diff --git "a/UNIVLA_LIBERO_VIDEO_BS192_8K/trainer_state.json" "b/UNIVLA_LIBERO_VIDEO_BS192_8K/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/UNIVLA_LIBERO_VIDEO_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": 11.443616555236428,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 17.9429,
+      "step": 10
+    },
+    {
+      "epoch": 1.1111111111111112,
+      "grad_norm": 23.24444973234202,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 9.8247,
+      "step": 20
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 9.360798162353898,
+      "learning_rate": 4.8e-05,
+      "loss": 6.2166,
+      "step": 30
+    },
+    {
+      "epoch": 2.2222222222222223,
+      "grad_norm": 9.20499089080509,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 5.6225,
+      "step": 40
+    },
+    {
+      "epoch": 2.7777777777777777,
+      "grad_norm": 4.493504392573152,
+      "learning_rate": 8e-05,
+      "loss": 5.3787,
+      "step": 50
+    },
+    {
+      "epoch": 3.3333333333333335,
+      "grad_norm": 5.684409842612429,
+      "learning_rate": 7.999970720328831e-05,
+      "loss": 5.2701,
+      "step": 60
+    },
+    {
+      "epoch": 3.888888888888889,
+      "grad_norm": 5.9749678734590015,
+      "learning_rate": 7.999882881772551e-05,
+      "loss": 5.0402,
+      "step": 70
+    },
+    {
+      "epoch": 4.444444444444445,
+      "grad_norm": 22.987031680673372,
+      "learning_rate": 7.999736485702831e-05,
+      "loss": 4.8761,
+      "step": 80
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 14.93644474404123,
+      "learning_rate": 7.999531534405768e-05,
+      "loss": 4.7698,
+      "step": 90
+    },
+    {
+      "epoch": 5.555555555555555,
+      "grad_norm": 6.9976911824376495,
+      "learning_rate": 7.999268031081841e-05,
+      "loss": 4.6613,
+      "step": 100
+    },
+    {
+      "epoch": 6.111111111111111,
+      "grad_norm": 19.732921314233987,
+      "learning_rate": 7.998945979845876e-05,
+      "loss": 4.5797,
+      "step": 110
+    },
+    {
+      "epoch": 6.666666666666667,
+      "grad_norm": 4.984793618539648,
+      "learning_rate": 7.998565385726968e-05,
+      "loss": 4.5138,
+      "step": 120
+    },
+    {
+      "epoch": 7.222222222222222,
+      "grad_norm": 10.872698660952349,
+      "learning_rate": 7.998126254668407e-05,
+      "loss": 4.4941,
+      "step": 130
+    },
+    {
+      "epoch": 7.777777777777778,
+      "grad_norm": 5.847254390834455,
+      "learning_rate": 7.997628593527586e-05,
+      "loss": 4.3622,
+      "step": 140
+    },
+    {
+      "epoch": 8.333333333333334,
+      "grad_norm": 20.73574596894628,
+      "learning_rate": 7.997072410075895e-05,
+      "loss": 4.2549,
+      "step": 150
+    },
+    {
+      "epoch": 8.88888888888889,
+      "grad_norm": 7.055924046356311,
+      "learning_rate": 7.996457712998597e-05,
+      "loss": 4.2908,
+      "step": 160
+    },
+    {
+      "epoch": 9.444444444444445,
+      "grad_norm": 4.302749003464693,
+      "learning_rate": 7.995784511894694e-05,
+      "loss": 4.1524,
+      "step": 170
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 10.92102938149444,
+      "learning_rate": 7.995052817276779e-05,
+      "loss": 4.1635,
+      "step": 180
+    },
+    {
+      "epoch": 10.555555555555555,
+      "grad_norm": 4.153956189015921,
+      "learning_rate": 7.994262640570861e-05,
+      "loss": 4.0984,
+      "step": 190
+    },
+    {
