "learning_rate": 7.970057509556505e-05,
+      "loss": 3.5506,
+      "step": 370
+    },
+    {
+      "epoch": 21.11111111111111,
+      "grad_norm": 5.815327274628244,
+      "learning_rate": 7.968159557295457e-05,
+      "loss": 3.5395,
+      "step": 380
+    },
+    {
+      "epoch": 21.666666666666668,
+      "grad_norm": 3.9735054901342406,
+      "learning_rate": 7.966203542906841e-05,
+      "loss": 3.5006,
+      "step": 390
+    },
+    {
+      "epoch": 22.22222222222222,
+      "grad_norm": 4.329744882798763,
+      "learning_rate": 7.964189496935435e-05,
+      "loss": 3.4282,
+      "step": 400
+    },
+    {
+      "epoch": 22.77777777777778,
+      "grad_norm": 9.576521423421042,
+      "learning_rate": 7.962117450832225e-05,
+      "loss": 3.3955,
+      "step": 410
+    },
+    {
+      "epoch": 23.333333333333332,
+      "grad_norm": 6.758969599276602,
+      "learning_rate": 7.959987436953921e-05,
+      "loss": 3.3677,
+      "step": 420
+    },
+    {
+      "epoch": 23.88888888888889,
+      "grad_norm": 6.70022000762921,
+      "learning_rate": 7.957799488562449e-05,
+      "loss": 3.4147,
+      "step": 430
+    },
+    {
+      "epoch": 24.444444444444443,
+      "grad_norm": 8.049603223795994,
+      "learning_rate": 7.955553639824423e-05,
+      "loss": 3.3966,
+      "step": 440
+    },
+    {
+      "epoch": 25.0,
+      "grad_norm": 3.599540887803097,
+      "learning_rate": 7.953249925810626e-05,
+      "loss": 3.3529,
+      "step": 450
+    },
+    {
+      "epoch": 25.555555555555557,
+      "grad_norm": 6.637753572667117,
+      "learning_rate": 7.950888382495452e-05,
+      "loss": 3.3147,
+      "step": 460
+    },
+    {
+      "epoch": 26.11111111111111,
+      "grad_norm": 4.6829201259216156,
+      "learning_rate": 7.948469046756344e-05,
+      "loss": 3.3358,
+      "step": 470
+    },
+    {
+      "epoch": 26.666666666666668,
+      "grad_norm": 5.2128150680057255,
+      "learning_rate": 7.945991956373229e-05,
+      "loss": 3.3251,
+      "step": 480
+    },
+    {
+      "epoch": 27.22222222222222,
+      "grad_norm": 3.8939615060542554,
+      "learning_rate": 7.943457150027913e-05,
+      "loss": 3.2542,
+      "step": 490
+    },
+    {
+      "epoch": 27.77777777777778,
+      "grad_norm": 6.497910383767135,
+      "learning_rate": 7.940864667303489e-05,
+      "loss": 3.2533,
+      "step": 500
+    },
+    {
+      "epoch": 28.333333333333332,
+      "grad_norm": 6.839880280367184,
+      "learning_rate": 7.938214548683712e-05,
+      "loss": 3.2514,
+      "step": 510
+    },
+    {
+      "epoch": 28.88888888888889,
+      "grad_norm": 3.888770829785339,
+      "learning_rate": 7.935506835552368e-05,
+      "loss": 3.2322,
+      "step": 520
+    },
+    {
+      "epoch": 29.444444444444443,
+      "grad_norm": 3.819439809777116,
+      "learning_rate": 7.932741570192633e-05,
+      "loss": 3.2223,
+      "step": 530
+    },
+    {
+      "epoch": 30.0,
+      "grad_norm": 8.124697635599855,
+      "learning_rate": 7.929918795786404e-05,
+      "loss": 3.2142,
+      "step": 540
+    },
+    {
+      "epoch": 30.555555555555557,
+      "grad_norm": 4.9424901926872415,
+      "learning_rate": 7.927038556413631e-05,
+      "loss": 3.1729,
+      "step": 550
+    },
+    {
+      "epoch": 31.11111111111111,
+      "grad_norm": 3.5572516614564353,
+      "learning_rate": 7.924100897051629e-05,
+      "loss": 3.1864,
+      "step": 560
+    },
+    {
+      "epoch": 31.666666666666668,
+      "grad_norm": 7.016836083089255,
+      "learning_rate": 7.921105863574369e-05,
+      "loss": 3.1884,
+      "step": 570
+    },
+    {
+      "epoch": 32.22222222222222,
+      "grad_norm": 3.689220996701065,
+      "learning_rate": 7.918053502751772e-05,
+      "loss": 3.1394,
+      "step": 580
+    },
+    {
+      "epoch": 32.77777777777778,
+      "grad_norm": 4.237286794694591,
+      "learning_rate": 7.914943862248966e-05,
+      "loss": 3.1173,
+      "step": 590
+    },
+    {
+      "epoch": 33.333333333333336,
+      "grad_norm": 6.465351483555105,
+      "learning_rate": 7.911776990625551e-05,
+      "loss": 3.0855,
+      "step": 600
+    },
+    {
+      "epoch": 33.888888888888886,
+      "grad_norm": 3.470090315025474,
+      "learning_rate": 7.908552937334843e-05,
+      "loss": 3.1065,
+      "step": 610
+    },
+    {
+      "epoch": 34.44444444444444,
+      "grad_norm": 4.297075856456074,
+      "learning_rate": 7.905271752723088e-05,
+      "loss": 3.1087,
+      "step": 620
+    },
+    {
+      "epoch": 35.0,
+      "grad_norm": 3.829969934689117,
+      "learning_rate": 7.901933488028692e-05,
+      "loss": 3.1313,
+      "step": 630
+    },
+    {
+      "epoch": 35.55555555555556,
+      "grad_norm": 3.04859038751192,
+      "learning_rate": 7.898538195381413e-05,
+      "loss": 3.0937,
+      "step": 640
+    },
+    {
+      "epoch": 36.111111111111114,
+      "grad_norm": 3.9192499745300764,
+      "learning_rate": 7.895085927801541e-05,
+      "loss": 3.1012,
+      "step": 650
+    },
+    {
+      "epoch": 36.666666666666664,
+      "grad_norm": 3.7938275450270864,
+      "learning_rate": 7.891576739199084e-05,
+      "loss": 3.0701,
+      "step": 660
+    },
+    {
+      "epoch": 37.22222222222222,
+      "grad_norm": 4.214462701438513,
+      "learning_rate": 7.888010684372915e-05,
+      "loss": 3.0746,
+      "step": 670
+    },
+    {
+      "epoch": 37.77777777777778,
+      "grad_norm": 3.5302589848856996,
+      "learning_rate": 7.884387819009922e-05,
+      "loss": 3.0314,
+      "step": 680
+    },
+    {
+      "epoch": 38.333333333333336,
+      "grad_norm": 3.349982327087245,
+      "learning_rate": 7.880708199684133e-05,
+      "loss": 3.0858,
+      "step": 690
+    },
+    {
+      "epoch": 38.888888888888886,
+      "grad_norm": 6.723071886210811,
+      "learning_rate": 7.87697188385584e-05,
+      "loss": 3.0457,
+      "step": 700
+    },
+    {
+      "epoch": 39.44444444444444,
+      "grad_norm": 4.113172509765762,
+      "learning_rate": 7.873178929870695e-05,
+      "loss": 2.9944,
+      "step": 710
+    },
+    {
+      "epoch": 40.0,
+      "grad_norm": 3.497427812639807,
+      "learning_rate": 7.869329396958801e-05,
+      "loss": 3.006,
+      "step": 720
+    },
+    {
+      "epoch": 40.55555555555556,
+      "grad_norm": 5.8220136985290045,
+      "learning_rate": 7.86542334523379e-05,
+      "loss": 3.0329,
+      "step": 730
+    },
+    {
+      "epoch": 41.111111111111114,
+      "grad_norm": 3.6682425204840143,
+      "learning_rate": 7.861460835691881e-05,
+      "loss": 2.9786,
+      "step": 740
+    },
+    {
+      "epoch": 41.666666666666664,
+      "grad_norm": 3.3462898376852253,
+      "learning_rate": 7.857441930210928e-05,
+      "loss": 2.9879,
+      "step": 750
+    },
+    {
+      "epoch": 42.22222222222222,
+      "grad_norm": 3.9246042185677656,
+      "learning_rate": 7.853366691549454e-05,
+      "loss": 2.963,
+      "step": 760
+    },
+    {
+      "epoch": 42.77777777777778,
+      "grad_norm": 4.288690714191476,
+      "learning_rate": 7.849235183345669e-05,
+      "loss": 3.0163,
+      "step": 770
+    },
+    {
+      "epoch": 43.333333333333336,
+      "grad_norm": 3.15485132027285,
+      "learning_rate": 7.845047470116485e-05,
+      "loss": 2.947,
+      "step": 780
+    },
+    {
+      "epoch": 43.888888888888886,
+      "grad_norm": 14.074942954237473,
+      "learning_rate": 7.840803617256494e-05,
+      "loss": 3.0502,
+      "step": 790
+    },
+    {
+      "epoch": 44.44444444444444,
+      "grad_norm": 4.148400258042579,
+      "learning_rate": 7.836503691036959e-05,
+      "loss": 3.1179,
+      "step": 800
+    },
+    {
+      "epoch": 45.0,
+      "grad_norm": 3.3580782494484764,
+      "learning_rate": 7.832147758604775e-05,
+      "loss": 3.0479,
+      "step": 810
+    },
+    {
+      "epoch": 45.55555555555556,
+      "grad_norm": 3.296947460372628,
+      "learning_rate": 7.827735887981415e-05,
+      "loss": 2.9871,
+      "step": 820
+    },
+    {
+      "epoch": 46.111111111111114,
+      "grad_norm": 3.317028477037915,
+      "learning_rate": 7.823268148061884e-05,
+      "loss": 2.9622,
+      "step": 830
+    },
+    {
+      "epoch": 46.666666666666664,
+      "grad_norm": 4.266805967401394,
+      "learning_rate": 7.818744608613617e-05,
+      "loss": 2.9287,
+      "step": 840
+    },
+    {
+      "epoch": 47.22222222222222,
+      "grad_norm": 2.9805058668232762,
+      "learning_rate": 7.814165340275418e-05,
+      "loss": 2.9585,
+      "step": 850
+    },
+    {
+      "epoch": 47.77777777777778,
+      "grad_norm": 3.0450009573698495,
+      "learning_rate": 7.809530414556335e-05,
+      "loss": 2.9407,
+      "step": 860
+    },
+    {
+      "epoch": 48.333333333333336,
+      "grad_norm": 4.50934774346947,
+      "learning_rate": 7.804839903834557e-05,
+      "loss": 2.9294,
+      "step": 870
+    },
+    {
+      "epoch": 48.888888888888886,
+      "grad_norm": 5.5632179739854,
+      "learning_rate": 7.800093881356278e-05,
+      "loss": 2.9301,
+      "step": 880
+    },
+    {
+      "epoch": 49.44444444444444,
+      "grad_norm": 3.023169690476701,
+      "learning_rate": 7.79529242123455e-05,
+      "loss": 2.9458,
+      "step": 890
+    },
+    {
+      "epoch": 50.0,
+      "grad_norm": 3.245904102133423,
+      "learning_rate": 7.790435598448135e-05,
+      "loss": 2.9213,
+      "step": 900
+    },
+    {
+      "epoch": 50.55555555555556,
+      "grad_norm": 2.8338775299143086,
+      "learning_rate": 7.785523488840324e-05,
+      "loss": 2.9409,
+      "step": 910
+    },
+    {
+      "epoch": 51.111111111111114,
+      "grad_norm": 3.550676233062673,
+      "learning_rate": 7.780556169117757e-05,
+      "loss": 2.9025,
+      "step": 920
+    },
+    {
+      "epoch": 51.666666666666664,
+      "grad_norm": 3.0407811970595833,
+      "learning_rate": 7.775533716849233e-05,
+      "loss": 2.926,
+      "step": 930
+    },
+    {
+      "epoch": 52.22222222222222,
+      "grad_norm": 3.5628872041923194,
+      "learning_rate": 7.770456210464481e-05,
+      "loss": 2.8804,
+      "step": 940
+    },
+    {
+      "epoch": 52.77777777777778,
+      "grad_norm": 2.6878750561558626,
+      "learning_rate": 7.765323729252954e-05,
+      "loss": 2.8726,
+      "step": 950
+    },
+    {
+      "epoch": 53.333333333333336,
+      "grad_norm": 3.2362781767136077,
+      "learning_rate": 7.760136353362576e-05,
+      "loss": 2.8688,
+      "step": 960
+    },
+    {
+      "epoch": 53.888888888888886,
+      "grad_norm": 2.953196368288579,
+      "learning_rate": 7.7548941637985e-05,
+      "loss": 2.9105,
+      "step": 970
+    },
+    {
+      "epoch": 54.44444444444444,
+      "grad_norm": 2.7777733665007425,
+      "learning_rate": 7.749597242421838e-05,
+      "loss": 2.8708,
+      "step": 980
+    },
+    {
+      "epoch": 55.0,
+      "grad_norm": 3.5697588668724873,
+      "learning_rate": 7.744245671948386e-05,
+      "loss": 2.8536,
+      "step": 990
+    },
+    {
+      "epoch": 55.55555555555556,
+      "grad_norm": 2.912195993692092,
+      "learning_rate": 7.738839535947331e-05,
+      "loss": 2.8433,
+      "step": 1000
+    },
+    {
+      "epoch": 56.111111111111114,
+      "grad_norm": 2.8401704833748607,
+      "learning_rate": 7.733378918839942e-05,
+      "loss": 2.8454,
+      "step": 1010
+    },
+    {
+      "epoch": 56.666666666666664,
+      "grad_norm": 2.9524314106234604,
+      "learning_rate": 7.727863905898261e-05,
+      "loss": 2.8417,
+      "step": 1020
+    },
+    {
+      "epoch": 57.22222222222222,
+      "grad_norm": 3.56533138794752,
+      "learning_rate": 7.72229458324376e-05,
+      "loss": 2.8254,
+      "step": 1030
+    },
+    {
+      "epoch": 57.77777777777778,
+      "grad_norm": 2.95726425482468,
+      "learning_rate": 7.716671037846008e-05,
+      "loss": 2.8686,
+      "step": 1040
+    },
+    {
+      "epoch": 58.333333333333336,
+      "grad_norm": 3.0130897103066743,
+      "learning_rate": 7.710993357521301e-05,
+      "loss": 2.8027,
+      "step": 1050
+    },
+    {
+      "epoch": 58.888888888888886,
+      "grad_norm": 2.916571225012564,
+      "learning_rate": 7.705261630931301e-05,
+      "loss": 2.8258,
+      "step": 1060
+    },
+    {
+      "epoch": 59.44444444444444,
+      "grad_norm": 3.004397507121576,
+      "learning_rate": 7.699475947581645e-05,
+      "loss": 2.8168,
+      "step": 1070
+    },
+    {
+      "epoch": 60.0,
+      "grad_norm": 2.824758644258909,
+      "learning_rate": 7.693636397820548e-05,
+      "loss": 2.8113,
+      "step": 1080
+    },
+    {
+      "epoch": 60.55555555555556,
+      "grad_norm": 2.8036972465433316,
+      "learning_rate": 7.687743072837396e-05,
+      "loss": 2.7665,
+      "step": 1090
+    },
+    {
+      "epoch": 61.111111111111114,
+      "grad_norm": 2.6295201575287646,
+      "learning_rate": 7.68179606466132e-05,
+      "loss": 2.8094,
+      "step": 1100
+    },
+    {
+      "epoch": 61.666666666666664,
+      "grad_norm": 2.7302271309013495,
+      "learning_rate": 7.675795466159753e-05,
+      "loss": 2.8014,
+      "step": 1110
+    },
+    {
+      "epoch": 62.22222222222222,
+      "grad_norm": 2.904590419975336,
+      "learning_rate": 7.669741371036993e-05,
+      "loss": 2.7912,
+      "step": 1120
+    },
+    {
+      "epoch": 62.77777777777778,
+      "grad_norm": 2.9374152638002746,
+      "learning_rate": 7.663633873832725e-05,
+      "loss": 2.7932,
+      "step": 1130
+    },
+    {
+      "epoch": 63.333333333333336,
+      "grad_norm": 2.706788331509178,
+      "learning_rate": 7.657473069920554e-05,
+      "loss": 2.7819,
+      "step": 1140
+    },
+    {
+      "epoch": 63.888888888888886,
+      "grad_norm": 2.6906334109291175,
+      "learning_rate": 7.651259055506515e-05,
+      "loss": 2.7956,
+      "step": 1150
+    },
+    {
+      "epoch": 64.44444444444444,
+      "grad_norm": 2.5220567451040634,
+      "learning_rate": 7.644991927627567e-05,
+      "loss": 2.7606,
+      "step": 1160
+    },
+    {
+      "epoch": 65.0,
+      "grad_norm": 3.0127377301511444,
+      "learning_rate": 7.638671784150078e-05,
+      "loss": 2.7321,
+      "step": 1170
+    },
+    {
+      "epoch": 65.55555555555556,
+      "grad_norm": 2.9139985187450828,
+      "learning_rate": 7.632298723768303e-05,
+      "loss": 2.7714,
+      "step": 1180
+    },
+    {
+      "epoch": 66.11111111111111,
+      "grad_norm": 2.5196427194939615,
+      "learning_rate": 7.625872846002834e-05,
+      "loss": 2.7589,
+      "step": 1190
+    },
+    {
+      "epoch": 66.66666666666667,
+      "grad_norm": 2.6769692561165805,
+      "learning_rate": 7.61939425119905e-05,
+      "loss": 2.7866,
+      "step": 1200
+    },
+    {
+      "epoch": 67.22222222222223,
+      "grad_norm": 2.9437128473678933,
+      "learning_rate": 7.612863040525552e-05,
+      "loss": 2.7357,
+      "step": 1210
+    },
+    {
+      "epoch": 67.77777777777777,
+      "grad_norm": 2.914308102973187,
+      "learning_rate": 7.606279315972582e-05,
+      "loss": 2.7684,
+      "step": 1220
+    },
+    {
+      "epoch": 68.33333333333333,
+      "grad_norm": 2.6750964994048214,
+      "learning_rate": 7.599643180350426e-05,
+      "loss": 2.7185,
+      "step": 1230
+    },
+    {
+      "epoch": 68.88888888888889,
+      "grad_norm": 3.02350279271795,
+      "learning_rate": 7.592954737287815e-05,
+      "loss": 2.7206,
+      "step": 1240
+    },
+    {
+      "epoch": 69.44444444444444,
+      "grad_norm": 2.608792542648536,
+      "learning_rate": 7.586214091230304e-05,
+      "loss": 2.7475,
+      "step": 1250
+    },
+    {
+      "epoch": 70.0,
+      "grad_norm": 2.628578199576311,
+      "learning_rate": 7.579421347438636e-05,
+      "loss": 2.7262,
+      "step": 1260
+    },
+    {
+      "epoch": 70.55555555555556,
+      "grad_norm": 2.6501663551657453,
+      "learning_rate": 7.57257661198711e-05,
+      "loss": 2.7211,
+      "step": 1270
+    },
+    {
+      "epoch": 71.11111111111111,
+      "grad_norm": 3.0838820166871947,
+      "learning_rate": 7.565679991761914e-05,
+      "loss": 2.7445,
+      "step": 1280
+    },
+    {
+      "epoch": 71.66666666666667,
+      "grad_norm": 3.338836767967965,
+      "learning_rate": 7.558731594459459e-05,
+      "loss": 2.7168,
+      "step": 1290
+    },
+    {
+      "epoch": 72.22222222222223,
+      "grad_norm": 3.3203686518689404,
+      "learning_rate": 7.551731528584699e-05,
+      "loss": 2.7302,
+      "step": 1300
+    },
+    {
+      "epoch": 72.77777777777777,
+      "grad_norm": 2.668338331145976,
+      "learning_rate": 7.544679903449437e-05,
+      "loss": 2.7044,
+      "step": 1310
+    },
+    {
+      "epoch": 73.33333333333333,
+      "grad_norm": 2.593806714276646,
+      "learning_rate": 7.537576829170611e-05,
+      "loss": 2.7227,
+      "step": 1320
+    },
+    {
+      "epoch": 73.88888888888889,
+      "grad_norm": 2.725919368936804,
+      "learning_rate": 7.530422416668585e-05,
+      "loss": 2.7161,
+      "step": 1330
+    },
+    {
+      "epoch": 74.44444444444444,
+      "grad_norm": 3.3534378460421874,
+      "learning_rate": 7.52321677766541e-05,
+      "loss": 2.6821,
+      "step": 1340
+    },
+    {
+      "epoch": 75.0,
+      "grad_norm": 2.6619907053234186,
+      "learning_rate": 7.515960024683081e-05,
+      "loss": 2.7196,
+      "step": 1350
+    },
+    {
+      "epoch": 75.55555555555556,
+      "grad_norm": 2.611514830209574,
+      "learning_rate": 7.508652271041779e-05,
+      "loss": 2.6739,
+      "step": 1360
+    },
+    {
+      "epoch": 76.11111111111111,
+      "grad_norm": 2.477041204258418,
+      "learning_rate": 7.501293630858103e-05,
+      "loss": 2.6829,
+      "step": 1370
+    },
+    {
+      "epoch": 76.66666666666667,
+      "grad_norm": 2.5499296066441386,
+      "learning_rate": 7.49388421904329e-05,
+      "loss": 2.7006,
+      "step": 1380
+    },
+    {
+      "epoch": 77.22222222222223,
+      "grad_norm": 3.044825720599332,
+      "learning_rate": 7.486424151301413e-05,
+      "loss": 2.6559,
+      "step": 1390
+    },
+    {
+      "epoch": 77.77777777777777,
+      "grad_norm": 2.920719509836431,
+      "learning_rate": 7.478913544127583e-05,
+      "loss": 2.7022,
+      "step": 1400
+    },
+    {
+      "epoch": 78.33333333333333,
+      "grad_norm": 2.60801186048116,
+      "learning_rate": 7.471352514806125e-05,
+      "loss": 2.6907,
+      "step": 1410
+    },
+    {
+      "epoch": 78.88888888888889,
+      "grad_norm": 2.9530436186703257,
+      "learning_rate": 7.463741181408744e-05,
+      "loss": 2.6958,
+      "step": 1420
+    },
+    {
+      "epoch": 79.44444444444444,
+      "grad_norm": 2.7345511243501317,
+      "learning_rate": 7.45607966279269e-05,
+      "loss": 2.6662,
+      "step": 1430
+    },
+    {
+      "epoch": 80.0,
+      "grad_norm": 3.1483501109828436,
+      "learning_rate": 7.448368078598894e-05,
+      "loss": 2.6674,
+      "step": 1440
+    },
+    {
+      "epoch": 80.55555555555556,
+      "grad_norm": 2.537373236134788,
+      "learning_rate": 7.440606549250101e-05,
+      "loss": 2.6653,
+      "step": 1450
+    },
+    {
+      "epoch": 81.11111111111111,
+      "grad_norm": 2.4904299670430103,
+      "learning_rate": 7.432795195948994e-05,
+      "loss": 2.6755,
+      "step": 1460
+    },
+    {
+      "epoch": 81.66666666666667,
+      "grad_norm": 2.5319617883652388,
+      "learning_rate": 7.424934140676293e-05,
+      "loss": 2.6476,
+      "step": 1470
+    },
+    {
+      "epoch": 82.22222222222223,
+      "grad_norm": 2.624148094359457,
+      "learning_rate": 7.41702350618886e-05,
+      "loss": 2.608,
+      "step": 1480
+    },
+    {
+      "epoch": 82.77777777777777,
+      "grad_norm": 3.0565164461596686,
+      "learning_rate": 7.409063416017778e-05,
+      "loss": 2.6412,
+      "step": 1490
+    },
+    {
+      "epoch": 83.33333333333333,
+      "grad_norm": 2.4970969510886007,
+      "learning_rate": 7.401053994466417e-05,
+      "loss": 2.6759,
+      "step": 1500
+    },
+    {
+      "epoch": 83.88888888888889,
+      "grad_norm": 2.47275729824135,
+      "learning_rate": 7.392995366608499e-05,
+      "loss": 2.6376,
+      "step": 1510
+    },
+    {
+      "epoch": 84.44444444444444,
+      "grad_norm": 2.8920639168309616,
+      "learning_rate": 7.384887658286145e-05,
+      "loss": 2.6204,
+      "step": 1520
+    },
+    {
+      "epoch": 85.0,
+      "grad_norm": 2.315314770714296,
+      "learning_rate": 7.376730996107905e-05,
+      "loss": 2.6212,
+      "step": 1530
+    },
+    {
+      "epoch": 85.55555555555556,
+      "grad_norm": 2.2970382736199353,
+      "learning_rate": 7.368525507446785e-05,
+      "loss": 2.6194,
+      "step": 1540
+    },
+    {
+      "epoch": 86.11111111111111,
+      "grad_norm": 2.643629543004188,
+      "learning_rate": 7.360271320438257e-05,
+      "loss": 2.6024,
+      "step": 1550
+    },
+    {
+      "epoch": 86.66666666666667,
+      "grad_norm": 2.602371685664894,
+      "learning_rate": 7.351968563978259e-05,
+      "loss": 2.6522,
+      "step": 1560
+    },
+    {
+      "epoch": 87.22222222222223,
+      "grad_norm": 2.695038270130096,
+      "learning_rate": 7.343617367721179e-05,
+      "loss": 2.6172,
+      "step": 1570
+    },
+    {
+      "epoch": 87.77777777777777,
+      "grad_norm": 2.679744886668,
+      "learning_rate": 7.335217862077831e-05,
+      "loss": 2.6107,
+      "step": 1580
+    },
+    {
+      "epoch": 88.33333333333333,
+      "grad_norm": 2.372212028431884,
+      "learning_rate": 7.326770178213425e-05,
+      "loss": 2.6308,
+      "step": 1590
+    },
+    {
+      "epoch": 88.88888888888889,
+      "grad_norm": 2.7099390307391578,
+      "learning_rate": 7.31827444804551e-05,
+      "loss": 2.628,
+      "step": 1600
+    },
+    {
+      "epoch": 89.44444444444444,
+      "grad_norm": 2.7981171646673073,
+      "learning_rate": 7.309730804241918e-05,
+      "loss": 2.6095,
+      "step": 1610
+    },
+    {
+      "epoch": 90.0,
+      "grad_norm": 2.8458678355622546,
+      "learning_rate": 7.301139380218691e-05,
+      "loss": 2.5848,
+      "step": 1620
+    },
+    {
+      "epoch": 90.55555555555556,
+      "grad_norm": 2.557825803168673,
+      "learning_rate": 7.292500310138003e-05,
+      "loss": 2.5921,
+      "step": 1630
+    },
+    {
+      "epoch": 91.11111111111111,
+      "grad_norm": 2.5860142998945372,
+      "learning_rate": 7.283813728906054e-05,
+      "loss": 2.6004,
+      "step": 1640
+    },
+    {
+      "epoch": 91.66666666666667,
+      "grad_norm": 2.4771778774158797,
+      "learning_rate": 7.275079772170975e-05,
+      "loss": 2.6083,
+      "step": 1650
+    },
+    {
+      "epoch": 92.22222222222223,
+      "grad_norm": 2.676200176774793,
+      "learning_rate": 7.266298576320701e-05,
+      "loss": 2.5751,
+      "step": 1660
+    },
+    {
+      "epoch": 92.77777777777777,
+      "grad_norm": 2.4295193040531884,
+      "learning_rate": 7.257470278480848e-05,
+      "loss": 2.5732,
+      "step": 1670
+    },
+    {
+      "epoch": 93.33333333333333,
+      "grad_norm": 2.311711769828301,
+      "learning_rate": 7.248595016512566e-05,
+      "loss": 2.5796,
+      "step": 1680
+    },
+    {
+      "epoch": 93.88888888888889,
+      "grad_norm": 2.7926463121079315,
+      "learning_rate": 7.239672929010391e-05,
+      "loss": 2.5632,
+      "step": 1690
+    },
+    {
+      "epoch": 94.44444444444444,
+      "grad_norm": 2.487622328007561,
+      "learning_rate": 7.230704155300076e-05,
+      "loss": 2.5688,
+      "step": 1700
+    },
+    {
+      "epoch": 95.0,
+      "grad_norm": 2.723537220445115,
+      "learning_rate": 7.221688835436417e-05,
+      "loss": 2.5487,
+      "step": 1710
+    },
+    {
+      "epoch": 95.55555555555556,
+      "grad_norm": 2.5459292475420425,
+      "learning_rate": 7.212627110201067e-05,
+      "loss": 2.5925,
+      "step": 1720
+    },
+    {
+      "epoch": 96.11111111111111,
+      "grad_norm": 3.032501906027199,
+      "learning_rate": 7.203519121100341e-05,
+      "loss": 2.5584,
+      "step": 1730
+    },
+    {
+      "epoch": 96.66666666666667,
+      "grad_norm": 2.395085281573907,
+      "learning_rate": 7.194365010362997e-05,
+      "loss": 2.5469,
+      "step": 1740
+    },
+    {
+      "epoch": 97.22222222222223,
+      "grad_norm": 2.4696802245801233,
+      "learning_rate": 7.185164920938025e-05,
+      "loss": 2.5525,
+      "step": 1750
+    },
+    {
+      "epoch": 97.77777777777777,
+      "grad_norm": 2.518117392635713,
+      "learning_rate": 7.175918996492408e-05,
+      "loss": 2.5746,
+      "step": 1760
+    },
+    {
+      "epoch": 98.33333333333333,
+      "grad_norm": 2.7831844503211203,
+      "learning_rate": 7.166627381408878e-05,
+      "loss": 2.5406,
+      "step": 1770
+    },
+    {
+      "epoch": 98.88888888888889,
+      "grad_norm": 2.411083469681673,
+      "learning_rate": 7.157290220783669e-05,
+      "loss": 2.5569,
+      "step": 1780
+    },
+    {
+      "epoch": 99.44444444444444,
+      "grad_norm": 2.3943866729650938,
+      "learning_rate": 7.147907660424242e-05,
+      "loss": 2.5579,
+      "step": 1790
+    },
+    {
+      "epoch": 100.0,
+      "grad_norm": 2.5753605673476803,
+      "learning_rate": 7.138479846847017e-05,
+      "loss": 2.5144,
+      "step": 1800
+    },
+    {
+      "epoch": 100.55555555555556,
+      "grad_norm": 2.331114826936834,
+      "learning_rate": 7.129006927275073e-05,
+      "loss": 2.5264,
+      "step": 1810
+    },
+    {
+      "epoch": 101.11111111111111,
+      "grad_norm": 2.387756122212614,
+      "learning_rate": 7.119489049635866e-05,
+      "loss": 2.5138,
+      "step": 1820
+    },
+    {
+      "epoch": 101.66666666666667,
+      "grad_norm": 2.63356872021627,
+      "learning_rate": 7.109926362558899e-05,
+      "loss": 2.5238,
+      "step": 1830
+    },
+    {
+      "epoch": 102.22222222222223,
+      "grad_norm": 2.484113571519236,
+      "learning_rate": 7.10031901537342e-05,
+      "loss": 2.5437,
+      "step": 1840
+    },
+    {
+      "epoch": 102.77777777777777,
+      "grad_norm": 2.6288736009766827,
+      "learning_rate": 7.090667158106076e-05,
+      "loss": 2.4851,
+      "step": 1850
+    },
+    {
+      "epoch": 103.33333333333333,
+      "grad_norm": 2.480369169383998,
+      "learning_rate": 7.080970941478578e-05,
+      "loss": 2.5129,
+      "step": 1860
+    },
+    {
+      "epoch": 103.88888888888889,
+      "grad_norm": 2.4342795396140606,
+      "learning_rate": 7.071230516905344e-05,
+      "loss": 2.5178,
+      "step": 1870
+    },
+    {
+      "epoch": 104.44444444444444,
+      "grad_norm": 2.436827835914139,
+      "learning_rate": 7.061446036491134e-05,
+      "loss": 2.5087,
+      "step": 1880
+    },
+    {
+      "epoch": 105.0,
+      "grad_norm": 2.6800788615850815,
+      "learning_rate": 7.051617653028682e-05,
+      "loss": 2.5161,
+      "step": 1890
+    },
+    {
+      "epoch": 105.55555555555556,
+      "grad_norm": 2.672615796284524,
+      "learning_rate": 7.041745519996298e-05,
+      "loss": 2.4889,
+      "step": 1900
+    },
+    {
+      "epoch": 106.11111111111111,
+      "grad_norm": 2.644439740760032,
+      "learning_rate": 7.031829791555481e-05,
+      "loss": 2.505,
+      "step": 1910
+    },
+    {
+      "epoch": 106.66666666666667,
+      "grad_norm": 2.5614743855208553,
+      "learning_rate": 7.021870622548507e-05,
+      "loss": 2.4728,
+      "step": 1920
+    },
+    {
+      "epoch": 107.22222222222223,
+      "grad_norm": 2.5034460159555003,
+      "learning_rate": 7.011868168496013e-05,
+      "loss": 2.5029,
+      "step": 1930
+    },
+    {
+      "epoch": 107.77777777777777,
+      "grad_norm": 2.492101590101853,
+      "learning_rate": 7.001822585594565e-05,
+      "loss": 2.5118,
+      "step": 1940
+    },
+    {
+      "epoch": 108.33333333333333,
+      "grad_norm": 2.58473228189801,
+      "learning_rate": 6.991734030714228e-05,
+      "loss": 2.5133,
+      "step": 1950
+    },
+    {
+      "epoch": 108.88888888888889,
+      "grad_norm": 2.4771004943852093,
+      "learning_rate": 6.981602661396103e-05,
+      "loss": 2.4884,
+      "step": 1960
+    },
+    {
+      "epoch": 109.44444444444444,
+      "grad_norm": 2.5509968735134776,
+      "learning_rate": 6.971428635849877e-05,
+      "loss": 2.491,
+      "step": 1970
+    },
+    {
+      "epoch": 110.0,
+      "grad_norm": 2.412647709252974,
+      "learning_rate": 6.961212112951348e-05,
+      "loss": 2.4943,
+      "step": 1980
+    },
+    {
+      "epoch": 110.55555555555556,
+      "grad_norm": 2.52315470052705,
+      "learning_rate": 6.950953252239947e-05,
+      "loss": 2.4613,
+      "step": 1990
+    },
+    {
+      "epoch": 111.11111111111111,
+      "grad_norm": 2.437207424504491,
+      "learning_rate": 6.940652213916241e-05,
+      "loss": 2.4327,
+      "step": 2000
+    },
+    {
+      "epoch": 111.66666666666667,
+      "grad_norm": 2.53179125238401,
+      "learning_rate": 6.930309158839443e-05,
+      "loss": 2.4594,
+      "step": 2010
+    },
+    {
+      "epoch": 112.22222222222223,
+      "grad_norm": 2.4974060906977558,
+      "learning_rate": 6.919924248524884e-05,
+      "loss": 2.4662,
+      "step": 2020
+    },
+    {
+      "epoch": 112.77777777777777,
+      "grad_norm": 2.5075943040847286,
+      "learning_rate": 6.9094976451415e-05,
+      "loss": 2.4576,
+      "step": 2030
+    },
+    {
+      "epoch": 113.33333333333333,
+      "grad_norm": 2.588050186708306,
+      "learning_rate": 6.899029511509305e-05,
+      "loss": 2.4667,
+      "step": 2040
+    },
+    {
+      "epoch": 113.88888888888889,
+      "grad_norm": 2.77082551271487,
+      "learning_rate": 6.888520011096835e-05,
+      "loss": 2.4616,
+      "step": 2050
+    },
+    {
+      "epoch": 114.44444444444444,
+      "grad_norm": 2.4338006533233547,
+      "learning_rate": 6.877969308018608e-05,
+      "loss": 2.4678,
+      "step": 2060
+    },
+    {
+      "epoch": 115.0,
+      "grad_norm": 2.5659851007363486,
+      "learning_rate": 6.867377567032551e-05,
+      "loss": 2.4622,
+      "step": 2070
+    },
+    {
+      "epoch": 115.55555555555556,
+      "grad_norm": 2.570343017396958,
+      "learning_rate": 6.856744953537433e-05,
+      "loss": 2.4239,
+      "step": 2080
+    },
+    {
+      "epoch": 116.11111111111111,
+      "grad_norm": 2.6921616865748197,
+      "learning_rate": 6.846071633570285e-05,
+      "loss": 2.4138,
+      "step": 2090
+    },
+    {
+      "epoch": 116.66666666666667,
+      "grad_norm": 2.5947506364732846,
+      "learning_rate": 6.835357773803795e-05,
+      "loss": 2.4439,
+      "step": 2100
+    },
+    {
+      "epoch": 117.22222222222223,
+      "grad_norm": 2.5565455967645447,
+      "learning_rate": 6.824603541543722e-05,
+      "loss": 2.4454,
+      "step": 2110
+    },
+    {
+      "epoch": 117.77777777777777,
+      "grad_norm": 2.3438651501660854,
+      "learning_rate": 6.81380910472627e-05,
+      "loss": 2.4189,
+      "step": 2120
+    },
+    {
+      "epoch": 118.33333333333333,
+      "grad_norm": 2.5794882204659717,
+      "learning_rate": 6.802974631915469e-05,
+      "loss": 2.4037,
+      "step": 2130
+    },
+    {
+      "epoch": 118.88888888888889,
+      "grad_norm": 2.443535985136445,
+      "learning_rate": 6.79210029230055e-05,
+      "loss": 2.4181,
+      "step": 2140
+    },
+    {
+      "epoch": 119.44444444444444,
+      "grad_norm": 2.4699802472023586,
+      "learning_rate": 6.781186255693291e-05,
+      "loss": 2.4031,
+      "step": 2150
+    },
+    {
+      "epoch": 120.0,
+      "grad_norm": 2.5895847749181793,
+      "learning_rate": 6.770232692525372e-05,
+      "loss": 2.4195,
+      "step": 2160
+    },
+    {
+      "epoch": 120.55555555555556,
+      "grad_norm": 2.508341605678702,
+      "learning_rate": 6.759239773845719e-05,
+      "loss": 2.3723,
+      "step": 2170
+    },
+    {
+      "epoch": 121.11111111111111,
+      "grad_norm": 2.3971434839469654,
+      "learning_rate": 6.74820767131782e-05,
+      "loss": 2.416,
+      "step": 2180
+    },
+    {
+      "epoch": 121.66666666666667,
+      "grad_norm": 2.4191773187462164,
+      "learning_rate": 6.73713655721705e-05,
+      "loss": 2.4258,
+      "step": 2190
+    },
+    {
+      "epoch": 122.22222222222223,
+      "grad_norm": 2.415767228256262,
+      "learning_rate": 6.726026604427991e-05,
+      "loss": 2.4211,
+      "step": 2200
+    },
+    {
+      "epoch": 122.77777777777777,
+      "grad_norm": 2.500744994739976,
+      "learning_rate": 6.714877986441712e-05,
+      "loss": 2.368,
+      "step": 2210
+    },
+    {
+      "epoch": 123.33333333333333,
+      "grad_norm": 2.6845336334287206,
+      "learning_rate": 6.703690877353082e-05,
+      "loss": 2.3583,
+      "step": 2220
+    },
+    {
+      "epoch": 123.88888888888889,
+      "grad_norm": 2.516495074386521,
+      "learning_rate": 6.69246545185803e-05,
+      "loss": 2.3652,
+      "step": 2230
+    },
+    {
+      "epoch": 124.44444444444444,
+      "grad_norm": 2.5654752253167183,
+      "learning_rate": 6.681201885250831e-05,
+      "loss": 2.3857,
+      "step": 2240
+    },
+    {
+      "epoch": 125.0,
+      "grad_norm": 2.3633257396703335,
+      "learning_rate": 6.66990035342137e-05,
+      "loss": 2.4044,
+      "step": 2250
+    },
+    {
+      "epoch": 125.55555555555556,
+      "grad_norm": 2.6321922046954924,
+      "learning_rate": 6.658561032852383e-05,
+      "loss": 2.3752,
+      "step": 2260
+    },
+    {
+      "epoch": 126.11111111111111,
+      "grad_norm": 2.502997794473052,
+      "learning_rate": 6.647184100616713e-05,
+      "loss": 2.3678,
+      "step": 2270
+    },
+    {
+      "epoch": 126.66666666666667,
+      "grad_norm": 2.4990096991855664,
+      "learning_rate": 6.635769734374537e-05,
+      "loss": 2.3774,
+      "step": 2280
+    },
+    {
+      "epoch": 127.22222222222223,
+      "grad_norm": 2.7434090613214206,
+      "learning_rate": 6.6243181123706e-05,
+      "loss": 2.3758,
+      "step": 2290