+      "epoch": 11.11111111111111,
+      "grad_norm": 11.914720140661023,
+      "learning_rate": 7.993413994116206e-05,
+      "loss": 4.0041,
+      "step": 200
+    },
+    {
+      "epoch": 11.666666666666666,
+      "grad_norm": 4.327135103687705,
+      "learning_rate": 7.992506891165125e-05,
+      "loss": 4.0425,
+      "step": 210
+    },
+    {
+      "epoch": 12.222222222222221,
+      "grad_norm": 4.096190213048536,
+      "learning_rate": 7.991541345882781e-05,
+      "loss": 3.928,
+      "step": 220
+    },
+    {
+      "epoch": 12.777777777777779,
+      "grad_norm": 7.934743808199343,
+      "learning_rate": 7.990517373346957e-05,
+      "loss": 3.8661,
+      "step": 230
+    },
+    {
+      "epoch": 13.333333333333334,
+      "grad_norm": 14.253677161422953,
+      "learning_rate": 7.989434989547832e-05,
+      "loss": 3.8257,
+      "step": 240
+    },
+    {
+      "epoch": 13.88888888888889,
+      "grad_norm": 7.913018145477381,
+      "learning_rate": 7.988294211387719e-05,
+      "loss": 3.7643,
+      "step": 250
+    },
+    {
+      "epoch": 14.444444444444445,
+      "grad_norm": 10.518653286213326,
+      "learning_rate": 7.987095056680811e-05,
+      "loss": 3.7103,
+      "step": 260
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 6.066158174157387,
+      "learning_rate": 7.985837544152896e-05,
+      "loss": 3.6302,
+      "step": 270
+    },
+    {
+      "epoch": 15.555555555555555,
+      "grad_norm": 4.235557250344666,
+      "learning_rate": 7.984521693441071e-05,
+      "loss": 3.6146,
+      "step": 280
+    },
+    {
+      "epoch": 16.11111111111111,
+      "grad_norm": 3.7489455965968186,
+      "learning_rate": 7.983147525093428e-05,
+      "loss": 3.5628,
+      "step": 290
+    },
+    {
+      "epoch": 16.666666666666668,
+      "grad_norm": 5.302587079273321,
+      "learning_rate": 7.981715060568741e-05,
+      "loss": 3.7421,
+      "step": 300
+    },
+    {
+      "epoch": 17.22222222222222,
+      "grad_norm": 6.258399202463245,
+      "learning_rate": 7.980224322236123e-05,
+      "loss": 3.6361,
+      "step": 310
+    },
+    {
+      "epoch": 17.77777777777778,
+      "grad_norm": 3.278173693939838,
+      "learning_rate": 7.978675333374685e-05,
+      "loss": 3.5182,
+      "step": 320
+    },
+    {
+      "epoch": 18.333333333333332,
+      "grad_norm": 8.914517390206436,
+      "learning_rate": 7.977068118173162e-05,
+      "loss": 3.4705,
+      "step": 330
+    },
+    {
+      "epoch": 18.88888888888889,
+      "grad_norm": 6.7939971794636875,
+      "learning_rate": 7.975402701729546e-05,
+      "loss": 3.5169,
+      "step": 340
+    },
+    {
+      "epoch": 19.444444444444443,
+      "grad_norm": 4.972779757156313,
+      "learning_rate": 7.97367911005069e-05,
+      "loss": 3.4205,
+      "step": 350
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 4.539979090369383,
+      "learning_rate": 7.971897370051896e-05,
+      "loss": 3.388,
+      "step": 360
+    },
+    {
+      "epoch": 20.555555555555557,
+      "grad_norm": 4.246916156987611,
+      "learning_rate": 7.970057509556505e-05,
+      "loss": 3.3472,
+      "step": 370
+    },
+    {
+      "epoch": 21.11111111111111,
+      "grad_norm": 10.911996284036759,
+      "learning_rate": 7.968159557295457e-05,
+      "loss": 3.3031,
+      "step": 380
+    },
+    {
+      "epoch": 21.666666666666668,
+      "grad_norm": 3.2486673704047813,
+      "learning_rate": 7.966203542906841e-05,
+      "loss": 3.3007,
+      "step": 390
+    },
+    {
+      "epoch": 22.22222222222222,
+      "grad_norm": 4.601663015203652,
+      "learning_rate": 7.964189496935435e-05,
+      "loss": 3.2456,
+      "step": 400
+    },
+    {
+      "epoch": 22.77777777777778,
+      "grad_norm": 3.3251375212947902,
+      "learning_rate": 7.962117450832225e-05,
+      "loss": 3.2387,
+      "step": 410
+    },
+    {
+      "epoch": 23.333333333333332,
+      "grad_norm": 8.001017982564733,
+      "learning_rate": 7.959987436953921e-05,
+      "loss": 3.2016,
+      "step": 420
+    },
+    {
+      "epoch": 23.88888888888889,
+      "grad_norm": 4.35688891329669,
+      "learning_rate": 7.957799488562449e-05,
+      "loss": 3.1819,
+      "step": 430
+    },
+    {
+      "epoch": 24.444444444444443,
+      "grad_norm": 5.039613016500285,
+      "learning_rate": 7.955553639824423e-05,
+      "loss": 3.1292,
+      "step": 440
+    },
+    {
+      "epoch": 25.0,
+      "grad_norm": 5.134919591940875,
+      "learning_rate": 7.953249925810626e-05,
+      "loss": 3.1823,
+      "step": 450
+    },
+    {
+      "epoch": 25.555555555555557,
+      "grad_norm": 7.15150474239977,
+      "learning_rate": 7.950888382495452e-05,
+      "loss": 3.131,
+      "step": 460
+    },
+    {
+      "epoch": 26.11111111111111,
+      "grad_norm": 3.172604932560225,
+      "learning_rate": 7.948469046756344e-05,
+      "loss": 3.1439,
+      "step": 470
+    },
+    {
+      "epoch": 26.666666666666668,
+      "grad_norm": 5.221707539638409,
+      "learning_rate": 7.945991956373229e-05,
+      "loss": 3.1096,
+      "step": 480
+    },
+    {
+      "epoch": 27.22222222222222,
+      "grad_norm": 3.49525470575019,
+      "learning_rate": 7.943457150027913e-05,
+      "loss": 3.0649,
+      "step": 490
+    },
+    {
+      "epoch": 27.77777777777778,
+      "grad_norm": 3.060897563617328,
+      "learning_rate": 7.940864667303489e-05,
+      "loss": 3.0233,
+      "step": 500
+    },
+    {
+      "epoch": 28.333333333333332,
+      "grad_norm": 6.7835354799085215,
+      "learning_rate": 7.938214548683712e-05,
+      "loss": 3.0597,
+      "step": 510
+    },
+    {
+      "epoch": 28.88888888888889,
+      "grad_norm": 4.520408129875502,
+      "learning_rate": 7.935506835552368e-05,
+      "loss": 3.0803,
+      "step": 520
+    },
+    {
+      "epoch": 29.444444444444443,
+      "grad_norm": 5.063817547432621,
+      "learning_rate": 7.932741570192633e-05,
+      "loss": 3.0258,
+      "step": 530
+    },
+    {
+      "epoch": 30.0,
+      "grad_norm": 3.8950867098949282,
+      "learning_rate": 7.929918795786404e-05,
+      "loss": 2.9988,
+      "step": 540
+    },
+    {
+      "epoch": 30.555555555555557,
+      "grad_norm": 7.470058182945552,
+      "learning_rate": 7.927038556413631e-05,
+      "loss": 3.0017,
+      "step": 550
+    },
+    {
+      "epoch": 31.11111111111111,
+      "grad_norm": 3.2096947262786344,
+      "learning_rate": 7.924100897051629e-05,
+      "loss": 3.0032,
+      "step": 560
+    },
+    {
+      "epoch": 31.666666666666668,
+      "grad_norm": 2.956864829778668,
+      "learning_rate": 7.921105863574369e-05,
+      "loss": 2.9584,
+      "step": 570
+    },
+    {
+      "epoch": 32.22222222222222,
+      "grad_norm": 3.7166557505722424,
+      "learning_rate": 7.918053502751772e-05,
+      "loss": 2.9194,
+      "step": 580
+    },
+    {