+    },
+    {
+      "epoch": 127.77777777777777,
+      "grad_norm": 2.510708666606988,
+      "learning_rate": 6.612829413431417e-05,
+      "loss": 2.375,
+      "step": 2300
+    },
+    {
+      "epoch": 128.33333333333334,
+      "grad_norm": 2.5624530136638812,
+      "learning_rate": 6.6013038169625e-05,
+      "loss": 2.3497,
+      "step": 2310
+    },
+    {
+      "epoch": 128.88888888888889,
+      "grad_norm": 2.660891079903747,
+      "learning_rate": 6.58974150294554e-05,
+      "loss": 2.348,
+      "step": 2320
+    },
+    {
+      "epoch": 129.44444444444446,
+      "grad_norm": 2.6030760248712994,
+      "learning_rate": 6.578142651935609e-05,
+      "loss": 2.3151,
+      "step": 2330
+    },
+    {
+      "epoch": 130.0,
+      "grad_norm": 2.580961389547689,
+      "learning_rate": 6.566507445058325e-05,
+      "loss": 2.3515,
+      "step": 2340
+    },
+    {
+      "epoch": 130.55555555555554,
+      "grad_norm": 2.5115563323811547,
+      "learning_rate": 6.554836064007041e-05,
+      "loss": 2.3381,
+      "step": 2350
+    },
+    {
+      "epoch": 131.11111111111111,
+      "grad_norm": 2.461051429654467,
+      "learning_rate": 6.543128691039996e-05,
+      "loss": 2.3535,
+      "step": 2360
+    },
+    {
+      "epoch": 131.66666666666666,
+      "grad_norm": 2.5533257917918624,
+      "learning_rate": 6.531385508977472e-05,
+      "loss": 2.3379,
+      "step": 2370
+    },
+    {
+      "epoch": 132.22222222222223,
+      "grad_norm": 2.3759297006848574,
+      "learning_rate": 6.519606701198941e-05,
+      "loss": 2.3239,
+      "step": 2380
+    },
+    {
+      "epoch": 132.77777777777777,
+      "grad_norm": 2.4962032097111733,
+      "learning_rate": 6.507792451640202e-05,
+      "loss": 2.3258,
+      "step": 2390
+    },
+    {
+      "epoch": 133.33333333333334,
+      "grad_norm": 2.5124855117000546,
+      "learning_rate": 6.495942944790501e-05,
+      "loss": 2.303,
+      "step": 2400
+    },
+    {
+      "epoch": 133.88888888888889,
+      "grad_norm": 2.525362585491815,
+      "learning_rate": 6.48405836568966e-05,
+      "loss": 2.3077,
+      "step": 2410
+    },
+    {
+      "epoch": 134.44444444444446,
+      "grad_norm": 2.6621767232860143,
+      "learning_rate": 6.472138899925184e-05,
+      "loss": 2.3063,
+      "step": 2420
+    },
+    {
+      "epoch": 135.0,
+      "grad_norm": 2.6511934507994717,
+      "learning_rate": 6.460184733629356e-05,
+      "loss": 2.3144,
+      "step": 2430
+    },
+    {
+      "epoch": 135.55555555555554,
+      "grad_norm": 2.942468864127099,
+      "learning_rate": 6.448196053476343e-05,
+      "loss": 2.2962,
+      "step": 2440
+    },
+    {
+      "epoch": 136.11111111111111,
+      "grad_norm": 2.540090124078266,
+      "learning_rate": 6.436173046679271e-05,
+      "loss": 2.3061,
+      "step": 2450
+    },
+    {
+      "epoch": 136.66666666666666,
+      "grad_norm": 2.455087447774662,
+      "learning_rate": 6.424115900987306e-05,
+      "loss": 2.259,
+      "step": 2460
+    },
+    {
+      "epoch": 137.22222222222223,
+      "grad_norm": 2.5124950010335354,
+      "learning_rate": 6.41202480468272e-05,
+      "loss": 2.2874,
+      "step": 2470
+    },
+    {
+      "epoch": 137.77777777777777,
+      "grad_norm": 3.3014010778521383,
+      "learning_rate": 6.399899946577953e-05,
+      "loss": 2.2855,
+      "step": 2480
+    },
+    {
+      "epoch": 138.33333333333334,
+      "grad_norm": 2.5140058151881925,
+      "learning_rate": 6.387741516012663e-05,
+      "loss": 2.2994,
+      "step": 2490
+    },
+    {
+      "epoch": 138.88888888888889,
+      "grad_norm": 2.6129392847773207,
+      "learning_rate": 6.375549702850768e-05,
+      "loss": 2.2986,
+      "step": 2500
+    },
+    {
+      "epoch": 139.44444444444446,
+      "grad_norm": 2.7750698390111124,
+      "learning_rate": 6.363324697477485e-05,
+      "loss": 2.2452,
+      "step": 2510
+    },
+    {
+      "epoch": 140.0,
+      "grad_norm": 2.460977027212322,
+      "learning_rate": 6.351066690796354e-05,
+      "loss": 2.2976,
+      "step": 2520
+    },
+    {
+      "epoch": 140.55555555555554,
+      "grad_norm": 2.5859659890625273,
+      "learning_rate": 6.338775874226258e-05,
+      "loss": 2.2974,
+      "step": 2530
+    },
+    {
+      "epoch": 141.11111111111111,
+      "grad_norm": 2.4983937825665783,
+      "learning_rate": 6.326452439698431e-05,
+      "loss": 2.2389,
+      "step": 2540
+    },
+    {
+      "epoch": 141.66666666666666,
+      "grad_norm": 2.5866033613022217,
+      "learning_rate": 6.314096579653469e-05,
+      "loss": 2.2452,
+      "step": 2550
+    },
+    {
+      "epoch": 142.22222222222223,
+      "grad_norm": 2.5427696982207606,
+      "learning_rate": 6.301708487038312e-05,
+      "loss": 2.2556,
+      "step": 2560
+    },
+    {
+      "epoch": 142.77777777777777,
+      "grad_norm": 2.5599564745302326,
+      "learning_rate": 6.289288355303245e-05,
+      "loss": 2.2533,
+      "step": 2570
+    },
+    {
+      "epoch": 143.33333333333334,
+      "grad_norm": 2.781581216013528,
+      "learning_rate": 6.276836378398864e-05,
+      "loss": 2.2404,
+      "step": 2580
+    },
+    {
+      "epoch": 143.88888888888889,
+      "grad_norm": 2.6558024309856334,
+      "learning_rate": 6.264352750773059e-05,
+      "loss": 2.2415,
+      "step": 2590
+    },
+    {
+      "epoch": 144.44444444444446,
+      "grad_norm": 2.7120590308581285,
+      "learning_rate": 6.251837667367967e-05,
+      "loss": 2.2222,
+      "step": 2600
+    },
+    {
+      "epoch": 145.0,
+      "grad_norm": 2.5903975180959797,
+      "learning_rate": 6.239291323616937e-05,
+      "loss": 2.2345,
+      "step": 2610
+    },
+    {
+      "epoch": 145.55555555555554,
+      "grad_norm": 2.725408708912341,
+      "learning_rate": 6.226713915441472e-05,
+      "loss": 2.253,
+      "step": 2620
+    },
+    {
+      "epoch": 146.11111111111111,
+      "grad_norm": 2.779857329680152,
+      "learning_rate": 6.214105639248174e-05,
+      "loss": 2.2335,
+      "step": 2630
+    },
+    {
+      "epoch": 146.66666666666666,
+      "grad_norm": 2.5558277816720456,
+      "learning_rate": 6.20146669192567e-05,
+      "loss": 2.2011,
+      "step": 2640
+    },
+    {
+      "epoch": 147.22222222222223,
+      "grad_norm": 2.9247900145708745,
+      "learning_rate": 6.188797270841548e-05,
+      "loss": 2.2058,
+      "step": 2650
+    },
+    {
+      "epoch": 147.77777777777777,
+      "grad_norm": 2.5994642806169916,
+      "learning_rate": 6.176097573839264e-05,
+      "loss": 2.193,
+      "step": 2660
+    },
+    {
+      "epoch": 148.33333333333334,
+      "grad_norm": 2.6993581291461135,
+      "learning_rate": 6.16336779923506e-05,
+      "loss": 2.1991,
+      "step": 2670
+    },
+    {
+      "epoch": 148.88888888888889,
+      "grad_norm": 2.6094487345436908,
+      "learning_rate": 6.150608145814863e-05,
+      "loss": 2.215,
+      "step": 2680
+    },
+    {
+      "epoch": 149.44444444444446,
+      "grad_norm": 2.6594028386273796,
+      "learning_rate": 6.137818812831183e-05,
+      "loss": 2.1927,
+      "step": 2690
+    },
+    {
+      "epoch": 150.0,
+      "grad_norm": 2.59346199734705,
+      "learning_rate": 6.125000000000001e-05,
+      "loss": 2.1894,
+      "step": 2700
+    },
+    {
+      "epoch": 150.55555555555554,
+      "grad_norm": 2.729285773864419,
+      "learning_rate": 6.11215190749765e-05,
+      "loss": 2.1905,
+      "step": 2710
+    },
+    {
+      "epoch": 151.11111111111111,
+      "grad_norm": 2.6021345729583074,
+      "learning_rate": 6.0992747359576896e-05,
+      "loss": 2.1727,
+      "step": 2720
+    },
+    {
+      "epoch": 151.66666666666666,
+      "grad_norm": 2.7536304958073416,
+      "learning_rate": 6.0863686864677734e-05,
+      "loss": 2.1743,
+      "step": 2730
+    },
+    {
+      "epoch": 152.22222222222223,
+      "grad_norm": 2.6883638457403305,
+      "learning_rate": 6.0734339605665026e-05,
+      "loss": 2.2146,
+      "step": 2740
+    },
+    {
+      "epoch": 152.77777777777777,
+      "grad_norm": 2.5219340377760786,
+      "learning_rate": 6.060470760240294e-05,
+      "loss": 2.1558,
+      "step": 2750
+    },
+    {
+      "epoch": 153.33333333333334,
+      "grad_norm": 2.648612317642482,
+      "learning_rate": 6.047479287920208e-05,
+      "loss": 2.1546,
+      "step": 2760
+    },
+    {
+      "epoch": 153.88888888888889,
+      "grad_norm": 2.937206010107733,
+      "learning_rate": 6.0344597464787954e-05,
+      "loss": 2.1807,
+      "step": 2770
+    },
+    {
+      "epoch": 154.44444444444446,
+      "grad_norm": 2.7347525199769027,
+      "learning_rate": 6.021412339226936e-05,
+      "loss": 2.1637,
+      "step": 2780
+    },
+    {
+      "epoch": 155.0,
+      "grad_norm": 2.6357110704674866,
+      "learning_rate": 6.008337269910651e-05,
+      "loss": 2.1523,
+      "step": 2790
+    },
+    {
+      "epoch": 155.55555555555554,
+      "grad_norm": 2.6896946839424674,
+      "learning_rate": 5.9952347427079305e-05,
+      "loss": 2.139,
+      "step": 2800
+    },
+    {
+      "epoch": 156.11111111111111,
+      "grad_norm": 2.696004143029717,
+      "learning_rate": 5.9821049622255403e-05,
+      "loss": 2.1411,
+      "step": 2810
+    },
+    {
+      "epoch": 156.66666666666666,
+      "grad_norm": 2.6762609344216286,
+      "learning_rate": 5.968948133495832e-05,
+      "loss": 2.1584,
+      "step": 2820
+    },
+    {
+      "epoch": 157.22222222222223,
+      "grad_norm": 2.7967418447157195,
+      "learning_rate": 5.955764461973535e-05,
+      "loss": 2.1804,
+      "step": 2830
+    },
+    {
+      "epoch": 157.77777777777777,
+      "grad_norm": 2.5755792244108977,
+      "learning_rate": 5.9425541535325495e-05,
+      "loss": 2.1291,
+      "step": 2840
+    },
+    {
+      "epoch": 158.33333333333334,
+      "grad_norm": 2.6970589549601547,
+      "learning_rate": 5.929317414462738e-05,
+      "loss": 2.1518,
+      "step": 2850
+    },
+    {
+      "epoch": 158.88888888888889,
+      "grad_norm": 2.654103028529104,
+      "learning_rate": 5.916054451466694e-05,
+      "loss": 2.1509,
+      "step": 2860
+    },
+    {
+      "epoch": 159.44444444444446,
+      "grad_norm": 2.7952822620170084,
+      "learning_rate": 5.9027654716565247e-05,
+      "loss": 2.121,
+      "step": 2870
+    },
+    {
+      "epoch": 160.0,
+      "grad_norm": 2.7726945768551965,
+      "learning_rate": 5.889450682550605e-05,
+      "loss": 2.1068,
+      "step": 2880
+    },
+    {
+      "epoch": 160.55555555555554,
+      "grad_norm": 2.67274933725095,
+      "learning_rate": 5.8761102920703476e-05,
+      "loss": 2.1176,
+      "step": 2890
+    },
+    {
+      "epoch": 161.11111111111111,
+      "grad_norm": 2.7725922491964003,
+      "learning_rate": 5.862744508536953e-05,
+      "loss": 2.1276,
+      "step": 2900
+    },
+    {
+      "epoch": 161.66666666666666,
+      "grad_norm": 2.950253498371094,
+      "learning_rate": 5.84935354066815e-05,
+      "loss": 2.1315,
+      "step": 2910
+    },
+    {
+      "epoch": 162.22222222222223,
+      "grad_norm": 2.6831662033190917,
+      "learning_rate": 5.835937597574944e-05,
+      "loss": 2.1042,
+      "step": 2920
+    },
+    {
+      "epoch": 162.77777777777777,
+      "grad_norm": 2.737649929173668,
+      "learning_rate": 5.822496888758352e-05,
+      "loss": 2.1018,
+      "step": 2930
+    },
+    {
+      "epoch": 163.33333333333334,
+      "grad_norm": 2.7363132255851053,
+      "learning_rate": 5.8090316241061225e-05,
+      "loss": 2.0902,
+      "step": 2940
+    },
+    {
+      "epoch": 163.88888888888889,
+      "grad_norm": 2.7390503068686285,
+      "learning_rate": 5.7955420138894706e-05,
+      "loss": 2.0631,
+      "step": 2950
+    },
+    {
+      "epoch": 164.44444444444446,
+      "grad_norm": 2.9877471247019134,
+      "learning_rate": 5.7820282687597816e-05,
+      "loss": 2.1025,
+      "step": 2960
+    },
+    {
+      "epoch": 165.0,
+      "grad_norm": 2.777542643661771,
+      "learning_rate": 5.7684905997453294e-05,
+      "loss": 2.0842,
+      "step": 2970
+    },
+    {
+      "epoch": 165.55555555555554,
+      "grad_norm": 2.6592661170768723,
+      "learning_rate": 5.7549292182479784e-05,
+      "loss": 2.0973,
+      "step": 2980
+    },
+    {
+      "epoch": 166.11111111111111,
+      "grad_norm": 2.7380918690362583,
+      "learning_rate": 5.741344336039887e-05,
+      "loss": 2.0579,
+      "step": 2990
+    },
+    {
+      "epoch": 166.66666666666666,
+      "grad_norm": 2.7891854224378387,
+      "learning_rate": 5.727736165260188e-05,
+      "loss": 2.0898,
+      "step": 3000
+    },
+    {
+      "epoch": 167.22222222222223,
+      "grad_norm": 2.7781047395497334,
+      "learning_rate": 5.714104918411696e-05,
+      "loss": 2.0221,
+      "step": 3010
+    },
+    {
+      "epoch": 167.77777777777777,
+      "grad_norm": 2.77884741745799,
+      "learning_rate": 5.7004508083575645e-05,
+      "loss": 2.0534,
+      "step": 3020
+    },
+    {
+      "epoch": 168.33333333333334,
+      "grad_norm": 2.779678586588523,
+      "learning_rate": 5.686774048317986e-05,
+      "loss": 2.0387,
+      "step": 3030
+    },
+    {
+      "epoch": 168.88888888888889,
+      "grad_norm": 2.671867950608619,
+      "learning_rate": 5.6730748518668455e-05,
+      "loss": 2.0441,
+      "step": 3040
+    },
+    {
+      "epoch": 169.44444444444446,
+      "grad_norm": 2.7957777716113426,
+      "learning_rate": 5.6593534329283935e-05,
+      "loss": 2.0442,
+      "step": 3050
+    },
+    {
+      "epoch": 170.0,
+      "grad_norm": 3.0244865864528645,
+      "learning_rate": 5.645610005773901e-05,
+      "loss": 2.0428,
+      "step": 3060
+    },
+    {
+      "epoch": 170.55555555555554,
+      "grad_norm": 2.8989285364149153,
+      "learning_rate": 5.631844785018314e-05,
+      "loss": 2.0317,
+      "step": 3070
+    },
+    {
+      "epoch": 171.11111111111111,
+      "grad_norm": 2.7839357930924047,
+      "learning_rate": 5.6180579856169076e-05,
+      "loss": 1.9943,
+      "step": 3080
+    },
+    {
+      "epoch": 171.66666666666666,
+      "grad_norm": 2.7333952102015155,
+      "learning_rate": 5.604249822861923e-05,
+      "loss": 2.0095,
+      "step": 3090
+    },
+    {
+      "epoch": 172.22222222222223,
+      "grad_norm": 2.6062779120910555,
+      "learning_rate": 5.5904205123792096e-05,
+      "loss": 2.0379,
+      "step": 3100
+    },
+    {
+      "epoch": 172.77777777777777,
+      "grad_norm": 2.821654446960896,
+      "learning_rate": 5.576570270124853e-05,
+      "loss": 2.0195,
+      "step": 3110
+    },
+    {
+      "epoch": 173.33333333333334,
+      "grad_norm": 2.8432806277411977,
+      "learning_rate": 5.562699312381807e-05,
+      "loss": 2.0468,
+      "step": 3120
+    },
+    {
+      "epoch": 173.88888888888889,
+      "grad_norm": 2.827441053779487,
+      "learning_rate": 5.548807855756516e-05,
+      "loss": 1.9872,
+      "step": 3130
+    },
+    {
+      "epoch": 174.44444444444446,
+      "grad_norm": 2.7016133397946396,
+      "learning_rate": 5.53489611717553e-05,
+      "loss": 2.0109,
+      "step": 3140
+    },
+    {
+      "epoch": 175.0,
+      "grad_norm": 2.6683785387729535,
+      "learning_rate": 5.5209643138821204e-05,
+      "loss": 1.987,
+      "step": 3150
+    },
+    {
+      "epoch": 175.55555555555554,
+      "grad_norm": 2.8009458783445864,
+      "learning_rate": 5.50701266343288e-05,
+      "loss": 1.9976,
+      "step": 3160
+    },
+    {
+      "epoch": 176.11111111111111,
+      "grad_norm": 2.771999534381205,
+      "learning_rate": 5.4930413836943396e-05,
+      "loss": 1.9937,
+      "step": 3170
+    },
+    {
+      "epoch": 176.66666666666666,
+      "grad_norm": 2.801448300997254,
+      "learning_rate": 5.47905069283955e-05,
+      "loss": 1.9611,
+      "step": 3180
+    },
+    {
+      "epoch": 177.22222222222223,
+      "grad_norm": 2.853988493395615,
+      "learning_rate": 5.465040809344691e-05,
+      "loss": 1.9403,
+      "step": 3190
+    },
+    {
+      "epoch": 177.77777777777777,
+      "grad_norm": 2.7786358323970433,
+      "learning_rate": 5.451011951985641e-05,
+      "loss": 1.9608,
+      "step": 3200
+    },
+    {
+      "epoch": 178.33333333333334,
+      "grad_norm": 2.815801928652619,
+      "learning_rate": 5.436964339834579e-05,
+      "loss": 1.9615,
+      "step": 3210
+    },
+    {
+      "epoch": 178.88888888888889,
+      "grad_norm": 2.8748830688580944,
+      "learning_rate": 5.4228981922565525e-05,
+      "loss": 1.9721,
+      "step": 3220
+    },
+    {
+      "epoch": 179.44444444444446,
+      "grad_norm": 2.780207835188233,
+      "learning_rate": 5.408813728906053e-05,
+      "loss": 1.9741,
+      "step": 3230
+    },
+    {
+      "epoch": 180.0,
+      "grad_norm": 2.9000987923993673,
+      "learning_rate": 5.394711169723593e-05,
+      "loss": 1.9858,
+      "step": 3240
+    },
+    {
+      "epoch": 180.55555555555554,
+      "grad_norm": 2.7927622472009443,
+      "learning_rate": 5.380590734932262e-05,
+      "loss": 1.9475,
+      "step": 3250
+    },
+    {
+      "epoch": 181.11111111111111,
+      "grad_norm": 2.724519942808679,
+      "learning_rate": 5.366452645034293e-05,
+      "loss": 1.9275,
+      "step": 3260
+    },
+    {
+      "epoch": 181.66666666666666,
+      "grad_norm": 2.7739204953275416,
+      "learning_rate": 5.352297120807619e-05,
+      "loss": 1.9331,
+      "step": 3270
+    },
+    {
+      "epoch": 182.22222222222223,
+      "grad_norm": 2.8751499302787726,
+      "learning_rate": 5.3381243833024236e-05,
+      "loss": 1.8986,
+      "step": 3280
+    },
+    {
+      "epoch": 182.77777777777777,
+      "grad_norm": 2.8279795741000497,
+      "learning_rate": 5.3239346538376895e-05,
+      "loss": 1.9376,
+      "step": 3290
+    },
+    {
+      "epoch": 183.33333333333334,
+      "grad_norm": 3.099559992665463,
+      "learning_rate": 5.3097281539977465e-05,
+      "loss": 1.9144,
+      "step": 3300
+    },
+    {
+      "epoch": 183.88888888888889,
+      "grad_norm": 2.723137921015265,
+      "learning_rate": 5.2955051056288006e-05,
+      "loss": 1.9104,
+      "step": 3310
+    },
+    {
+      "epoch": 184.44444444444446,
+      "grad_norm": 2.85696142506152,
+      "learning_rate": 5.281265730835483e-05,
+      "loss": 1.9126,
+      "step": 3320
+    },
+    {
+      "epoch": 185.0,
+      "grad_norm": 2.7468657405553993,
+      "learning_rate": 5.267010251977371e-05,
+      "loss": 1.896,
+      "step": 3330
+    },
+    {
+      "epoch": 185.55555555555554,
+      "grad_norm": 2.901436706346961,
+      "learning_rate": 5.252738891665525e-05,
+      "loss": 1.9171,
+      "step": 3340
+    },
+    {
+      "epoch": 186.11111111111111,
+      "grad_norm": 2.9327788052785224,
+      "learning_rate": 5.2384518727590035e-05,
+      "loss": 1.9078,
+      "step": 3350
+    },
+    {
+      "epoch": 186.66666666666666,
+      "grad_norm": 2.8428408668145253,
+      "learning_rate": 5.224149418361388e-05,
+      "loss": 1.9112,
+      "step": 3360
+    },
+    {
+      "epoch": 187.22222222222223,
+      "grad_norm": 2.9078664898664424,
+      "learning_rate": 5.2098317518173004e-05,
+      "loss": 1.9064,
+      "step": 3370
+    },
+    {
+      "epoch": 187.77777777777777,
+      "grad_norm": 2.8603482283823176,
+      "learning_rate": 5.195499096708908e-05,
+      "loss": 1.8736,
+      "step": 3380
+    },
+    {
+      "epoch": 188.33333333333334,
+      "grad_norm": 2.955136053370996,
+      "learning_rate": 5.1811516768524385e-05,
+      "loss": 1.8711,
+      "step": 3390
+    },
+    {
+      "epoch": 188.88888888888889,
+      "grad_norm": 2.9564016420211434,
+      "learning_rate": 5.166789716294686e-05,
+      "loss": 1.8655,
+      "step": 3400
+    },
+    {
+      "epoch": 189.44444444444446,
+      "grad_norm": 2.7829976109704506,
+      "learning_rate": 5.152413439309508e-05,
+      "loss": 1.8722,
+      "step": 3410
+    },
+    {
+      "epoch": 190.0,
+      "grad_norm": 2.9193162781271718,
+      "learning_rate": 5.1380230703943226e-05,
+      "loss": 1.8642,
+      "step": 3420
+    },
+    {
+      "epoch": 190.55555555555554,
+      "grad_norm": 2.7994391765251776,
+      "learning_rate": 5.1236188342666084e-05,
+      "loss": 1.8579,
+      "step": 3430
+    },
+    {
+      "epoch": 191.11111111111111,
+      "grad_norm": 2.941999033737802,
+      "learning_rate": 5.109200955860392e-05,
+      "loss": 1.8542,
+      "step": 3440
+    },
+    {
+      "epoch": 191.66666666666666,
+      "grad_norm": 3.005161930741914,
+      "learning_rate": 5.0947696603227316e-05,
+      "loss": 1.847,
+      "step": 3450
+    },
+    {
+      "epoch": 192.22222222222223,
+      "grad_norm": 2.848523503698547,
+      "learning_rate": 5.0803251730102097e-05,
+      "loss": 1.8427,
+      "step": 3460
+    },
+    {
+      "epoch": 192.77777777777777,
+      "grad_norm": 2.8921938376251637,
+      "learning_rate": 5.065867719485405e-05,
+      "loss": 1.825,
+      "step": 3470
+    },
+    {
+      "epoch": 193.33333333333334,
+      "grad_norm": 2.9309556174742744,
+      "learning_rate": 5.051397525513377e-05,
+      "loss": 1.8348,
+      "step": 3480
+    },
+    {
+      "epoch": 193.88888888888889,
+      "grad_norm": 2.83715457914387,
+      "learning_rate": 5.036914817058136e-05,
+      "loss": 1.8061,
+      "step": 3490
+    },
+    {
+      "epoch": 194.44444444444446,
+      "grad_norm": 2.9683744494164213,
+      "learning_rate": 5.022419820279119e-05,
+      "loss": 1.7958,
+      "step": 3500
+    },
+    {
+      "epoch": 195.0,
+      "grad_norm": 2.813384107866755,
+      "learning_rate": 5.0079127615276534e-05,
+      "loss": 1.8013,
+      "step": 3510
+    },
+    {
+      "epoch": 195.55555555555554,
+      "grad_norm": 2.8314280742366025,
+      "learning_rate": 4.993393867343423e-05,
+      "loss": 1.7859,
+      "step": 3520
+    },
+    {
+      "epoch": 196.11111111111111,
+      "grad_norm": 2.820045268324459,
+      "learning_rate": 4.978863364450935e-05,
+      "loss": 1.805,
+      "step": 3530
+    },
+    {
+      "epoch": 196.66666666666666,
+      "grad_norm": 2.917557734659385,
+      "learning_rate": 4.964321479755974e-05,
+      "loss": 1.8434,
+      "step": 3540
+    },
+    {
+      "epoch": 197.22222222222223,
+      "grad_norm": 2.892890825697841,
+      "learning_rate": 4.949768440342059e-05,
+      "loss": 1.8175,
+      "step": 3550
+    },
+    {
+      "epoch": 197.77777777777777,
+      "grad_norm": 2.8507267811642643,
+      "learning_rate": 4.935204473466903e-05,
+      "loss": 1.7984,
+      "step": 3560
+    },
+    {
+      "epoch": 198.33333333333334,
+      "grad_norm": 2.8731418077874697,
+      "learning_rate": 4.920629806558859e-05,
+      "loss": 1.7951,
+      "step": 3570
+    },
+    {
+      "epoch": 198.88888888888889,
+      "grad_norm": 2.8414892915050363,
+      "learning_rate": 4.906044667213368e-05,
+      "loss": 1.7751,
+      "step": 3580
+    },
+    {
+      "epoch": 199.44444444444446,
+      "grad_norm": 2.977522087111335,
+      "learning_rate": 4.891449283189408e-05,
+      "loss": 1.798,
+      "step": 3590
+    },
+    {
+      "epoch": 200.0,
+      "grad_norm": 2.8607777974395776,
+      "learning_rate": 4.876843882405937e-05,
+      "loss": 1.7546,
+      "step": 3600
+    },
+    {
+      "epoch": 200.55555555555554,
+      "grad_norm": 2.9199011352210764,
+      "learning_rate": 4.862228692938334e-05,
+      "loss": 1.7497,
+      "step": 3610
+    },
+    {
+      "epoch": 201.11111111111111,
+      "grad_norm": 2.8934146101767504,
+      "learning_rate": 4.8476039430148316e-05,
+      "loss": 1.759,
+      "step": 3620
+    },
+    {
+      "epoch": 201.66666666666666,
+      "grad_norm": 2.9283950599702733,
+      "learning_rate": 4.83296986101296e-05,
+      "loss": 1.7277,
+      "step": 3630
+    },
+    {
+      "epoch": 202.22222222222223,
+      "grad_norm": 2.9566811438732525,
+      "learning_rate": 4.818326675455978e-05,
+      "loss": 1.7519,
+      "step": 3640
+    },
+    {
+      "epoch": 202.77777777777777,
+      "grad_norm": 2.970436179502361,
+      "learning_rate": 4.803674615009305e-05,
+      "loss": 1.7352,
+      "step": 3650
+    },
+    {
+      "epoch": 203.33333333333334,
+      "grad_norm": 2.939465615889158,
+      "learning_rate": 4.789013908476944e-05,
+      "loss": 1.7205,
+      "step": 3660
+    },
+    {
+      "epoch": 203.88888888888889,
+      "grad_norm": 2.9556859129487045,
+      "learning_rate": 4.7743447847979194e-05,
+      "loss": 1.7137,
+      "step": 3670
+    },
+    {
+      "epoch": 204.44444444444446,
+      "grad_norm": 2.8444437710774473,
+      "learning_rate": 4.7596674730426936e-05,
+      "loss": 1.7214,
+      "step": 3680
+    },
+    {
+      "epoch": 205.0,
+      "grad_norm": 2.9718422295240736,
+      "learning_rate": 4.744982202409592e-05,
+      "loss": 1.7329,
+      "step": 3690
+    },
+    {
+      "epoch": 205.55555555555554,
+      "grad_norm": 2.9689981457250605,
+      "learning_rate": 4.730289202221226e-05,
+      "loss": 1.6969,
+      "step": 3700
+    },
+    {
+      "epoch": 206.11111111111111,
+      "grad_norm": 2.911539330262856,
+      "learning_rate": 4.7155887019209106e-05,
+      "loss": 1.7079,
+      "step": 3710
+    },
+    {
+      "epoch": 206.66666666666666,
+      "grad_norm": 2.958060525126641,
+      "learning_rate": 4.700880931069078e-05,
+      "loss": 1.7096,
+      "step": 3720
+    },
+    {
+      "epoch": 207.22222222222223,
+      "grad_norm": 3.0407717882031413,
+      "learning_rate": 4.6861661193397e-05,
+      "loss": 1.7193,
+      "step": 3730
+    },
+    {
+      "epoch": 207.77777777777777,
+      "grad_norm": 2.828965204525847,
+      "learning_rate": 4.671444496516696e-05,
+      "loss": 1.7116,
+      "step": 3740
+    },
+    {
+      "epoch": 208.33333333333334,
+      "grad_norm": 3.0146503355800633,
+      "learning_rate": 4.6567162924903464e-05,
+      "loss": 1.6964,
+      "step": 3750
+    },
+    {
+      "epoch": 208.88888888888889,
+      "grad_norm": 2.856498730803932,
+      "learning_rate": 4.6419817372537015e-05,
+      "loss": 1.6755,
+      "step": 3760
+    },
+    {
+      "epoch": 209.44444444444446,
+      "grad_norm": 2.9776138491448383,
+      "learning_rate": 4.627241060898991e-05,
+      "loss": 1.7132,
+      "step": 3770
+    },
+    {
+      "epoch": 210.0,
+      "grad_norm": 2.899723230506249,
+      "learning_rate": 4.612494493614034e-05,
+      "loss": 1.6674,
+      "step": 3780
+    },
+    {
+      "epoch": 210.55555555555554,
+      "grad_norm": 2.9381395009098257,
+      "learning_rate": 4.597742265678638e-05,
+      "loss": 1.6759,
+      "step": 3790
+    },
+    {
+      "epoch": 211.11111111111111,
+      "grad_norm": 3.146875794002429,
+      "learning_rate": 4.582984607461006e-05,
+      "loss": 1.6659,
+      "step": 3800
+    },
+    {
+      "epoch": 211.66666666666666,
+      "grad_norm": 3.0595930666645104,
+      "learning_rate": 4.568221749414142e-05,
+      "loss": 1.6468,
+      "step": 3810
+    },
+    {
+      "epoch": 212.22222222222223,
+      "grad_norm": 2.9528440311527553,
+      "learning_rate": 4.553453922072248e-05,
+      "loss": 1.6578,
+      "step": 3820
+    },
+    {
+      "epoch": 212.77777777777777,
+      "grad_norm": 2.893829548939085,
+      "learning_rate": 4.538681356047126e-05,
+      "loss": 1.6602,
+      "step": 3830
+    },
+    {
+      "epoch": 213.33333333333334,
+      "grad_norm": 3.0478684590808482,
+      "learning_rate": 4.523904282024575e-05,
+      "loss": 1.6803,
+      "step": 3840
+    },
+    {
+      "epoch": 213.88888888888889,
+      "grad_norm": 2.84224921228185,
+      "learning_rate": 4.509122930760793e-05,
+      "loss": 1.6053,
+      "step": 3850
+    },
+    {
+      "epoch": 214.44444444444446,
+      "grad_norm": 2.927574270399235,
+      "learning_rate": 4.494337533078768e-05,
+      "loss": 1.6171,
+      "step": 3860
+    },
+    {
+      "epoch": 215.0,
+      "grad_norm": 2.9336062658375655,
+      "learning_rate": 4.479548319864678e-05,
+      "loss": 1.6529,
+      "step": 3870
+    },
+    {
+      "epoch": 215.55555555555554,
+      "grad_norm": 3.0013485102948345,
+      "learning_rate": 4.464755522064282e-05,
+      "loss": 1.6476,
+      "step": 3880
+    },
+    {
+      "epoch": 216.11111111111111,
+      "grad_norm": 2.947592260261155,
+      "learning_rate": 4.449959370679317e-05,
+      "loss": 1.6447,
+      "step": 3890
+    },
+    {
+      "epoch": 216.66666666666666,
+      "grad_norm": 3.0745695851867256,
+      "learning_rate": 4.435160096763887e-05,
+      "loss": 1.5955,
+      "step": 3900
+    },
+    {
+      "epoch": 217.22222222222223,
+      "grad_norm": 3.02305044609477,
+      "learning_rate": 4.4203579314208594e-05,
+      "loss": 1.6066,
+      "step": 3910
+    },
+    {
+      "epoch": 217.77777777777777,
+      "grad_norm": 3.0464866977628504,
+      "learning_rate": 4.40555310579825e-05,
+      "loss": 1.6255,
+      "step": 3920
+    },
+    {
+      "epoch": 218.33333333333334,
+      "grad_norm": 3.0134907023373625,
+      "learning_rate": 4.390745851085622e-05,
+      "loss": 1.5877,
+      "step": 3930
+    },
+    {
+      "epoch": 218.88888888888889,
+      "grad_norm": 3.077957381123827,
+      "learning_rate": 4.375936398510467e-05,
+      "loss": 1.599,
+      "step": 3940
+    },
+    {
+      "epoch": 219.44444444444446,
+      "grad_norm": 2.975229199980975,
+      "learning_rate": 4.3611249793346e-05,
+      "loss": 1.5876,
+      "step": 3950
+    },
+    {
+      "epoch": 220.0,
+      "grad_norm": 2.961577736273063,
+      "learning_rate": 4.3463118248505436e-05,
+      "loss": 1.5897,
+      "step": 3960
+    },
+    {
+      "epoch": 220.55555555555554,
+      "grad_norm": 3.2573780529415304,
+      "learning_rate": 4.3314971663779214e-05,
+      "loss": 1.5872,
+      "step": 3970
+    },
+    {
+      "epoch": 221.11111111111111,
+      "grad_norm": 3.0729242529479097,
+      "learning_rate": 4.316681235259841e-05,
+      "loss": 1.5676,
+      "step": 3980
+    },
+    {
+      "epoch": 221.66666666666666,
+      "grad_norm": 3.0174302148216054,
+      "learning_rate": 4.301864262859285e-05,
+      "loss": 1.5765,
+      "step": 3990
+    },
+    {
+      "epoch": 222.22222222222223,
+      "grad_norm": 3.17640385266762,
+      "learning_rate": 4.287046480555496e-05,
+      "loss": 1.5958,
+      "step": 4000
+    },
+    {
+      "epoch": 222.77777777777777,
+      "grad_norm": 2.91931758483602,
+      "learning_rate": 4.272228119740365e-05,
+      "loss": 1.5522,
+      "step": 4010
+    },
+    {
+      "epoch": 223.33333333333334,
+      "grad_norm": 2.9523088245693203,
+      "learning_rate": 4.257409411814814e-05,
+      "loss": 1.5643,
+      "step": 4020
+    },
+    {
+      "epoch": 223.88888888888889,
+      "grad_norm": 3.0376089194278606,
+      "learning_rate": 4.242590588185188e-05,
+      "loss": 1.5518,
+      "step": 4030
+    },
+    {
+      "epoch": 224.44444444444446,
+      "grad_norm": 2.8958667963596496,
+      "learning_rate": 4.227771880259636e-05,
+      "loss": 1.5362,
+      "step": 4040
+    },
+    {
+      "epoch": 225.0,
+      "grad_norm": 2.9902587730164845,
+      "learning_rate": 4.2129535194445045e-05,
+      "loss": 1.5684,
+      "step": 4050
+    },
+    {
+      "epoch": 225.55555555555554,
+      "grad_norm": 3.069533853585607,
+      "learning_rate": 4.1981357371407156e-05,
+      "loss": 1.5546,
+      "step": 4060
+    },
+    {
+      "epoch": 226.11111111111111,
+      "grad_norm": 2.848179982273905,
+      "learning_rate": 4.18331876474016e-05,
+      "loss": 1.5433,
+      "step": 4070
+    },
+    {
+      "epoch": 226.66666666666666,
+      "grad_norm": 3.066725793213224,
+      "learning_rate": 4.1685028336220806e-05,
+      "loss": 1.5312,
+      "step": 4080
+    },
+    {
+      "epoch": 227.22222222222223,
+      "grad_norm": 3.034242706086393,
+      "learning_rate": 4.1536881751494577e-05,
+      "loss": 1.5191,
+      "step": 4090
+    },
+    {
+      "epoch": 227.77777777777777,
+      "grad_norm": 2.926737447011644,
+      "learning_rate": 4.138875020665401e-05,
+      "loss": 1.5351,
+      "step": 4100
+    },
+    {
+      "epoch": 228.33333333333334,
+      "grad_norm": 3.144361560105088,
+      "learning_rate": 4.1240636014895337e-05,
+      "loss": 1.5205,
+      "step": 4110
+    },
+    {
+      "epoch": 228.88888888888889,
+      "grad_norm": 2.968946751549714,
+      "learning_rate": 4.109254148914379e-05,
+      "loss": 1.4961,
+      "step": 4120
+    },
+    {
+      "epoch": 229.44444444444446,
+      "grad_norm": 3.0203478894476734,
+      "learning_rate": 4.094446894201751e-05,
+      "loss": 1.4969,
+      "step": 4130
+    },
+    {
+      "epoch": 230.0,
+      "grad_norm": 3.0681797220242206,
+      "learning_rate": 4.0796420685791426e-05,
+      "loss": 1.4893,
+      "step": 4140
+    },
+    {
+      "epoch": 230.55555555555554,
+      "grad_norm": 3.031883252253927,
+      "learning_rate": 4.064839903236115e-05,
+      "loss": 1.4907,
+      "step": 4150
+    },
+    {
+      "epoch": 231.11111111111111,
+      "grad_norm": 2.9412164107584045,
+      "learning_rate": 4.050040629320684e-05,
+      "loss": 1.4863,
+      "step": 4160
+    },
+    {
+      "epoch": 231.66666666666666,
+      "grad_norm": 2.9993611291277285,
+      "learning_rate": 4.035244477935719e-05,
+      "loss": 1.4877,
+      "step": 4170
+    },
+    {
+      "epoch": 232.22222222222223,
+      "grad_norm": 3.0750895603442485,
+      "learning_rate": 4.020451680135323e-05,
+      "loss": 1.4705,
+      "step": 4180
+    },
+    {
+      "epoch": 232.77777777777777,
+      "grad_norm": 3.040543928173114,
+      "learning_rate": 4.005662466921234e-05,
+      "loss": 1.5049,
+      "step": 4190
+    },
+    {
+      "epoch": 233.33333333333334,
+      "grad_norm": 3.020268635201872,
+      "learning_rate": 3.9908770692392085e-05,
+      "loss": 1.4929,
+      "step": 4200
+    },