+      "epoch": 32.77777777777778,
+      "grad_norm": 3.1945993984805763,
+      "learning_rate": 7.914943862248966e-05,
+      "loss": 2.9151,
+      "step": 590
+    },
+    {
+      "epoch": 33.333333333333336,
+      "grad_norm": 4.201284920463043,
+      "learning_rate": 7.911776990625551e-05,
+      "loss": 2.8868,
+      "step": 600
+    },
+    {
+      "epoch": 33.888888888888886,
+      "grad_norm": 2.4970784282291367,
+      "learning_rate": 7.908552937334843e-05,
+      "loss": 2.8751,
+      "step": 610
+    },
+    {
+      "epoch": 34.44444444444444,
+      "grad_norm": 2.6552292601602656,
+      "learning_rate": 7.905271752723088e-05,
+      "loss": 2.8761,
+      "step": 620
+    },
+    {
+      "epoch": 35.0,
+      "grad_norm": 3.3511736450830942,
+      "learning_rate": 7.901933488028692e-05,
+      "loss": 2.8826,
+      "step": 630
+    },
+    {
+      "epoch": 35.55555555555556,
+      "grad_norm": 3.476694580312304,
+      "learning_rate": 7.898538195381413e-05,
+      "loss": 2.8924,
+      "step": 640
+    },
+    {
+      "epoch": 36.111111111111114,
+      "grad_norm": 3.084515130319896,
+      "learning_rate": 7.895085927801541e-05,
+      "loss": 2.8806,
+      "step": 650
+    },
+    {
+      "epoch": 36.666666666666664,
+      "grad_norm": 3.746191888663341,
+      "learning_rate": 7.891576739199084e-05,
+      "loss": 2.8441,
+      "step": 660
+    },
+    {
+      "epoch": 37.22222222222222,
+      "grad_norm": 4.912108210114176,
+      "learning_rate": 7.888010684372915e-05,
+      "loss": 2.8286,
+      "step": 670
+    },
+    {
+      "epoch": 37.77777777777778,
+      "grad_norm": 2.7904075136262896,
+      "learning_rate": 7.884387819009922e-05,
+      "loss": 2.8044,
+      "step": 680
+    },
+    {
+      "epoch": 38.333333333333336,
+      "grad_norm": 2.492065426697779,
+      "learning_rate": 7.880708199684133e-05,
+      "loss": 2.8257,
+      "step": 690
+    },
+    {
+      "epoch": 38.888888888888886,
+      "grad_norm": 3.3996757969837605,
+      "learning_rate": 7.87697188385584e-05,
+      "loss": 2.7869,
+      "step": 700
+    },
+    {
+      "epoch": 39.44444444444444,
+      "grad_norm": 5.06255594858407,
+      "learning_rate": 7.873178929870695e-05,
+      "loss": 2.8123,
+      "step": 710
+    },
+    {
+      "epoch": 40.0,
+      "grad_norm": 2.5537313836119835,
+      "learning_rate": 7.869329396958801e-05,
+      "loss": 2.7796,
+      "step": 720
+    },
+    {
+      "epoch": 40.55555555555556,
+      "grad_norm": 2.5810944073868645,
+      "learning_rate": 7.86542334523379e-05,
+      "loss": 2.7743,
+      "step": 730
+    },
+    {
+      "epoch": 41.111111111111114,
+      "grad_norm": 3.489897179605547,
+      "learning_rate": 7.861460835691881e-05,
+      "loss": 2.7929,
+      "step": 740
+    },
+    {
+      "epoch": 41.666666666666664,
+      "grad_norm": 3.1054123975331267,
+      "learning_rate": 7.857441930210928e-05,
+      "loss": 2.7762,
+      "step": 750
+    },
+    {
+      "epoch": 42.22222222222222,
+      "grad_norm": 2.8147468916559646,
+      "learning_rate": 7.853366691549454e-05,
+      "loss": 2.7158,
+      "step": 760
+    },
+    {
+      "epoch": 42.77777777777778,
+      "grad_norm": 2.604598281696218,
+      "learning_rate": 7.849235183345669e-05,
+      "loss": 2.7439,
+      "step": 770
+    },
+    {
+      "epoch": 43.333333333333336,
+      "grad_norm": 3.1964430173165392,