+    {
+      "epoch": 233.88888888888889,
+      "grad_norm": 3.1822270180253462,
+      "learning_rate": 3.976095717975426e-05,
+      "loss": 1.4503,
+      "step": 4210
+    },
+    {
+      "epoch": 234.44444444444446,
+      "grad_norm": 2.997125201873957,
+      "learning_rate": 3.9613186439528754e-05,
+      "loss": 1.454,
+      "step": 4220
+    },
+    {
+      "epoch": 235.0,
+      "grad_norm": 3.0438572733498694,
+      "learning_rate": 3.946546077927753e-05,
+      "loss": 1.4718,
+      "step": 4230
+    },
+    {
+      "epoch": 235.55555555555554,
+      "grad_norm": 3.0258582229732225,
+      "learning_rate": 3.931778250585859e-05,
+      "loss": 1.4276,
+      "step": 4240
+    },
+    {
+      "epoch": 236.11111111111111,
+      "grad_norm": 3.0835268286398767,
+      "learning_rate": 3.917015392538995e-05,
+      "loss": 1.45,
+      "step": 4250
+    },
+    {
+      "epoch": 236.66666666666666,
+      "grad_norm": 2.968676114417747,
+      "learning_rate": 3.9022577343213635e-05,
+      "loss": 1.4285,
+      "step": 4260
+    },
+    {
+      "epoch": 237.22222222222223,
+      "grad_norm": 2.9114127294873358,
+      "learning_rate": 3.887505506385967e-05,
+      "loss": 1.4668,
+      "step": 4270
+    },
+    {
+      "epoch": 237.77777777777777,
+      "grad_norm": 2.9470962263382856,
+      "learning_rate": 3.87275893910101e-05,
+      "loss": 1.4415,
+      "step": 4280
+    },
+    {
+      "epoch": 238.33333333333334,
+      "grad_norm": 3.1263128955955435,
+      "learning_rate": 3.8580182627463004e-05,
+      "loss": 1.4624,
+      "step": 4290
+    },
+    {
+      "epoch": 238.88888888888889,
+      "grad_norm": 3.1185409934545425,
+      "learning_rate": 3.843283707509655e-05,
+      "loss": 1.4281,
+      "step": 4300
+    },
+    {
+      "epoch": 239.44444444444446,
+      "grad_norm": 2.9140042460246365,
+      "learning_rate": 3.8285555034833046e-05,
+      "loss": 1.4191,
+      "step": 4310
+    },
+    {
+      "epoch": 240.0,
+      "grad_norm": 2.853922998278429,
+      "learning_rate": 3.8138338806603005e-05,
+      "loss": 1.4359,
+      "step": 4320
+    },
+    {
+      "epoch": 240.55555555555554,
+      "grad_norm": 2.9723591599653765,
+      "learning_rate": 3.799119068930923e-05,
+      "loss": 1.3928,
+      "step": 4330
+    },
+    {
+      "epoch": 241.11111111111111,
+      "grad_norm": 3.0770582490353577,
+      "learning_rate": 3.784411298079091e-05,
+      "loss": 1.4073,
+      "step": 4340
+    },
+    {
+      "epoch": 241.66666666666666,
+      "grad_norm": 2.8464337774360065,
+      "learning_rate": 3.769710797778775e-05,
+      "loss": 1.4489,
+      "step": 4350
+    },
+    {
+      "epoch": 242.22222222222223,
+      "grad_norm": 2.9242033631645845,
+      "learning_rate": 3.7550177975904086e-05,
+      "loss": 1.3971,
+      "step": 4360
+    },
+    {
+      "epoch": 242.77777777777777,
+      "grad_norm": 2.887207320271527,
+      "learning_rate": 3.740332526957308e-05,
+      "loss": 1.4153,
+      "step": 4370
+    },
+    {
+      "epoch": 243.33333333333334,
+      "grad_norm": 3.082116032777149,
+      "learning_rate": 3.725655215202081e-05,
+      "loss": 1.3609,
+      "step": 4380
+    },
+    {
+      "epoch": 243.88888888888889,
+      "grad_norm": 3.1066998039028206,
+      "learning_rate": 3.7109860915230564e-05,
+      "loss": 1.4014,
+      "step": 4390
+    },
+    {
+      "epoch": 244.44444444444446,
+      "grad_norm": 3.0599052180153072,
+      "learning_rate": 3.696325384990696e-05,
+      "loss": 1.3499,
+      "step": 4400
+    },
+    {
+      "epoch": 245.0,
+      "grad_norm": 2.9585070133716003,
+      "learning_rate": 3.681673324544022e-05,
+      "loss": 1.4098,
+      "step": 4410
+    },
+    {
+      "epoch": 245.55555555555554,
+      "grad_norm": 2.9766176073474897,
+      "learning_rate": 3.6670301389870404e-05,
+      "loss": 1.3436,
+      "step": 4420
+    },
+    {
+      "epoch": 246.11111111111111,
+      "grad_norm": 3.0080311088881,
+      "learning_rate": 3.65239605698517e-05,
+      "loss": 1.3685,
+      "step": 4430
+    },
+    {
+      "epoch": 246.66666666666666,
+      "grad_norm": 3.1430219167393667,
+      "learning_rate": 3.637771307061668e-05,
+      "loss": 1.3722,
+      "step": 4440
+    },
+    {
+      "epoch": 247.22222222222223,
+      "grad_norm": 2.816808663876402,
+      "learning_rate": 3.6231561175940636e-05,
+      "loss": 1.3796,
+      "step": 4450
+    },
+    {
+      "epoch": 247.77777777777777,
+      "grad_norm": 3.1438964202941047,
+      "learning_rate": 3.608550716810593e-05,
+      "loss": 1.3392,
+      "step": 4460
+    },
+    {
+      "epoch": 248.33333333333334,
+      "grad_norm": 2.9673236833751293,
+      "learning_rate": 3.593955332786634e-05,
+      "loss": 1.3413,
+      "step": 4470
+    },
+    {
+      "epoch": 248.88888888888889,
+      "grad_norm": 2.899391202492822,
+      "learning_rate": 3.579370193441142e-05,
+      "loss": 1.3314,
+      "step": 4480
+    },
+    {
+      "epoch": 249.44444444444446,
+      "grad_norm": 3.139663625060467,
+      "learning_rate": 3.5647955265330975e-05,
+      "loss": 1.3171,
+      "step": 4490
+    },
+    {
+      "epoch": 250.0,
+      "grad_norm": 3.1374599013482802,
+      "learning_rate": 3.5502315596579415e-05,
+      "loss": 1.3497,
+      "step": 4500
+    },
+    {
+      "epoch": 250.55555555555554,
+      "grad_norm": 2.868179025114189,
+      "learning_rate": 3.535678520244027e-05,
+      "loss": 1.2739,
+      "step": 4510
+    },
+    {
+      "epoch": 251.11111111111111,
+      "grad_norm": 3.0480330393945407,
+      "learning_rate": 3.521136635549066e-05,
+      "loss": 1.3107,
+      "step": 4520
+    },
+    {
+      "epoch": 251.66666666666666,
+      "grad_norm": 2.9642362311430235,
+      "learning_rate": 3.506606132656578e-05,
+      "loss": 1.3483,
+      "step": 4530
+    },
+    {
+      "epoch": 252.22222222222223,
+      "grad_norm": 2.9700101085452286,
+      "learning_rate": 3.492087238472348e-05,
+      "loss": 1.3292,
+      "step": 4540
+    },
+    {
+      "epoch": 252.77777777777777,
+      "grad_norm": 2.9292226193665436,
+      "learning_rate": 3.477580179720882e-05,
+      "loss": 1.2735,
+      "step": 4550
+    },
+    {
+      "epoch": 253.33333333333334,
+      "grad_norm": 2.8727989063835557,
+      "learning_rate": 3.4630851829418646e-05,
+      "loss": 1.3242,
+      "step": 4560
+    },
+    {
+      "epoch": 253.88888888888889,
+      "grad_norm": 3.0928301744282702,
+      "learning_rate": 3.4486024744866246e-05,
+      "loss": 1.2886,
+      "step": 4570
+    },
+    {
+      "epoch": 254.44444444444446,
+      "grad_norm": 3.090940644873239,
+      "learning_rate": 3.4341322805145965e-05,
+      "loss": 1.2714,
+      "step": 4580
+    },
+    {
+      "epoch": 255.0,
+      "grad_norm": 3.0086868402486733,
+      "learning_rate": 3.4196748269897916e-05,
+      "loss": 1.2883,
+      "step": 4590
+    },
+    {
+      "epoch": 255.55555555555554,
+      "grad_norm": 2.8933971412349773,
+      "learning_rate": 3.405230339677269e-05,
+      "loss": 1.2899,
+      "step": 4600
+    },
+    {
+      "epoch": 256.1111111111111,
+      "grad_norm": 2.874713385849318,
+      "learning_rate": 3.390799044139609e-05,
+      "loss": 1.258,
+      "step": 4610
+    },
+    {
+      "epoch": 256.6666666666667,
+      "grad_norm": 3.09839258666389,
+      "learning_rate": 3.376381165733392e-05,
+      "loss": 1.2793,
+      "step": 4620
+    },
+    {
+      "epoch": 257.22222222222223,
+      "grad_norm": 2.9550535978823502,
+      "learning_rate": 3.3619769296056794e-05,
+      "loss": 1.2509,
+      "step": 4630
+    },
+    {
+      "epoch": 257.77777777777777,
+      "grad_norm": 2.97393025631631,
+      "learning_rate": 3.347586560690495e-05,
+      "loss": 1.2741,
+      "step": 4640
+    },
+    {
+      "epoch": 258.3333333333333,
+      "grad_norm": 3.020839787618823,
+      "learning_rate": 3.3332102837053155e-05,
+      "loss": 1.2746,
+      "step": 4650
+    },
+    {
+      "epoch": 258.8888888888889,
+      "grad_norm": 3.002393562405945,
+      "learning_rate": 3.318848323147563e-05,
+      "loss": 1.2725,
+      "step": 4660
+    },
+    {
+      "epoch": 259.44444444444446,
+      "grad_norm": 2.8833136394320618,
+      "learning_rate": 3.3045009032910934e-05,
+      "loss": 1.2414,
+      "step": 4670
+    },
+    {
+      "epoch": 260.0,
+      "grad_norm": 2.802068225767872,
+      "learning_rate": 3.2901682481827016e-05,
+      "loss": 1.2529,
+      "step": 4680
+    },
+    {
+      "epoch": 260.55555555555554,
+      "grad_norm": 2.950696643104212,
+      "learning_rate": 3.2758505816386125e-05,
+      "loss": 1.256,
+      "step": 4690
+    },
+    {
+      "epoch": 261.1111111111111,
+      "grad_norm": 3.0296431539672883,
+      "learning_rate": 3.261548127240998e-05,
+      "loss": 1.2287,
+      "step": 4700
+    },
+    {
+      "epoch": 261.6666666666667,
+      "grad_norm": 2.943231712767288,
+      "learning_rate": 3.247261108334476e-05,
+      "loss": 1.2594,
+      "step": 4710
+    },
+    {
+      "epoch": 262.22222222222223,
+      "grad_norm": 2.9333730073615567,
+      "learning_rate": 3.23298974802263e-05,
+      "loss": 1.2277,
+      "step": 4720
+    },
+    {
+      "epoch": 262.77777777777777,
+      "grad_norm": 3.160015013453877,
+      "learning_rate": 3.218734269164519e-05,
+      "loss": 1.2597,
+      "step": 4730
+    },
+    {
+      "epoch": 263.3333333333333,
+      "grad_norm": 2.9754263243825436,
+      "learning_rate": 3.204494894371201e-05,
+      "loss": 1.2405,
+      "step": 4740
+    },
+    {
+      "epoch": 263.8888888888889,
+      "grad_norm": 2.984720260061695,
+      "learning_rate": 3.190271846002255e-05,
+      "loss": 1.2073,
+      "step": 4750
+    },
+    {
+      "epoch": 264.44444444444446,
+      "grad_norm": 2.9158880738597195,
+      "learning_rate": 3.1760653461623104e-05,
+      "loss": 1.2112,
+      "step": 4760
+    },
+    {
+      "epoch": 265.0,
+      "grad_norm": 2.9485902230190772,
+      "learning_rate": 3.161875616697578e-05,
+      "loss": 1.2284,
+      "step": 4770
+    },
+    {
+      "epoch": 265.55555555555554,
+      "grad_norm": 3.007425972018738,
+      "learning_rate": 3.147702879192382e-05,
+      "loss": 1.2056,
+      "step": 4780
+    },
+    {
+      "epoch": 266.1111111111111,
+      "grad_norm": 2.988670732634813,
+      "learning_rate": 3.1335473549657086e-05,
+      "loss": 1.2204,
+      "step": 4790
+    },
+    {
+      "epoch": 266.6666666666667,
+      "grad_norm": 2.7767776362976693,
+      "learning_rate": 3.11940926506774e-05,
+      "loss": 1.1956,
+      "step": 4800
+    },
+    {
+      "epoch": 267.22222222222223,
+      "grad_norm": 3.066881898569449,
+      "learning_rate": 3.105288830276408e-05,
+      "loss": 1.2192,
+      "step": 4810
+    },
+    {
+      "epoch": 267.77777777777777,
+      "grad_norm": 3.1256617036253704,
+      "learning_rate": 3.091186271093947e-05,
+      "loss": 1.1917,
+      "step": 4820
+    },
+    {
+      "epoch": 268.3333333333333,
+      "grad_norm": 3.0128127509525457,
+      "learning_rate": 3.077101807743449e-05,
+      "loss": 1.2127,
+      "step": 4830
+    },
+    {
+      "epoch": 268.8888888888889,
+      "grad_norm": 3.0811792702130187,
+      "learning_rate": 3.063035660165422e-05,
+      "loss": 1.2015,
+      "step": 4840
+    },
+    {
+      "epoch": 269.44444444444446,
+      "grad_norm": 2.9128259893865613,
+      "learning_rate": 3.04898804801436e-05,
+      "loss": 1.1665,
+      "step": 4850
+    },
+    {
+      "epoch": 270.0,
+      "grad_norm": 2.934277329124476,
+      "learning_rate": 3.0349591906553102e-05,
+      "loss": 1.1374,
+      "step": 4860
+    },
+    {
+      "epoch": 270.55555555555554,
+      "grad_norm": 2.9547255281112266,
+      "learning_rate": 3.0209493071604496e-05,
+      "loss": 1.173,
+      "step": 4870
+    },
+    {
+      "epoch": 271.1111111111111,
+      "grad_norm": 2.925854589054956,
+      "learning_rate": 3.006958616305662e-05,
+      "loss": 1.1657,
+      "step": 4880
+    },
+    {
+      "epoch": 271.6666666666667,
+      "grad_norm": 2.978845236628889,
+      "learning_rate": 2.9929873365671204e-05,
+      "loss": 1.1459,
+      "step": 4890
+    },
+    {
+      "epoch": 272.22222222222223,
+      "grad_norm": 2.9139520455461643,
+      "learning_rate": 2.979035686117882e-05,
+      "loss": 1.1622,
+      "step": 4900
+    },
+    {
+      "epoch": 272.77777777777777,
+      "grad_norm": 3.226624273489324,
+      "learning_rate": 2.96510388282447e-05,
+      "loss": 1.1415,
+      "step": 4910
+    },
+    {
+      "epoch": 273.3333333333333,
+      "grad_norm": 3.0375944774361483,
+      "learning_rate": 2.951192144243485e-05,
+      "loss": 1.1358,
+      "step": 4920
+    },
+    {
+      "epoch": 273.8888888888889,
+      "grad_norm": 2.9349787829280456,
+      "learning_rate": 2.937300687618194e-05,
+      "loss": 1.1391,
+      "step": 4930
+    },
+    {
+      "epoch": 274.44444444444446,
+      "grad_norm": 3.038083583201836,
+      "learning_rate": 2.9234297298751482e-05,
+      "loss": 1.1168,
+      "step": 4940
+    },
+    {
+      "epoch": 275.0,
+      "grad_norm": 2.97456080296235,
+      "learning_rate": 2.909579487620792e-05,
+      "loss": 1.1384,
+      "step": 4950
+    },
+    {
+      "epoch": 275.55555555555554,
+      "grad_norm": 2.821097985779195,
+      "learning_rate": 2.8957501771380773e-05,
+      "loss": 1.1444,
+      "step": 4960
+    },
+    {
+      "epoch": 276.1111111111111,
+      "grad_norm": 2.9853339926290303,
+      "learning_rate": 2.8819420143830936e-05,
+      "loss": 1.1409,
+      "step": 4970
+    },
+    {
+      "epoch": 276.6666666666667,
+      "grad_norm": 2.939766354521605,
+      "learning_rate": 2.8681552149816874e-05,
+      "loss": 1.1318,
+      "step": 4980
+    },
+    {
+      "epoch": 277.22222222222223,
+      "grad_norm": 2.826600649245691,
+      "learning_rate": 2.854389994226101e-05,
+      "loss": 1.1417,
+      "step": 4990
+    },
+    {
+      "epoch": 277.77777777777777,
+      "grad_norm": 2.9047293017223104,
+      "learning_rate": 2.840646567071608e-05,
+      "loss": 1.1431,
+      "step": 5000
+    },
+    {
+      "epoch": 278.3333333333333,
+      "grad_norm": 2.819797204618978,
+      "learning_rate": 2.8269251481331554e-05,
+      "loss": 1.1059,
+      "step": 5010
+    },
+    {
+      "epoch": 278.8888888888889,
+      "grad_norm": 3.012549220272663,
+      "learning_rate": 2.8132259516820147e-05,
+      "loss": 1.1449,
+      "step": 5020
+    },
+    {
+      "epoch": 279.44444444444446,
+      "grad_norm": 3.014129900537137,
+      "learning_rate": 2.7995491916424372e-05,
+      "loss": 1.1256,
+      "step": 5030
+    },
+    {
+      "epoch": 280.0,
+      "grad_norm": 2.975675035074844,
+      "learning_rate": 2.785895081588307e-05,
+      "loss": 1.0984,
+      "step": 5040
+    },
+    {
+      "epoch": 280.55555555555554,
+      "grad_norm": 3.13220285015499,
+      "learning_rate": 2.7722638347398127e-05,
+      "loss": 1.1073,
+      "step": 5050
+    },
+    {
+      "epoch": 281.1111111111111,
+      "grad_norm": 3.0127325071144018,
+      "learning_rate": 2.7586556639601152e-05,
+      "loss": 1.1074,
+      "step": 5060
+    },
+    {
+      "epoch": 281.6666666666667,
+      "grad_norm": 3.0501408216141837,
+      "learning_rate": 2.7450707817520226e-05,
+      "loss": 1.096,
+      "step": 5070
+    },
+    {
+      "epoch": 282.22222222222223,
+      "grad_norm": 3.06167404558777,
+      "learning_rate": 2.7315094002546726e-05,
+      "loss": 1.1015,
+      "step": 5080
+    },
+    {
+      "epoch": 282.77777777777777,
+      "grad_norm": 3.037988781946602,
+      "learning_rate": 2.71797173124022e-05,
+      "loss": 1.0905,
+      "step": 5090
+    },
+    {
+      "epoch": 283.3333333333333,
+      "grad_norm": 2.9782806003586457,
+      "learning_rate": 2.7044579861105303e-05,
+      "loss": 1.0663,
+      "step": 5100
+    },
+    {
+      "epoch": 283.8888888888889,
+      "grad_norm": 2.965923047539464,
+      "learning_rate": 2.6909683758938784e-05,
+      "loss": 1.0716,
+      "step": 5110
+    },
+    {
+      "epoch": 284.44444444444446,
+      "grad_norm": 3.0777094991419185,
+      "learning_rate": 2.6775031112416504e-05,
+      "loss": 1.1025,
+      "step": 5120
+    },
+    {
+      "epoch": 285.0,
+      "grad_norm": 3.1221743206303074,
+      "learning_rate": 2.6640624024250583e-05,
+      "loss": 1.0839,
+      "step": 5130
+    },
+    {
+      "epoch": 285.55555555555554,
+      "grad_norm": 3.334648365659446,
+      "learning_rate": 2.6506464593318518e-05,
+      "loss": 1.1158,
+      "step": 5140
+    },
+    {
+      "epoch": 286.1111111111111,
+      "grad_norm": 2.8474017161621443,
+      "learning_rate": 2.6372554914630496e-05,
+      "loss": 1.0466,
+      "step": 5150
+    },
+    {
+      "epoch": 286.6666666666667,
+      "grad_norm": 2.796883865427639,
+      "learning_rate": 2.623889707929653e-05,
+      "loss": 1.0858,
+      "step": 5160
+    },
+    {
+      "epoch": 287.22222222222223,
+      "grad_norm": 2.8324270668976514,
+      "learning_rate": 2.610549317449397e-05,
+      "loss": 1.0645,
+      "step": 5170
+    },
+    {
+      "epoch": 287.77777777777777,
+      "grad_norm": 3.0740195843600766,
+      "learning_rate": 2.5972345283434773e-05,
+      "loss": 1.027,
+      "step": 5180
+    },
+    {
+      "epoch": 288.3333333333333,
+      "grad_norm": 3.0863236197817865,
+      "learning_rate": 2.5839455485333073e-05,
+      "loss": 1.0716,
+      "step": 5190
+    },
+    {
+      "epoch": 288.8888888888889,
+      "grad_norm": 3.015186495350761,
+      "learning_rate": 2.5706825855372648e-05,
+      "loss": 1.0683,
+      "step": 5200
+    },
+    {
+      "epoch": 289.44444444444446,
+      "grad_norm": 3.1217365772615246,
+      "learning_rate": 2.557445846467452e-05,
+      "loss": 1.0587,
+      "step": 5210
+    },
+    {
+      "epoch": 290.0,
+      "grad_norm": 2.9772203574515155,
+      "learning_rate": 2.5442355380264677e-05,
+      "loss": 1.0344,
+      "step": 5220
+    },
+    {
+      "epoch": 290.55555555555554,
+      "grad_norm": 2.958927327729786,
+      "learning_rate": 2.531051866504169e-05,
+      "loss": 1.0525,
+      "step": 5230
+    },
+    {
+      "epoch": 291.1111111111111,
+      "grad_norm": 2.95255964407889,
+      "learning_rate": 2.517895037774461e-05,
+      "loss": 1.0176,
+      "step": 5240
+    },
+    {
+      "epoch": 291.6666666666667,
+      "grad_norm": 3.107826038789483,
+      "learning_rate": 2.5047652572920708e-05,
+      "loss": 1.0368,
+      "step": 5250
+    },
+    {
+      "epoch": 292.22222222222223,
+      "grad_norm": 2.8975447027725783,
+      "learning_rate": 2.4916627300893508e-05,
+      "loss": 1.0216,
+      "step": 5260
+    },
+    {
+      "epoch": 292.77777777777777,
+      "grad_norm": 2.827433380367477,
+      "learning_rate": 2.478587660773065e-05,
+      "loss": 1.0063,
+      "step": 5270
+    },
+    {
+      "epoch": 293.3333333333333,
+      "grad_norm": 2.91972459617843,
+      "learning_rate": 2.465540253521206e-05,
+      "loss": 1.0002,
+      "step": 5280
+    },
+    {
+      "epoch": 293.8888888888889,
+      "grad_norm": 2.9222881820577076,
+      "learning_rate": 2.452520712079795e-05,
+      "loss": 1.0328,
+      "step": 5290
+    },
+    {
+      "epoch": 294.44444444444446,
+      "grad_norm": 3.0137664918515545,
+      "learning_rate": 2.4395292397597077e-05,
+      "loss": 1.0132,
+      "step": 5300
+    },
+    {
+      "epoch": 295.0,
+      "grad_norm": 3.0917196065340797,
+      "learning_rate": 2.4265660394334987e-05,
+      "loss": 1.0252,
+      "step": 5310
+    },
+    {
+      "epoch": 295.55555555555554,
+      "grad_norm": 3.2152195800025596,
+      "learning_rate": 2.413631313532229e-05,
+      "loss": 0.9939,
+      "step": 5320
+    },
+    {
+      "epoch": 296.1111111111111,
+      "grad_norm": 2.7996887885575172,
+      "learning_rate": 2.4007252640423117e-05,
+      "loss": 0.9986,
+      "step": 5330
+    },
+    {
+      "epoch": 296.6666666666667,
+      "grad_norm": 2.6727828775541935,
+      "learning_rate": 2.387848092502351e-05,
+      "loss": 1.0083,
+      "step": 5340
+    },
+    {
+      "epoch": 297.22222222222223,
+      "grad_norm": 3.0476104631189513,
+      "learning_rate": 2.375000000000001e-05,
+      "loss": 0.9898,
+      "step": 5350
+    },
+    {
+      "epoch": 297.77777777777777,
+      "grad_norm": 2.8794718255401612,
+      "learning_rate": 2.3621811871688184e-05,
+      "loss": 0.9864,
+      "step": 5360
+    },
+    {
+      "epoch": 298.3333333333333,
+      "grad_norm": 2.937549509483458,
+      "learning_rate": 2.3493918541851386e-05,
+      "loss": 0.9833,
+      "step": 5370
+    },
+    {
+      "epoch": 298.8888888888889,
+      "grad_norm": 3.0748966540826332,
+      "learning_rate": 2.336632200764942e-05,
+      "loss": 0.9644,
+      "step": 5380
+    },
+    {
+      "epoch": 299.44444444444446,
+      "grad_norm": 2.9508957295004983,
+      "learning_rate": 2.323902426160737e-05,
+      "loss": 0.9767,
+      "step": 5390
+    },
+    {
+      "epoch": 300.0,
+      "grad_norm": 3.0396742021252736,
+      "learning_rate": 2.311202729158454e-05,
+      "loss": 1.0101,
+      "step": 5400
+    },
+    {
+      "epoch": 300.55555555555554,
+      "grad_norm": 2.986949189935583,
+      "learning_rate": 2.298533308074331e-05,
+      "loss": 0.9841,
+      "step": 5410
+    },
+    {
+      "epoch": 301.1111111111111,
+      "grad_norm": 2.885470521793822,
+      "learning_rate": 2.2858943607518288e-05,
+      "loss": 0.9463,
+      "step": 5420
+    },
+    {
+      "epoch": 301.6666666666667,
+      "grad_norm": 2.991928208871907,
+      "learning_rate": 2.2732860845585298e-05,
+      "loss": 0.9889,
+      "step": 5430
+    },
+    {
+      "epoch": 302.22222222222223,
+      "grad_norm": 2.969817280912554,
+      "learning_rate": 2.2607086763830653e-05,
+      "loss": 0.9882,
+      "step": 5440
+    },
+    {
+      "epoch": 302.77777777777777,
+      "grad_norm": 2.9833808561265918,
+      "learning_rate": 2.2481623326320366e-05,
+      "loss": 0.9552,
+      "step": 5450
+    },
+    {
+      "epoch": 303.3333333333333,
+      "grad_norm": 2.822937227481602,
+      "learning_rate": 2.235647249226944e-05,
+      "loss": 0.9347,
+      "step": 5460
+    },
+    {
+      "epoch": 303.8888888888889,
+      "grad_norm": 2.9381411238314956,
+      "learning_rate": 2.223163621601138e-05,
+      "loss": 0.9423,
+      "step": 5470
+    },
+    {
+      "epoch": 304.44444444444446,
+      "grad_norm": 2.8432215943765278,
+      "learning_rate": 2.210711644696756e-05,
+      "loss": 0.9429,
+      "step": 5480
+    },
+    {
+      "epoch": 305.0,
+      "grad_norm": 2.981934673133606,
+      "learning_rate": 2.1982915129616885e-05,
+      "loss": 0.9327,
+      "step": 5490
+    },
+    {
+      "epoch": 305.55555555555554,
+      "grad_norm": 2.8347461394362683,
+      "learning_rate": 2.185903420346532e-05,
+      "loss": 0.9364,
+      "step": 5500
+    },
+    {
+      "epoch": 306.1111111111111,
+      "grad_norm": 2.7963268232983345,
+      "learning_rate": 2.1735475603015695e-05,
+      "loss": 0.903,
+      "step": 5510
+    },
+    {
+      "epoch": 306.6666666666667,
+      "grad_norm": 2.9526206902910612,
+      "learning_rate": 2.161224125773743e-05,
+      "loss": 0.9255,
+      "step": 5520
+    },
+    {
+      "epoch": 307.22222222222223,
+      "grad_norm": 2.8677331057105677,
+      "learning_rate": 2.1489333092036477e-05,
+      "loss": 0.9343,
+      "step": 5530
+    },
+    {
+      "epoch": 307.77777777777777,
+      "grad_norm": 2.9562309925847394,
+      "learning_rate": 2.1366753025225178e-05,
+      "loss": 0.9457,
+      "step": 5540
+    },
+    {
+      "epoch": 308.3333333333333,
+      "grad_norm": 3.0299978914662695,
+      "learning_rate": 2.1244502971492338e-05,
+      "loss": 0.9555,
+      "step": 5550
+    },
+    {
+      "epoch": 308.8888888888889,
+      "grad_norm": 2.9306606456682016,
+      "learning_rate": 2.1122584839873397e-05,
+      "loss": 0.9139,
+      "step": 5560
+    },
+    {
+      "epoch": 309.44444444444446,
+      "grad_norm": 2.9597039786085384,
+      "learning_rate": 2.1001000534220486e-05,
+      "loss": 0.9077,
+      "step": 5570
+    },
+    {
+      "epoch": 310.0,
+      "grad_norm": 2.9133993831220897,
+      "learning_rate": 2.087975195317282e-05,
+      "loss": 0.8925,
+      "step": 5580
+    },
+    {
+      "epoch": 310.55555555555554,
+      "grad_norm": 2.8822479108699786,
+      "learning_rate": 2.0758840990126954e-05,
+      "loss": 0.9062,
+      "step": 5590
+    },
+    {
+      "epoch": 311.1111111111111,
+      "grad_norm": 2.847237764228229,
+      "learning_rate": 2.063826953320731e-05,
+      "loss": 0.9222,
+      "step": 5600
+    },
+    {
+      "epoch": 311.6666666666667,
+      "grad_norm": 2.94652825719469,
+      "learning_rate": 2.0518039465236593e-05,
+      "loss": 0.9279,
+      "step": 5610
+    },
+    {
+      "epoch": 312.22222222222223,
+      "grad_norm": 2.889891495301209,
+      "learning_rate": 2.0398152663706455e-05,
+      "loss": 0.9088,
+      "step": 5620
+    },
+    {
+      "epoch": 312.77777777777777,
+      "grad_norm": 3.014916529472665,
+      "learning_rate": 2.0278611000748184e-05,
+      "loss": 0.9158,
+      "step": 5630
+    },
+    {
+      "epoch": 313.3333333333333,
+      "grad_norm": 2.864561952453352,
+      "learning_rate": 2.0159416343103403e-05,
+      "loss": 0.9324,
+      "step": 5640
+    },
+    {
+      "epoch": 313.8888888888889,
+      "grad_norm": 2.7472918787569225,
+      "learning_rate": 2.0040570552094998e-05,
+      "loss": 0.8872,
+      "step": 5650
+    },
+    {
+      "epoch": 314.44444444444446,
+      "grad_norm": 2.866237223048124,
+      "learning_rate": 1.992207548359799e-05,
+      "loss": 0.8644,
+      "step": 5660
+    },
+    {
+      "epoch": 315.0,
+      "grad_norm": 2.933569205799871,
+      "learning_rate": 1.9803932988010596e-05,
+      "loss": 0.8975,
+      "step": 5670
+    },
+    {
+      "epoch": 315.55555555555554,
+      "grad_norm": 2.963074086895836,
+      "learning_rate": 1.968614491022529e-05,
+      "loss": 0.8809,
+      "step": 5680
+    },
+    {
+      "epoch": 316.1111111111111,
+      "grad_norm": 2.794427408107888,
+      "learning_rate": 1.956871308960006e-05,
+      "loss": 0.8573,
+      "step": 5690
+    },
+    {
+      "epoch": 316.6666666666667,
+      "grad_norm": 2.9765046682271517,
+      "learning_rate": 1.945163935992961e-05,
+      "loss": 0.8782,
+      "step": 5700
+    },
+    {
+      "epoch": 317.22222222222223,
+      "grad_norm": 2.99540597116115,
+      "learning_rate": 1.933492554941676e-05,
+      "loss": 0.886,
+      "step": 5710
+    },
+    {
+      "epoch": 317.77777777777777,
+      "grad_norm": 2.9659005393845366,
+      "learning_rate": 1.9218573480643932e-05,
+      "loss": 0.8977,
+      "step": 5720
+    },
+    {
+      "epoch": 318.3333333333333,
+      "grad_norm": 2.9835577839503795,
+      "learning_rate": 1.9102584970544597e-05,
+      "loss": 0.8773,
+      "step": 5730
+    },
+    {
+      "epoch": 318.8888888888889,
+      "grad_norm": 2.7903593238048794,
+      "learning_rate": 1.8986961830375014e-05,
+      "loss": 0.8539,
+      "step": 5740
+    },
+    {
+      "epoch": 319.44444444444446,
+      "grad_norm": 2.9625515124514243,
+      "learning_rate": 1.8871705865685837e-05,
+      "loss": 0.8902,
+      "step": 5750
+    },
+    {
+      "epoch": 320.0,
+      "grad_norm": 2.7446173094372286,
+      "learning_rate": 1.875681887629402e-05,
+      "loss": 0.8352,
+      "step": 5760
+    },
+    {
+      "epoch": 320.55555555555554,
+      "grad_norm": 2.927055621213949,
+      "learning_rate": 1.864230265625464e-05,
+      "loss": 0.8574,
+      "step": 5770
+    },
+    {
+      "epoch": 321.1111111111111,
+      "grad_norm": 2.7057398678687616,
+      "learning_rate": 1.852815899383288e-05,
+      "loss": 0.8455,
+      "step": 5780
+    },
+    {
+      "epoch": 321.6666666666667,
+      "grad_norm": 3.087924751995556,
+      "learning_rate": 1.8414389671476188e-05,
+      "loss": 0.8514,
+      "step": 5790
+    },
+    {
+      "epoch": 322.22222222222223,
+      "grad_norm": 2.7052327135995164,
+      "learning_rate": 1.8300996465786317e-05,
+      "loss": 0.8184,
+      "step": 5800
+    },
+    {
+      "epoch": 322.77777777777777,
+      "grad_norm": 2.783387466482922,
+      "learning_rate": 1.8187981147491703e-05,
+      "loss": 0.8352,
+      "step": 5810
+    },
+    {
+      "epoch": 323.3333333333333,
+      "grad_norm": 2.9300800518257897,
+      "learning_rate": 1.8075345481419723e-05,
+      "loss": 0.8398,
+      "step": 5820
+    },
+    {
+      "epoch": 323.8888888888889,
+      "grad_norm": 2.7704714106096846,
+      "learning_rate": 1.7963091226469206e-05,
+      "loss": 0.8105,
+      "step": 5830
+    },
+    {
+      "epoch": 324.44444444444446,
+      "grad_norm": 2.8557601341689556,
+      "learning_rate": 1.7851220135582877e-05,
+      "loss": 0.8544,
+      "step": 5840
+    },
+    {
+      "epoch": 325.0,
+      "grad_norm": 3.017679255991434,
+      "learning_rate": 1.773973395572011e-05,
+      "loss": 0.8251,
+      "step": 5850
+    },
+    {
+      "epoch": 325.55555555555554,
+      "grad_norm": 2.769424587073558,
+      "learning_rate": 1.7628634427829513e-05,
+      "loss": 0.839,
+      "step": 5860
+    },
+    {
+      "epoch": 326.1111111111111,
+      "grad_norm": 2.8986640286414604,
+      "learning_rate": 1.751792328682183e-05,
+      "loss": 0.8386,
+      "step": 5870
+    },
+    {
+      "epoch": 326.6666666666667,
+      "grad_norm": 2.7262643033456566,
+      "learning_rate": 1.740760226154283e-05,
+      "loss": 0.8296,
+      "step": 5880
+    },
+    {
+      "epoch": 327.22222222222223,
+      "grad_norm": 2.9172530720088408,
+      "learning_rate": 1.7297673074746285e-05,
+      "loss": 0.8267,
+      "step": 5890
+    },
+    {
+      "epoch": 327.77777777777777,
+      "grad_norm": 2.9872264720881865,
+      "learning_rate": 1.7188137443067116e-05,
+      "loss": 0.8337,
+      "step": 5900
+    },
+    {
+      "epoch": 328.3333333333333,
+      "grad_norm": 2.788496029385015,
+      "learning_rate": 1.707899707699451e-05,
+      "loss": 0.8347,
+      "step": 5910
+    },
+    {
+      "epoch": 328.8888888888889,
+      "grad_norm": 2.8927620900733655,
+      "learning_rate": 1.697025368084532e-05,
+      "loss": 0.7881,
+      "step": 5920
+    },
+    {
+      "epoch": 329.44444444444446,
+      "grad_norm": 2.82466664330731,
+      "learning_rate": 1.686190895273733e-05,
+      "loss": 0.826,
+      "step": 5930
+    },
+    {
+      "epoch": 330.0,
+      "grad_norm": 2.72042339171236,
+      "learning_rate": 1.6753964584562798e-05,
+      "loss": 0.7828,
+      "step": 5940
+    },
+    {
+      "epoch": 330.55555555555554,
+      "grad_norm": 2.8438335029163664,
+      "learning_rate": 1.6646422261962067e-05,
+      "loss": 0.8282,
+      "step": 5950
+    },
+    {
+      "epoch": 331.1111111111111,
+      "grad_norm": 2.7521529006732317,
+      "learning_rate": 1.653928366429717e-05,
+      "loss": 0.8017,
+      "step": 5960
+    },
+    {
+      "epoch": 331.6666666666667,
+      "grad_norm": 2.9813997307050335,
+      "learning_rate": 1.643255046462568e-05,
+      "loss": 0.8004,
+      "step": 5970
+    },
+    {
+      "epoch": 332.22222222222223,
+      "grad_norm": 2.8082460233282736,
+      "learning_rate": 1.6326224329674502e-05,
+      "loss": 0.7762,
+      "step": 5980
+    },
+    {
+      "epoch": 332.77777777777777,
+      "grad_norm": 2.8447838884543195,
+      "learning_rate": 1.6220306919813932e-05,
+      "loss": 0.8017,
+      "step": 5990
+    },
+    {
+      "epoch": 333.3333333333333,
+      "grad_norm": 2.812120539180332,
+      "learning_rate": 1.611479988903165e-05,
+      "loss": 0.788,
+      "step": 6000
+    },
+    {
+      "epoch": 333.8888888888889,
+      "grad_norm": 2.9347454710620875,
+      "learning_rate": 1.6009704884906963e-05,
+      "loss": 0.792,
+      "step": 6010
+    },
+    {
+      "epoch": 334.44444444444446,
+      "grad_norm": 2.821133311862373,
+      "learning_rate": 1.5905023548585002e-05,
+      "loss": 0.793,
+      "step": 6020
+    },
+    {
+      "epoch": 335.0,
+      "grad_norm": 2.877165600712079,
+      "learning_rate": 1.580075751475117e-05,
+      "loss": 0.7737,
+      "step": 6030
+    },
+    {
+      "epoch": 335.55555555555554,
+      "grad_norm": 2.689428125734847,
+      "learning_rate": 1.5696908411605575e-05,
+      "loss": 0.7828,
+      "step": 6040
+    },
+    {
+      "epoch": 336.1111111111111,
+      "grad_norm": 2.784873231094917,
+      "learning_rate": 1.5593477860837575e-05,
+      "loss": 0.7788,
+      "step": 6050
+    },
+    {
+      "epoch": 336.6666666666667,
+      "grad_norm": 2.7096678649240213,
+      "learning_rate": 1.5490467477600537e-05,
+      "loss": 0.7827,
+      "step": 6060
+    },
+    {
+      "epoch": 337.22222222222223,
+      "grad_norm": 2.8130972334095916,
+      "learning_rate": 1.5387878870486533e-05,
+      "loss": 0.771,
+      "step": 6070
+    },
+    {
+      "epoch": 337.77777777777777,
+      "grad_norm": 2.945121617291416,
+      "learning_rate": 1.5285713641501237e-05,
+      "loss": 0.7537,
+      "step": 6080
+    },
+    {
+      "epoch": 338.3333333333333,
+      "grad_norm": 2.750799583020774,
+      "learning_rate": 1.518397338603898e-05,
+      "loss": 0.7549,
+      "step": 6090
+    },
+    {
+      "epoch": 338.8888888888889,
+      "grad_norm": 2.784370386395983,
+      "learning_rate": 1.508265969285772e-05,
+      "loss": 0.7435,
+      "step": 6100
+    },
+    {
+      "epoch": 339.44444444444446,
+      "grad_norm": 2.81282745150536,
+      "learning_rate": 1.4981774144054346e-05,
+      "loss": 0.7824,
+      "step": 6110
+    },
+    {
+      "epoch": 340.0,