+      "learning_rate": 7.845047470116485e-05,
+      "loss": 2.7179,
+      "step": 780
+    },
+    {
+      "epoch": 43.888888888888886,
+      "grad_norm": 3.2807474387270146,
+      "learning_rate": 7.840803617256494e-05,
+      "loss": 2.7602,
+      "step": 790
+    },
+    {
+      "epoch": 44.44444444444444,
+      "grad_norm": 2.564829999856057,
+      "learning_rate": 7.836503691036959e-05,
+      "loss": 2.7486,
+      "step": 800
+    },
+    {
+      "epoch": 45.0,
+      "grad_norm": 2.3842738938799792,
+      "learning_rate": 7.832147758604775e-05,
+      "loss": 2.6844,
+      "step": 810
+    },
+    {
+      "epoch": 45.55555555555556,
+      "grad_norm": 2.5957162596956525,
+      "learning_rate": 7.827735887981415e-05,
+      "loss": 2.7037,
+      "step": 820
+    },
+    {
+      "epoch": 46.111111111111114,
+      "grad_norm": 4.366176099132706,
+      "learning_rate": 7.823268148061884e-05,
+      "loss": 2.6937,
+      "step": 830
+    },
+    {
+      "epoch": 46.666666666666664,
+      "grad_norm": 2.542146752047913,
+      "learning_rate": 7.818744608613617e-05,
+      "loss": 2.7072,
+      "step": 840
+    },
+    {
+      "epoch": 47.22222222222222,
+      "grad_norm": 2.282775995587985,
+      "learning_rate": 7.814165340275418e-05,
+      "loss": 2.7012,
+      "step": 850
+    },
+    {
+      "epoch": 47.77777777777778,
+      "grad_norm": 2.4972644144165983,
+      "learning_rate": 7.809530414556335e-05,
+      "loss": 2.6787,
+      "step": 860
+    },
+    {
+      "epoch": 48.333333333333336,
+      "grad_norm": 3.6324282863440165,
+      "learning_rate": 7.804839903834557e-05,
+      "loss": 2.6731,
+      "step": 870
+    },
+    {
+      "epoch": 48.888888888888886,
+      "grad_norm": 3.7509619432724093,
+      "learning_rate": 7.800093881356278e-05,
+      "loss": 2.6783,
+      "step": 880
+    },
+    {
+      "epoch": 49.44444444444444,
+      "grad_norm": 2.966452261502853,
+      "learning_rate": 7.79529242123455e-05,
+      "loss": 2.729,
+      "step": 890
+    },
+    {
+      "epoch": 50.0,
+      "grad_norm": 2.359789073052784,
+      "learning_rate": 7.790435598448135e-05,
+      "loss": 2.6445,
+      "step": 900
+    },
+    {
+      "epoch": 50.55555555555556,
+      "grad_norm": 2.769470472435637,
+      "learning_rate": 7.785523488840324e-05,
+      "loss": 2.6502,
+      "step": 910
+    },
+    {
+      "epoch": 51.111111111111114,
+      "grad_norm": 2.5216131554182954,
+      "learning_rate": 7.780556169117757e-05,
+      "loss": 2.6239,
+      "step": 920
+    },
+    {
+      "epoch": 51.666666666666664,
+      "grad_norm": 5.099804421956975,
+      "learning_rate": 7.775533716849233e-05,
+      "loss": 2.6084,
+      "step": 930
+    },
+    {
+      "epoch": 52.22222222222222,
+      "grad_norm": 2.2151324371631516,
+      "learning_rate": 7.770456210464481e-05,
+      "loss": 2.5866,
+      "step": 940
+    },
+    {
+      "epoch": 52.77777777777778,
+      "grad_norm": 2.2846921270128226,
+      "learning_rate": 7.765323729252954e-05,
+      "loss": 2.616,
+      "step": 950
+    },
+    {
+      "epoch": 53.333333333333336,
+      "grad_norm": 2.5129423824312873,
+      "learning_rate": 7.760136353362576e-05,
+      "loss": 2.5647,
+      "step": 960
+    },
+    {
+      "epoch": 53.888888888888886,
+      "grad_norm": 2.5058947684805606,
+      "learning_rate": 7.7548941637985e-05,
+      "loss": 2.6088,
+      "step": 970
+    },