+      "grad_norm": 2.7052825079320995,
+      "learning_rate": 1.4881318315039875e-05,
+      "loss": 0.7607,
+      "step": 6120
+    },
+    {
+      "epoch": 340.55555555555554,
+      "grad_norm": 2.9219548015610584,
+      "learning_rate": 1.4781293774514936e-05,
+      "loss": 0.778,
+      "step": 6130
+    },
+    {
+      "epoch": 341.1111111111111,
+      "grad_norm": 2.8182673343425506,
+      "learning_rate": 1.4681702084445201e-05,
+      "loss": 0.7551,
+      "step": 6140
+    },
+    {
+      "epoch": 341.6666666666667,
+      "grad_norm": 2.92289071609676,
+      "learning_rate": 1.458254480003702e-05,
+      "loss": 0.739,
+      "step": 6150
+    },
+    {
+      "epoch": 342.22222222222223,
+      "grad_norm": 2.7214103924115856,
+      "learning_rate": 1.448382346971319e-05,
+      "loss": 0.7448,
+      "step": 6160
+    },
+    {
+      "epoch": 342.77777777777777,
+      "grad_norm": 2.871325259054803,
+      "learning_rate": 1.438553963508866e-05,
+      "loss": 0.7414,
+      "step": 6170
+    },
+    {
+      "epoch": 343.3333333333333,
+      "grad_norm": 2.801729134977625,
+      "learning_rate": 1.428769483094658e-05,
+      "loss": 0.7169,
+      "step": 6180
+    },
+    {
+      "epoch": 343.8888888888889,
+      "grad_norm": 2.9124921806787136,
+      "learning_rate": 1.4190290585214231e-05,
+      "loss": 0.7569,
+      "step": 6190
+    },
+    {
+      "epoch": 344.44444444444446,
+      "grad_norm": 2.7035627947725422,
+      "learning_rate": 1.4093328418939246e-05,
+      "loss": 0.7237,
+      "step": 6200
+    },
+    {
+      "epoch": 345.0,
+      "grad_norm": 2.688322872780844,
+      "learning_rate": 1.3996809846265802e-05,
+      "loss": 0.745,
+      "step": 6210
+    },
+    {
+      "epoch": 345.55555555555554,
+      "grad_norm": 2.8913998776653975,
+      "learning_rate": 1.390073637441101e-05,
+      "loss": 0.7366,
+      "step": 6220
+    },
+    {
+      "epoch": 346.1111111111111,
+      "grad_norm": 2.6432723820102324,
+      "learning_rate": 1.3805109503641356e-05,
+      "loss": 0.7263,
+      "step": 6230
+    },
+    {
+      "epoch": 346.6666666666667,
+      "grad_norm": 2.9825456215118344,
+      "learning_rate": 1.3709930727249264e-05,
+      "loss": 0.7432,
+      "step": 6240
+    },
+    {
+      "epoch": 347.22222222222223,
+      "grad_norm": 2.7182717615198664,
+      "learning_rate": 1.3615201531529845e-05,
+      "loss": 0.7137,
+      "step": 6250
+    },
+    {
+      "epoch": 347.77777777777777,
+      "grad_norm": 2.9278420294270977,
+      "learning_rate": 1.3520923395757577e-05,
+      "loss": 0.7429,
+      "step": 6260
+    },
+    {
+      "epoch": 348.3333333333333,
+      "grad_norm": 2.915202963357083,
+      "learning_rate": 1.342709779216332e-05,
+      "loss": 0.7284,
+      "step": 6270
+    },
+    {
+      "epoch": 348.8888888888889,
+      "grad_norm": 2.746900979744846,
+      "learning_rate": 1.3333726185911221e-05,
+      "loss": 0.745,
+      "step": 6280
+    },
+    {
+      "epoch": 349.44444444444446,
+      "grad_norm": 2.916266622810717,
+      "learning_rate": 1.3240810035075934e-05,
+      "loss": 0.7109,
+      "step": 6290
+    },
+    {
+      "epoch": 350.0,
+      "grad_norm": 2.7171800838147697,
+      "learning_rate": 1.3148350790619758e-05,
+      "loss": 0.7365,
+      "step": 6300
+    },
+    {
+      "epoch": 350.55555555555554,
+      "grad_norm": 2.870664561883793,
+      "learning_rate": 1.3056349896370031e-05,
+      "loss": 0.7165,
+      "step": 6310
+    },
+    {
+      "epoch": 351.1111111111111,
+      "grad_norm": 2.772392054417928,
+      "learning_rate": 1.2964808788996606e-05,
+      "loss": 0.7032,
+      "step": 6320
+    },
+    {
+      "epoch": 351.6666666666667,
+      "grad_norm": 2.9073093903531064,
+      "learning_rate": 1.2873728897989336e-05,
+      "loss": 0.7276,
+      "step": 6330
+    },
+    {
+      "epoch": 352.22222222222223,
+      "grad_norm": 2.8394059758109953,
+      "learning_rate": 1.2783111645635851e-05,
+      "loss": 0.6919,
+      "step": 6340
+    },
+    {
+      "epoch": 352.77777777777777,
+      "grad_norm": 2.622035305137583,
+      "learning_rate": 1.2692958446999248e-05,
+      "loss": 0.7074,
+      "step": 6350
+    },
+    {
+      "epoch": 353.3333333333333,
+      "grad_norm": 2.896416218959587,
+      "learning_rate": 1.2603270709896092e-05,
+      "loss": 0.7316,
+      "step": 6360
+    },
+    {
+      "epoch": 353.8888888888889,
+      "grad_norm": 2.8702391306180663,
+      "learning_rate": 1.2514049834874331e-05,
+      "loss": 0.7065,
+      "step": 6370
+    },
+    {
+      "epoch": 354.44444444444446,
+      "grad_norm": 2.9084413532064737,
+      "learning_rate": 1.2425297215191518e-05,
+      "loss": 0.6905,
+      "step": 6380
+    },
+    {
+      "epoch": 355.0,
+      "grad_norm": 2.8951659130985354,
+      "learning_rate": 1.2337014236793001e-05,
+      "loss": 0.695,
+      "step": 6390
+    },
+    {
+      "epoch": 355.55555555555554,
+      "grad_norm": 2.790641615456117,
+      "learning_rate": 1.2249202278290264e-05,
+      "loss": 0.6659,
+      "step": 6400
+    },
+    {
+      "epoch": 356.1111111111111,
+      "grad_norm": 2.804789336754505,
+      "learning_rate": 1.2161862710939476e-05,
+      "loss": 0.7117,
+      "step": 6410
+    },
+    {
+      "epoch": 356.6666666666667,
+      "grad_norm": 2.763414005609754,
+      "learning_rate": 1.2074996898619982e-05,
+      "loss": 0.6788,
+      "step": 6420
+    },
+    {
+      "epoch": 357.22222222222223,
+      "grad_norm": 2.8966769812268223,
+      "learning_rate": 1.1988606197813097e-05,
+      "loss": 0.7103,
+      "step": 6430
+    },
+    {
+      "epoch": 357.77777777777777,
+      "grad_norm": 2.8711658287083286,
+      "learning_rate": 1.1902691957580829e-05,
+      "loss": 0.6801,
+      "step": 6440
+    },
+    {
+      "epoch": 358.3333333333333,
+      "grad_norm": 2.550419499915517,
+      "learning_rate": 1.181725551954491e-05,
+      "loss": 0.6762,
+      "step": 6450
+    },
+    {
+      "epoch": 358.8888888888889,
+      "grad_norm": 2.5633529546091705,
+      "learning_rate": 1.1732298217865753e-05,
+      "loss": 0.6815,
+      "step": 6460
+    },
+    {
+      "epoch": 359.44444444444446,
+      "grad_norm": 3.0112556071722096,
+      "learning_rate": 1.1647821379221695e-05,
+      "loss": 0.6774,
+      "step": 6470
+    },
+    {
+      "epoch": 360.0,
+      "grad_norm": 2.8739336772956574,
+      "learning_rate": 1.1563826322788229e-05,
+      "loss": 0.6664,
+      "step": 6480
+    },
+    {
+      "epoch": 360.55555555555554,
+      "grad_norm": 2.6734058835307675,
+      "learning_rate": 1.148031436021742e-05,
+      "loss": 0.6534,
+      "step": 6490
+    },
+    {
+      "epoch": 361.1111111111111,
+      "grad_norm": 2.741569079794837,
+      "learning_rate": 1.1397286795617434e-05,
+      "loss": 0.6832,
+      "step": 6500
+    },
+    {
+      "epoch": 361.6666666666667,
+      "grad_norm": 2.9360052323255768,
+      "learning_rate": 1.1314744925532153e-05,
+      "loss": 0.6568,
+      "step": 6510
+    },
+    {
+      "epoch": 362.22222222222223,
+      "grad_norm": 2.8512844002340536,
+      "learning_rate": 1.123269003892096e-05,
+      "loss": 0.6692,
+      "step": 6520
+    },
+    {
+      "epoch": 362.77777777777777,
+      "grad_norm": 3.033220574794864,
+      "learning_rate": 1.1151123417138555e-05,
+      "loss": 0.6729,
+      "step": 6530
+    },
+    {
+      "epoch": 363.3333333333333,
+      "grad_norm": 2.827834477966812,
+      "learning_rate": 1.1070046333915014e-05,
+      "loss": 0.6398,
+      "step": 6540
+    },
+    {
+      "epoch": 363.8888888888889,
+      "grad_norm": 2.9539937351188783,
+      "learning_rate": 1.0989460055335845e-05,
+      "loss": 0.6409,
+      "step": 6550
+    },
+    {
+      "epoch": 364.44444444444446,
+      "grad_norm": 2.7953159526294007,
+      "learning_rate": 1.0909365839822227e-05,
+      "loss": 0.6406,
+      "step": 6560
+    },
+    {
+      "epoch": 365.0,
+      "grad_norm": 2.7894481744897255,
+      "learning_rate": 1.0829764938111404e-05,
+      "loss": 0.6634,
+      "step": 6570
+    },
+    {
+      "epoch": 365.55555555555554,
+      "grad_norm": 2.813838047054896,
+      "learning_rate": 1.0750658593237084e-05,
+      "loss": 0.6688,
+      "step": 6580
+    },
+    {
+      "epoch": 366.1111111111111,
+      "grad_norm": 2.884819841230544,
+      "learning_rate": 1.0672048040510083e-05,
+      "loss": 0.6542,
+      "step": 6590
+    },
+    {
+      "epoch": 366.6666666666667,
+      "grad_norm": 2.7835962064798045,
+      "learning_rate": 1.0593934507498998e-05,
+      "loss": 0.6646,
+      "step": 6600
+    },
+    {
+      "epoch": 367.22222222222223,
+      "grad_norm": 2.9568443491235463,
+      "learning_rate": 1.051631921401107e-05,
+      "loss": 0.6488,
+      "step": 6610
+    },
+    {
+      "epoch": 367.77777777777777,
+      "grad_norm": 2.9165140021197478,
+      "learning_rate": 1.0439203372073096e-05,
+      "loss": 0.6486,
+      "step": 6620
+    },
+    {
+      "epoch": 368.3333333333333,
+      "grad_norm": 3.0764659147663638,
+      "learning_rate": 1.0362588185912561e-05,
+      "loss": 0.6567,
+      "step": 6630
+    },
+    {
+      "epoch": 368.8888888888889,
+      "grad_norm": 2.6813972735028258,
+      "learning_rate": 1.0286474851938762e-05,
+      "loss": 0.6292,
+      "step": 6640
+    },
+    {
+      "epoch": 369.44444444444446,
+      "grad_norm": 2.9418457027197804,
+      "learning_rate": 1.0210864558724168e-05,
+      "loss": 0.6441,
+      "step": 6650
+    },
+    {
+      "epoch": 370.0,
+      "grad_norm": 2.8125066121341544,
+      "learning_rate": 1.013575848698587e-05,
+      "loss": 0.6494,
+      "step": 6660
+    },
+    {
+      "epoch": 370.55555555555554,
+      "grad_norm": 2.9225546306532633,
+      "learning_rate": 1.0061157809567101e-05,
+      "loss": 0.6155,
+      "step": 6670
+    },
+    {
+      "epoch": 371.1111111111111,
+      "grad_norm": 2.612613062260089,
+      "learning_rate": 9.987063691418975e-06,
+      "loss": 0.6385,
+      "step": 6680
+    },
+    {
+      "epoch": 371.6666666666667,
+      "grad_norm": 2.8319054726918775,
+      "learning_rate": 9.91347728958222e-06,
+      "loss": 0.6394,
+      "step": 6690
+    },
+    {
+      "epoch": 372.22222222222223,
+      "grad_norm": 2.7340304784352605,
+      "learning_rate": 9.840399753169205e-06,
+      "loss": 0.6191,
+      "step": 6700
+    },
+    {
+      "epoch": 372.77777777777777,
+      "grad_norm": 2.7919406946163163,
+      "learning_rate": 9.767832223345916e-06,
+      "loss": 0.6342,
+      "step": 6710
+    },
+    {
+      "epoch": 373.3333333333333,
+      "grad_norm": 2.7706127126719435,
+      "learning_rate": 9.695775833314159e-06,
+      "loss": 0.6403,
+      "step": 6720
+    },
+    {
+      "epoch": 373.8888888888889,
+      "grad_norm": 2.7971277389134506,
+      "learning_rate": 9.624231708293902e-06,
+      "loss": 0.6321,
+      "step": 6730
+    },
+    {
+      "epoch": 374.44444444444446,
+      "grad_norm": 2.577292568025829,
+      "learning_rate": 9.553200965505643e-06,
+      "loss": 0.6039,
+      "step": 6740
+    },
+    {
+      "epoch": 375.0,
+      "grad_norm": 2.7401178706356424,
+      "learning_rate": 9.482684714153013e-06,
+      "loss": 0.6327,
+      "step": 6750
+    },
+    {
+      "epoch": 375.55555555555554,
+      "grad_norm": 2.7554109399115814,
+      "learning_rate": 9.41268405540541e-06,
+      "loss": 0.602,
+      "step": 6760
+    },
+    {
+      "epoch": 376.1111111111111,
+      "grad_norm": 2.8533690700194483,
+      "learning_rate": 9.343200082380868e-06,
+      "loss": 0.617,
+      "step": 6770
+    },
+    {
+      "epoch": 376.6666666666667,
+      "grad_norm": 2.8445222246277786,
+      "learning_rate": 9.274233880128898e-06,
+      "loss": 0.6302,
+      "step": 6780
+    },
+    {
+      "epoch": 377.22222222222223,
+      "grad_norm": 2.9577964689583065,
+      "learning_rate": 9.205786525613647e-06,
+      "loss": 0.6241,
+      "step": 6790
+    },
+    {
+      "epoch": 377.77777777777777,
+      "grad_norm": 2.5828334468262155,
+      "learning_rate": 9.13785908769698e-06,
+      "loss": 0.6187,
+      "step": 6800
+    },
+    {
+      "epoch": 378.3333333333333,
+      "grad_norm": 2.838979555470915,
+      "learning_rate": 9.070452627121857e-06,
+      "loss": 0.6212,
+      "step": 6810
+    },
+    {
+      "epoch": 378.8888888888889,
+      "grad_norm": 2.5910199976642208,
+      "learning_rate": 9.00356819649575e-06,
+      "loss": 0.5848,
+      "step": 6820
+    },
+    {
+      "epoch": 379.44444444444446,
+      "grad_norm": 2.8751809436414955,
+      "learning_rate": 8.937206840274185e-06,
+      "loss": 0.6233,
+      "step": 6830
+    },
+    {
+      "epoch": 380.0,
+      "grad_norm": 2.7800721664118297,
+      "learning_rate": 8.871369594744485e-06,
+      "loss": 0.5943,
+      "step": 6840
+    },
+    {
+      "epoch": 380.55555555555554,
+      "grad_norm": 2.7187237793107446,
+      "learning_rate": 8.806057488009507e-06,
+      "loss": 0.604,
+      "step": 6850
+    },
+    {
+      "epoch": 381.1111111111111,
+      "grad_norm": 2.8786798230262662,
+      "learning_rate": 8.741271539971674e-06,
+      "loss": 0.6238,
+      "step": 6860
+    },
+    {
+      "epoch": 381.6666666666667,
+      "grad_norm": 2.865893329133973,
+      "learning_rate": 8.677012762316979e-06,
+      "loss": 0.6132,
+      "step": 6870
+    },
+    {
+      "epoch": 382.22222222222223,
+      "grad_norm": 2.8051210031232316,
+      "learning_rate": 8.613282158499221e-06,
+      "loss": 0.611,
+      "step": 6880
+    },
+    {
+      "epoch": 382.77777777777777,
+      "grad_norm": 2.754483036755567,
+      "learning_rate": 8.550080723724344e-06,
+      "loss": 0.5919,
+      "step": 6890
+    },
+    {
+      "epoch": 383.3333333333333,
+      "grad_norm": 2.8570644027292724,
+      "learning_rate": 8.487409444934853e-06,
+      "loss": 0.6108,
+      "step": 6900
+    },
+    {
+      "epoch": 383.8888888888889,
+      "grad_norm": 2.7169434256281386,
+      "learning_rate": 8.425269300794467e-06,
+      "loss": 0.5928,
+      "step": 6910
+    },
+    {
+      "epoch": 384.44444444444446,
+      "grad_norm": 2.896855583249299,
+      "learning_rate": 8.363661261672756e-06,
+      "loss": 0.5903,
+      "step": 6920
+    },
+    {
+      "epoch": 385.0,
+      "grad_norm": 2.846043426965124,
+      "learning_rate": 8.302586289630078e-06,
+      "loss": 0.6054,
+      "step": 6930
+    },
+    {
+      "epoch": 385.55555555555554,
+      "grad_norm": 2.8455182959987746,
+      "learning_rate": 8.242045338402464e-06,
+      "loss": 0.596,
+      "step": 6940
+    },
+    {
+      "epoch": 386.1111111111111,
+      "grad_norm": 2.632978394486001,
+      "learning_rate": 8.182039353386808e-06,
+      "loss": 0.5864,
+      "step": 6950
+    },
+    {
+      "epoch": 386.6666666666667,
+      "grad_norm": 2.806716057832868,
+      "learning_rate": 8.122569271626042e-06,
+      "loss": 0.6074,
+      "step": 6960
+    },
+    {
+      "epoch": 387.22222222222223,
+      "grad_norm": 2.617582285697988,
+      "learning_rate": 8.063636021794522e-06,
+      "loss": 0.5783,
+      "step": 6970
+    },
+    {
+      "epoch": 387.77777777777777,
+      "grad_norm": 2.826185963245596,
+      "learning_rate": 8.005240524183561e-06,
+      "loss": 0.5749,
+      "step": 6980
+    },
+    {
+      "epoch": 388.3333333333333,
+      "grad_norm": 2.766603199870456,
+      "learning_rate": 7.94738369068699e-06,
+      "loss": 0.5853,
+      "step": 6990
+    },
+    {
+      "epoch": 388.8888888888889,
+      "grad_norm": 2.699434440154436,
+      "learning_rate": 7.890066424786995e-06,
+      "loss": 0.577,
+      "step": 7000
+    },
+    {
+      "epoch": 389.44444444444446,
+      "grad_norm": 2.8440501295733815,
+      "learning_rate": 7.833289621539925e-06,
+      "loss": 0.5872,
+      "step": 7010
+    },
+    {
+      "epoch": 390.0,
+      "grad_norm": 2.7682732440720543,
+      "learning_rate": 7.777054167562405e-06,
+      "loss": 0.5697,
+      "step": 7020
+    },
+    {
+      "epoch": 390.55555555555554,
+      "grad_norm": 2.6857153711345765,
+      "learning_rate": 7.721360941017406e-06,
+      "loss": 0.5781,
+      "step": 7030
+    },
+    {
+      "epoch": 391.1111111111111,
+      "grad_norm": 2.639350538418199,
+      "learning_rate": 7.66621081160059e-06,
+      "loss": 0.5595,
+      "step": 7040
+    },
+    {
+      "epoch": 391.6666666666667,
+      "grad_norm": 2.692937274302927,
+      "learning_rate": 7.611604640526709e-06,
+      "loss": 0.5612,
+      "step": 7050
+    },
+    {
+      "epoch": 392.22222222222223,
+      "grad_norm": 2.646228240170974,
+      "learning_rate": 7.557543280516148e-06,
+      "loss": 0.567,
+      "step": 7060
+    },
+    {
+      "epoch": 392.77777777777777,
+      "grad_norm": 2.637918111916942,
+      "learning_rate": 7.50402757578163e-06,
+      "loss": 0.5695,