+    {
+      "epoch": 54.44444444444444,
+      "grad_norm": 2.3357415260603003,
+      "learning_rate": 7.749597242421838e-05,
+      "loss": 2.5933,
+      "step": 980
+    },
+    {
+      "epoch": 55.0,
+      "grad_norm": 2.2717760444928956,
+      "learning_rate": 7.744245671948386e-05,
+      "loss": 2.5602,
+      "step": 990
+    },
+    {
+      "epoch": 55.55555555555556,
+      "grad_norm": 2.393666047299851,
+      "learning_rate": 7.738839535947331e-05,
+      "loss": 2.6021,
+      "step": 1000
+    },
+    {
+      "epoch": 56.111111111111114,
+      "grad_norm": 3.6219481413913455,
+      "learning_rate": 7.733378918839942e-05,
+      "loss": 2.5959,
+      "step": 1010
+    },
+    {
+      "epoch": 56.666666666666664,
+      "grad_norm": 2.285785086420793,
+      "learning_rate": 7.727863905898261e-05,
+      "loss": 2.5468,
+      "step": 1020
+    },
+    {
+      "epoch": 57.22222222222222,
+      "grad_norm": 2.3177908093354422,
+      "learning_rate": 7.72229458324376e-05,
+      "loss": 2.5997,
+      "step": 1030
+    },
+    {
+      "epoch": 57.77777777777778,
+      "grad_norm": 2.366559741689681,
+      "learning_rate": 7.716671037846008e-05,
+      "loss": 2.5609,
+      "step": 1040
+    },
+    {
+      "epoch": 58.333333333333336,
+      "grad_norm": 2.42129942914918,
+      "learning_rate": 7.710993357521301e-05,
+      "loss": 2.5348,
+      "step": 1050
+    },
+    {
+      "epoch": 58.888888888888886,
+      "grad_norm": 2.814708351353641,
+      "learning_rate": 7.705261630931301e-05,
+      "loss": 2.548,
+      "step": 1060
+    },
+    {
+      "epoch": 59.44444444444444,
+      "grad_norm": 2.286890346812349,
+      "learning_rate": 7.699475947581645e-05,
+      "loss": 2.5384,
+      "step": 1070
+    },
+    {
+      "epoch": 60.0,
+      "grad_norm": 2.80847498784567,
+      "learning_rate": 7.693636397820548e-05,
+      "loss": 2.4881,
+      "step": 1080
+    },
+    {
+      "epoch": 60.55555555555556,
+      "grad_norm": 2.300196315845816,
+      "learning_rate": 7.687743072837396e-05,
+      "loss": 2.4919,
+      "step": 1090
+    },
+    {
+      "epoch": 61.111111111111114,
+      "grad_norm": 2.4009666403566436,
+      "learning_rate": 7.68179606466132e-05,
+      "loss": 2.5098,
+      "step": 1100
+    },
+    {
+      "epoch": 61.666666666666664,
+      "grad_norm": 9.48305687079573,
+      "learning_rate": 7.675795466159753e-05,
+      "loss": 2.53,
+      "step": 1110
+    },
+    {
+      "epoch": 62.22222222222222,
+      "grad_norm": 2.596093187954865,
+      "learning_rate": 7.669741371036993e-05,
+      "loss": 2.5293,
+      "step": 1120
+    },
+    {
+      "epoch": 62.77777777777778,
+      "grad_norm": 2.147522221584612,
+      "learning_rate": 7.663633873832725e-05,
+      "loss": 2.5022,
+      "step": 1130
+    },
+    {
+      "epoch": 63.333333333333336,
+      "grad_norm": 2.202611552259382,
+      "learning_rate": 7.657473069920554e-05,
+      "loss": 2.4937,
+      "step": 1140
+    },
+    {
+      "epoch": 63.888888888888886,
+      "grad_norm": 2.1881040556536426,
+      "learning_rate": 7.651259055506515e-05,
+      "loss": 2.4747,
+      "step": 1150
+    },
+    {
+      "epoch": 64.44444444444444,
+      "grad_norm": 2.2958507719491106,
+      "learning_rate": 7.644991927627567e-05,
+      "loss": 2.4539,
+      "step": 1160
+    },
+    {
+      "epoch": 65.0,
+      "grad_norm": 2.282416267655807,
+      "learning_rate": 7.638671784150078e-05,
+      "loss": 2.4461,
+      "step": 1170
+    },
+    {
+      "epoch": 65.55555555555556,
+      "grad_norm": 2.2613733848197843,
+      "learning_rate": 7.632298723768303e-05,
+      "loss": 2.4803,
+      "step": 1180
+    },
+    {
+      "epoch": 66.11111111111111,
+      "grad_norm": 2.2081444017695406,
+      "learning_rate": 7.625872846002834e-05,
+      "loss": 2.4678,
+      "step": 1190
+    },
+    {
+      "epoch": 66.66666666666667,
+      "grad_norm": 2.219065361714099,
+      "learning_rate": 7.61939425119905e-05,
+      "loss": 2.4499,
+      "step": 1200
+    },
+    {
+      "epoch": 67.22222222222223,
+      "grad_norm": 2.3336979263290565,
+      "learning_rate": 7.612863040525552e-05,
+      "loss": 2.4388,
+      "step": 1210
+    },
+    {
+      "epoch": 67.77777777777777,
+      "grad_norm": 2.0148806121518925,
+      "learning_rate": 7.606279315972582e-05,
+      "loss": 2.4442,
+      "step": 1220
+    },
+    {
+      "epoch": 68.33333333333333,
+      "grad_norm": 2.3872289533695055,
+      "learning_rate": 7.599643180350426e-05,
+      "loss": 2.4436,
+      "step": 1230
+    },
+    {
+      "epoch": 68.88888888888889,
+      "grad_norm": 2.132644702612647,
+      "learning_rate": 7.592954737287815e-05,
+      "loss": 2.4234,
+      "step": 1240
+    },
+    {
+      "epoch": 69.44444444444444,
+      "grad_norm": 2.272555045624701,
+      "learning_rate": 7.586214091230304e-05,
+      "loss": 2.4163,
+      "step": 1250
+    },
+    {
+      "epoch": 70.0,
+      "grad_norm": 2.555055737568658,
+      "learning_rate": 7.579421347438636e-05,
+      "loss": 2.4144,
+      "step": 1260
+    },
+    {
+      "epoch": 70.55555555555556,
+      "grad_norm": 2.40263650242411,
+      "learning_rate": 7.57257661198711e-05,
+      "loss": 2.4168,
+      "step": 1270
+    },
+    {
+      "epoch": 71.11111111111111,
+      "grad_norm": 2.3251411723274873,
+      "learning_rate": 7.565679991761914e-05,
+      "loss": 2.4238,
+      "step": 1280
+    },
+    {
+      "epoch": 71.66666666666667,
+      "grad_norm": 2.2359134739425324,
+      "learning_rate": 7.558731594459459e-05,
+      "loss": 2.3987,
+      "step": 1290
+    },
+    {
+      "epoch": 72.22222222222223,
+      "grad_norm": 2.140451800168365,
+      "learning_rate": 7.551731528584699e-05,
+      "loss": 2.3918,
+      "step": 1300
+    },
+    {
+      "epoch": 72.77777777777777,
+      "grad_norm": 2.1380915442299395,
+      "learning_rate": 7.544679903449437e-05,
+      "loss": 2.3695,
+      "step": 1310
+    },
+    {
+      "epoch": 73.33333333333333,
+      "grad_norm": 2.4343385980159042,
+      "learning_rate": 7.537576829170611e-05,
+      "loss": 2.3496,
+      "step": 1320
+    },
+    {
+      "epoch": 73.88888888888889,
+      "grad_norm": 2.23364960959876,
+      "learning_rate": 7.530422416668585e-05,
+      "loss": 2.3591,
+      "step": 1330
+    },
+    {
+      "epoch": 74.44444444444444,
+      "grad_norm": 2.8257157828351684,
+      "learning_rate": 7.52321677766541e-05,
+      "loss": 2.332,
+      "step": 1340
+    },
+    {
+      "epoch": 75.0,
+      "grad_norm": 2.4267027942706934,
+      "learning_rate": 7.515960024683081e-05,
+      "loss": 2.3507,
+      "step": 1350
+    },
+    {
+      "epoch": 75.55555555555556,
+      "grad_norm": 2.1994630938722644,
+      "learning_rate": 7.508652271041779e-05,
+      "loss": 2.3203,
+      "step": 1360