+      "step": 7070
+    },
+    {
+      "epoch": 393.3333333333333,
+      "grad_norm": 2.688015555677362,
+      "learning_rate": 7.451058362015012e-06,
+      "loss": 0.578,
+      "step": 7080
+    },
+    {
+      "epoch": 393.8888888888889,
+      "grad_norm": 2.5645011903161707,
+      "learning_rate": 7.398636466374252e-06,
+      "loss": 0.5575,
+      "step": 7090
+    },
+    {
+      "epoch": 394.44444444444446,
+      "grad_norm": 2.972626895233265,
+      "learning_rate": 7.3467627074704676e-06,
+      "loss": 0.5817,
+      "step": 7100
+    },
+    {
+      "epoch": 395.0,
+      "grad_norm": 2.860270709050585,
+      "learning_rate": 7.295437895355191e-06,
+      "loss": 0.551,
+      "step": 7110
+    },
+    {
+      "epoch": 395.55555555555554,
+      "grad_norm": 2.6787784024155576,
+      "learning_rate": 7.244662831507682e-06,
+      "loss": 0.5576,
+      "step": 7120
+    },
+    {
+      "epoch": 396.1111111111111,
+      "grad_norm": 2.8897870473050387,
+      "learning_rate": 7.194438308822426e-06,
+      "loss": 0.5563,
+      "step": 7130
+    },
+    {
+      "epoch": 396.6666666666667,
+      "grad_norm": 2.6141621426450063,
+      "learning_rate": 7.144765111596779e-06,
+      "loss": 0.578,
+      "step": 7140
+    },
+    {
+      "epoch": 397.22222222222223,
+      "grad_norm": 2.7693278207033796,
+      "learning_rate": 7.095644015518659e-06,
+      "loss": 0.5434,
+      "step": 7150
+    },
+    {
+      "epoch": 397.77777777777777,
+      "grad_norm": 2.6814667158420864,
+      "learning_rate": 7.0470757876545026e-06,
+      "loss": 0.545,
+      "step": 7160
+    },
+    {
+      "epoch": 398.3333333333333,
+      "grad_norm": 2.7744477487175296,
+      "learning_rate": 6.999061186437224e-06,
+      "loss": 0.5535,
+      "step": 7170
+    },
+    {
+      "epoch": 398.8888888888889,
+      "grad_norm": 2.7369126234062353,
+      "learning_rate": 6.951600961654432e-06,
+      "loss": 0.542,
+      "step": 7180
+    },
+    {
+      "epoch": 399.44444444444446,
+      "grad_norm": 2.583087468496482,
+      "learning_rate": 6.9046958544366604e-06,
+      "loss": 0.5409,
+      "step": 7190
+    },
+    {
+      "epoch": 400.0,
+      "grad_norm": 2.7132184110043536,
+      "learning_rate": 6.8583465972458365e-06,
+      "loss": 0.5656,
+      "step": 7200
+    },
+    {
+      "epoch": 400.55555555555554,
+      "grad_norm": 2.6385953427615645,
+      "learning_rate": 6.81255391386384e-06,
+      "loss": 0.5489,
+      "step": 7210
+    },
+    {
+      "epoch": 401.1111111111111,
+      "grad_norm": 2.7963972482804422,
+      "learning_rate": 6.767318519381177e-06,
+      "loss": 0.5765,
+      "step": 7220
+    },
+    {
+      "epoch": 401.6666666666667,
+      "grad_norm": 2.8913317666900147,
+      "learning_rate": 6.722641120185843e-06,
+      "loss": 0.5291,
+      "step": 7230
+    },
+    {
+      "epoch": 402.22222222222223,
+      "grad_norm": 2.7991058147963535,
+      "learning_rate": 6.678522413952258e-06,
+      "loss": 0.5565,
+      "step": 7240
+    },
+    {
+      "epoch": 402.77777777777777,
+      "grad_norm": 2.7454729965791085,
+      "learning_rate": 6.63496308963041e-06,
+      "loss": 0.5441,
+      "step": 7250
+    },
+    {
+      "epoch": 403.3333333333333,
+      "grad_norm": 2.72165270662981,
+      "learning_rate": 6.5919638274350614e-06,
+      "loss": 0.5516,
+      "step": 7260
+    },
+    {
+      "epoch": 403.8888888888889,
+      "grad_norm": 2.7195468480233522,
+      "learning_rate": 6.549525298835157e-06,
+      "loss": 0.5456,
+      "step": 7270
+    },
+    {
+      "epoch": 404.44444444444446,
+      "grad_norm": 2.647449140195266,
+      "learning_rate": 6.50764816654331e-06,
+      "loss": 0.5465,
+      "step": 7280
+    },
+    {
+      "epoch": 405.0,
+      "grad_norm": 2.767835950990054,
+      "learning_rate": 6.4663330845054775e-06,
+      "loss": 0.5488,
+      "step": 7290
+    },
+    {
+      "epoch": 405.55555555555554,
+      "grad_norm": 2.497583556586515,
+      "learning_rate": 6.425580697890733e-06,
+      "loss": 0.5491,
+      "step": 7300
+    },
+    {
+      "epoch": 406.1111111111111,
+      "grad_norm": 2.6155962761289646,
+      "learning_rate": 6.385391643081197e-06,
+      "loss": 0.5268,
+      "step": 7310
+    },
+    {
+      "epoch": 406.6666666666667,
+      "grad_norm": 2.680320108995982,
+      "learning_rate": 6.34576654766211e-06,
+      "loss": 0.5489,
+      "step": 7320
+    },
+    {
+      "epoch": 407.22222222222223,
+      "grad_norm": 2.6284274477205773,
+      "learning_rate": 6.306706030412002e-06,
+      "loss": 0.5337,
+      "step": 7330
+    },
+    {
+      "epoch": 407.77777777777777,
+      "grad_norm": 2.6927732143787986,
+      "learning_rate": 6.268210701293069e-06,
+      "loss": 0.5509,
+      "step": 7340
+    },
+    {
+      "epoch": 408.3333333333333,
+      "grad_norm": 2.559253403752575,
+      "learning_rate": 6.23028116144161e-06,
+      "loss": 0.5496,
+      "step": 7350
+    },
+    {
+      "epoch": 408.8888888888889,
+      "grad_norm": 2.5081958893625678,
+      "learning_rate": 6.192918003158677e-06,
+      "loss": 0.5133,
+      "step": 7360
+    },
+    {
+      "epoch": 409.44444444444446,
+      "grad_norm": 2.534004220051251,
+      "learning_rate": 6.156121809900792e-06,
+      "loss": 0.4953,
+      "step": 7370
+    },
+    {
+      "epoch": 410.0,
+      "grad_norm": 2.7899342090957564,
+      "learning_rate": 6.1198931562708504e-06,
+      "loss": 0.5258,
+      "step": 7380
+    },
+    {
+      "epoch": 410.55555555555554,
+      "grad_norm": 2.786730306786951,
+      "learning_rate": 6.084232608009166e-06,
+      "loss": 0.5211,
+      "step": 7390
+    },
+    {
+      "epoch": 411.1111111111111,
+      "grad_norm": 2.538047611171766,
+      "learning_rate": 6.049140721984598e-06,
+      "loss": 0.5279,
+      "step": 7400
+    },
+    {
+      "epoch": 411.6666666666667,
+      "grad_norm": 2.830534525019375,
+      "learning_rate": 6.014618046185887e-06,
+      "loss": 0.547,
+      "step": 7410
+    },
+    {
+      "epoch": 412.22222222222223,
+      "grad_norm": 2.5429736253073467,
+      "learning_rate": 5.980665119713077e-06,
+      "loss": 0.5159,
+      "step": 7420
+    },
+    {
+      "epoch": 412.77777777777777,
+      "grad_norm": 2.531166546234726,
+      "learning_rate": 5.947282472769127e-06,
+      "loss": 0.5055,
+      "step": 7430
+    },
+    {
+      "epoch": 413.3333333333333,
+      "grad_norm": 2.7091377726429657,
+      "learning_rate": 5.914470626651588e-06,
+      "loss": 0.4933,
+      "step": 7440
+    },
+    {
+      "epoch": 413.8888888888889,
+      "grad_norm": 2.770821640640101,
+      "learning_rate": 5.882230093744495e-06,
+      "loss": 0.5171,
+      "step": 7450
+    },
+    {
+      "epoch": 414.44444444444446,
+      "grad_norm": 2.6847000618028978,
+      "learning_rate": 5.8505613775103555e-06,
+      "loss": 0.4937,
+      "step": 7460
+    },
+    {
+      "epoch": 415.0,
+      "grad_norm": 2.745730727444083,
+      "learning_rate": 5.819464972482287e-06,
+      "loss": 0.5074,
+      "step": 7470
+    },
+    {
+      "epoch": 415.55555555555554,
+      "grad_norm": 2.6242149632373506,
+      "learning_rate": 5.788941364256304e-06,
+      "loss": 0.4918,
+      "step": 7480
+    },
+    {
+      "epoch": 416.1111111111111,
+      "grad_norm": 2.7233746537241474,
+      "learning_rate": 5.758991029483714e-06,
+      "loss": 0.5278,
+      "step": 7490
+    },
+    {
+      "epoch": 416.6666666666667,
+      "grad_norm": 2.5643715885445544,
+      "learning_rate": 5.7296144358636925e-06,
+      "loss": 0.4966,
+      "step": 7500
+    },
+    {
+      "epoch": 417.22222222222223,
+      "grad_norm": 2.682906283481785,
+      "learning_rate": 5.700812042135969e-06,
+      "loss": 0.5126,
+      "step": 7510
+    },
+    {
+      "epoch": 417.77777777777777,
+      "grad_norm": 2.742571942369028,
+      "learning_rate": 5.672584298073685e-06,
+      "loss": 0.522,
+      "step": 7520
+    },
+    {
+      "epoch": 418.3333333333333,
+      "grad_norm": 2.7996087379928554,
+      "learning_rate": 5.644931644476331e-06,
+      "loss": 0.5155,
+      "step": 7530
+    },
+    {
+      "epoch": 418.8888888888889,
+      "grad_norm": 2.5961986152071015,
+      "learning_rate": 5.617854513162896e-06,
+      "loss": 0.4972,
+      "step": 7540
+    },
+    {
+      "epoch": 419.44444444444446,
+      "grad_norm": 2.7430442927863012,
+      "learning_rate": 5.5913533269651185e-06,
+      "loss": 0.5203,
+      "step": 7550
+    },
+    {
+      "epoch": 420.0,
+      "grad_norm": 2.760347319877515,
+      "learning_rate": 5.565428499720873e-06,
+      "loss": 0.4957,
+      "step": 7560
+    },
+    {
+      "epoch": 420.55555555555554,
+      "grad_norm": 2.6756870670867587,
+      "learning_rate": 5.540080436267716e-06,
+      "loss": 0.5274,
+      "step": 7570
+    },
+    {
+      "epoch": 421.1111111111111,
+      "grad_norm": 2.590524989499604,
+      "learning_rate": 5.515309532436559e-06,
+      "loss": 0.5215,
+      "step": 7580
+    },
+    {
+      "epoch": 421.6666666666667,
+      "grad_norm": 2.5621487446531637,
+      "learning_rate": 5.491116175045491e-06,
+      "loss": 0.503,
+      "step": 7590
+    },
+    {
+      "epoch": 422.22222222222223,
+      "grad_norm": 2.575305853891702,
+      "learning_rate": 5.467500741893746e-06,
+      "loss": 0.4951,
+      "step": 7600
+    },
+    {
+      "epoch": 422.77777777777777,
+      "grad_norm": 2.654692877842046,
+      "learning_rate": 5.444463601755772e-06,
+      "loss": 0.4821,
+      "step": 7610
+    },
+    {
+      "epoch": 423.3333333333333,
+      "grad_norm": 2.781368295961238,
+      "learning_rate": 5.422005114375519e-06,
+      "loss": 0.5198,
+      "step": 7620
+    },
+    {
+      "epoch": 423.8888888888889,
+      "grad_norm": 2.8118485438011,
+      "learning_rate": 5.400125630460786e-06,
+      "loss": 0.495,
+      "step": 7630
+    },
+    {
+      "epoch": 424.44444444444446,
+      "grad_norm": 2.795252409236338,
+      "learning_rate": 5.37882549167776e-06,
+      "loss": 0.5229,
+      "step": 7640
+    },
+    {
+      "epoch": 425.0,
+      "grad_norm": 2.3883490605534163,
+      "learning_rate": 5.358105030645666e-06,
+      "loss": 0.4733,
+      "step": 7650
+    },
+    {
+      "epoch": 425.55555555555554,
+      "grad_norm": 2.7384162031540744,
+      "learning_rate": 5.337964570931595e-06,
+      "loss": 0.4988,
+      "step": 7660
+    },
+    {
+      "epoch": 426.1111111111111,
+      "grad_norm": 2.552065001960365,
+      "learning_rate": 5.31840442704543e-06,
+      "loss": 0.4777,
+      "step": 7670
+    },
+    {
+      "epoch": 426.6666666666667,
+      "grad_norm": 2.6193294129970566,
+      "learning_rate": 5.299424904434947e-06,
+      "loss": 0.4865,
+      "step": 7680
+    },
+    {
+      "epoch": 427.22222222222223,
+      "grad_norm": 2.6106971543349933,
+      "learning_rate": 5.281026299481048e-06,
+      "loss": 0.4764,
+      "step": 7690
+    },
+    {
+      "epoch": 427.77777777777777,
+      "grad_norm": 2.757888144517409,
+      "learning_rate": 5.263208899493116e-06,
+      "loss": 0.4845,
+      "step": 7700
+    },
+    {
+      "epoch": 428.3333333333333,
+      "grad_norm": 3.0188954067553766,
+      "learning_rate": 5.245972982704546e-06,
+      "loss": 0.4835,
+      "step": 7710
+    },
+    {
+      "epoch": 428.8888888888889,
+      "grad_norm": 2.558215117648491,
+      "learning_rate": 5.229318818268392e-06,
+      "loss": 0.4971,
+      "step": 7720
+    },
+    {
+      "epoch": 429.44444444444446,
+      "grad_norm": 2.566720046276857,
+      "learning_rate": 5.2132466662531654e-06,
+      "loss": 0.4786,
+      "step": 7730
+    },
+    {
+      "epoch": 430.0,
+      "grad_norm": 2.7927892240433865,
+      "learning_rate": 5.197756777638772e-06,
+      "loss": 0.5024,
+      "step": 7740
+    },
+    {
+      "epoch": 430.55555555555554,
+      "grad_norm": 2.5587102266545965,
+      "learning_rate": 5.182849394312596e-06,
+      "loss": 0.4957,
+      "step": 7750
+    },
+    {
+      "epoch": 431.1111111111111,
+      "grad_norm": 2.6402876598965412,
+      "learning_rate": 5.1685247490657234e-06,
+      "loss": 0.4839,
+      "step": 7760
+    },
+    {
+      "epoch": 431.6666666666667,
+      "grad_norm": 2.8683833401133545,
+      "learning_rate": 5.154783065589303e-06,
+      "loss": 0.4802,
+      "step": 7770
+    },
+    {
+      "epoch": 432.22222222222223,
+      "grad_norm": 2.9947095953711287,
+      "learning_rate": 5.141624558471049e-06,
+      "loss": 0.4626,
+      "step": 7780
+    },
+    {
+      "epoch": 432.77777777777777,
+      "grad_norm": 2.645846469476085,
+      "learning_rate": 5.1290494331919025e-06,
+      "loss": 0.492,
+      "step": 7790
+    },
+    {
+      "epoch": 433.3333333333333,
+      "grad_norm": 2.5495497810993437,
+      "learning_rate": 5.117057886122817e-06,
+      "loss": 0.481,
+      "step": 7800
+    },
+    {
+      "epoch": 433.8888888888889,
+      "grad_norm": 2.7517164249030195,
+      "learning_rate": 5.105650104521688e-06,
+      "loss": 0.4944,
+      "step": 7810
+    },
+    {
+      "epoch": 434.44444444444446,
+      "grad_norm": 2.768363415874053,
+      "learning_rate": 5.094826266530429e-06,
+      "loss": 0.4732,
+      "step": 7820
+    },
+    {
+      "epoch": 435.0,
+      "grad_norm": 2.8950166900100958,
+      "learning_rate": 5.0845865411722e-06,
+      "loss": 0.4795,
+      "step": 7830
+    },
+    {
+      "epoch": 435.55555555555554,
+      "grad_norm": 2.5118072161957428,
+      "learning_rate": 5.074931088348753e-06,
+      "loss": 0.4841,
+      "step": 7840
+    },
+    {
+      "epoch": 436.1111111111111,
+      "grad_norm": 2.459145708071279,
+      "learning_rate": 5.06586005883795e-06,
+      "loss": 0.477,
+      "step": 7850
+    },
+    {
+      "epoch": 436.6666666666667,
+      "grad_norm": 2.562951583148991,
+      "learning_rate": 5.057373594291389e-06,
+      "loss": 0.471,
+      "step": 7860
+    },
+    {
+      "epoch": 437.22222222222223,
+      "grad_norm": 2.7214780253038815,
+      "learning_rate": 5.049471827232223e-06,
+      "loss": 0.4791,
+      "step": 7870
+    },
+    {
+      "epoch": 437.77777777777777,
+      "grad_norm": 2.5544098402782973,
+      "learning_rate": 5.042154881053054e-06,
+      "loss": 0.443,
+      "step": 7880
+    },
+    {
+      "epoch": 438.3333333333333,
+      "grad_norm": 2.655628894669282,
+      "learning_rate": 5.0354228700140305e-06,
+      "loss": 0.4749,
+      "step": 7890
+    },
+    {
+      "epoch": 438.8888888888889,
+      "grad_norm": 2.8925240541874038,
+      "learning_rate": 5.0292758992410565e-06,
+      "loss": 0.4843,
+      "step": 7900
+    },
+    {
+      "epoch": 439.44444444444446,
+      "grad_norm": 2.684336019639309,
+      "learning_rate": 5.023714064724151e-06,
+      "loss": 0.4548,
+      "step": 7910
+    },
+    {
+      "epoch": 440.0,
+      "grad_norm": 2.819419402422093,
+      "learning_rate": 5.018737453315944e-06,
+      "loss": 0.4702,
+      "step": 7920
+    },
+    {
+      "epoch": 440.55555555555554,
+      "grad_norm": 2.791918491753208,
+      "learning_rate": 5.014346142730328e-06,
+      "loss": 0.4825,
+      "step": 7930
+    },
+    {
+      "epoch": 441.1111111111111,
+      "grad_norm": 2.8268336092786077,
+      "learning_rate": 5.010540201541244e-06,
+      "loss": 0.4691,
+      "step": 7940
+    },
+    {
+      "epoch": 441.6666666666667,
+      "grad_norm": 2.5252259710437768,
+      "learning_rate": 5.007319689181595e-06,
+      "loss": 0.4513,
+      "step": 7950
+    },
+    {
+      "epoch": 442.22222222222223,
+      "grad_norm": 2.6507762383847315,
+      "learning_rate": 5.004684655942338e-06,
+      "loss": 0.4628,
+      "step": 7960
+    },
+    {
+      "epoch": 442.77777777777777,
+      "grad_norm": 2.7600387794423975,
+      "learning_rate": 5.0026351429716924e-06,
+      "loss": 0.4717,
+      "step": 7970
+    },
+    {
+      "epoch": 443.3333333333333,
+      "grad_norm": 2.643279056673312,
+      "learning_rate": 5.0011711822744915e-06,
+      "loss": 0.4631,
+      "step": 7980
+    },
+    {
+      "epoch": 443.8888888888889,
+      "grad_norm": 2.913324584802314,
+      "learning_rate": 5.000292796711689e-06,
+      "loss": 0.4704,
+      "step": 7990
+    },
+    {
+      "epoch": 444.44444444444446,
+      "grad_norm": 2.828079476174524,
+      "learning_rate": 5e-06,
+      "loss": 0.4382,
+      "step": 8000
+    },
+    {
+      "epoch": 444.44444444444446,
+      "step": 8000,
+      "total_flos": 3925623177216000.0,
+      "train_loss": 1.7424663036465644,
+      "train_runtime": 126880.6464,
+      "train_samples_per_second": 12.106,
+      "train_steps_per_second": 0.063
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 445,
+  "save_steps": 4000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3925623177216000.0,
+  "train_batch_size": 6,
+  "trial_name": null,
+  "trial_params": null
+}