+    },
+    {
+      "epoch": 76.11111111111111,
+      "grad_norm": 2.237729096005724,
+      "learning_rate": 7.501293630858103e-05,
+      "loss": 2.3199,
+      "step": 1370
+    },
+    {
+      "epoch": 76.66666666666667,
+      "grad_norm": 2.118517581041144,
+      "learning_rate": 7.49388421904329e-05,
+      "loss": 2.3428,
+      "step": 1380
+    },
+    {
+      "epoch": 77.22222222222223,
+      "grad_norm": 2.19508670175106,
+      "learning_rate": 7.486424151301413e-05,
+      "loss": 2.3104,
+      "step": 1390
+    },
+    {
+      "epoch": 77.77777777777777,
+      "grad_norm": 2.269786091611358,
+      "learning_rate": 7.478913544127583e-05,
+      "loss": 2.3584,
+      "step": 1400
+    },
+    {
+      "epoch": 78.33333333333333,
+      "grad_norm": 2.371950601169516,
+      "learning_rate": 7.471352514806125e-05,
+      "loss": 2.2984,
+      "step": 1410
+    },
+    {
+      "epoch": 78.88888888888889,
+      "grad_norm": 2.187016133880758,
+      "learning_rate": 7.463741181408744e-05,
+      "loss": 2.3186,
+      "step": 1420
+    },
+    {
+      "epoch": 79.44444444444444,
+      "grad_norm": 2.480456831295184,
+      "learning_rate": 7.45607966279269e-05,
+      "loss": 2.3182,
+      "step": 1430
+    },
+    {
+      "epoch": 80.0,
+      "grad_norm": 2.077693578925958,
+      "learning_rate": 7.448368078598894e-05,
+      "loss": 2.2832,
+      "step": 1440
+    },
+    {
+      "epoch": 80.55555555555556,
+      "grad_norm": 3.0301574625829057,
+      "learning_rate": 7.440606549250101e-05,
+      "loss": 2.2995,
+      "step": 1450
+    },
+    {
+      "epoch": 81.11111111111111,
+      "grad_norm": 2.481371137222743,
+      "learning_rate": 7.432795195948994e-05,
+      "loss": 2.2786,
+      "step": 1460
+    },
+    {
+      "epoch": 81.66666666666667,
+      "grad_norm": 2.701438156981817,
+      "learning_rate": 7.424934140676293e-05,
+      "loss": 2.2892,
+      "step": 1470
+    },
+    {
+      "epoch": 82.22222222222223,
+      "grad_norm": 2.312641036398088,
+      "learning_rate": 7.41702350618886e-05,
+      "loss": 2.2504,
+      "step": 1480
+    },
+    {
+      "epoch": 82.77777777777777,
+      "grad_norm": 2.3523078827038177,
+      "learning_rate": 7.409063416017778e-05,
+      "loss": 2.2475,
+      "step": 1490
+    },
+    {
+      "epoch": 83.33333333333333,
+      "grad_norm": 2.648877402712022,
+      "learning_rate": 7.401053994466417e-05,
+      "loss": 2.2604,
+      "step": 1500
+    },
+    {
+      "epoch": 83.88888888888889,
+      "grad_norm": 2.280831912672915,
+      "learning_rate": 7.392995366608499e-05,
+      "loss": 2.2395,
+      "step": 1510
+    },
+    {
+      "epoch": 84.44444444444444,
+      "grad_norm": 2.3086628395237576,
+      "learning_rate": 7.384887658286145e-05,
+      "loss": 2.2243,
+      "step": 1520
+    },
+    {
+      "epoch": 85.0,
+      "grad_norm": 2.201845703969594,
+      "learning_rate": 7.376730996107905e-05,
+      "loss": 2.2136,
+      "step": 1530
+    },
+    {
+      "epoch": 85.55555555555556,
+      "grad_norm": 2.281121759860022,
+      "learning_rate": 7.368525507446785e-05,
+      "loss": 2.2332,
+      "step": 1540
+    },
+    {
+      "epoch": 86.11111111111111,
+      "grad_norm": 2.1372221470542963,
+      "learning_rate": 7.360271320438257e-05,
+      "loss": 2.2096,
+      "step": 1550
+    },
+    {
+      "epoch": 86.66666666666667,
+      "grad_norm": 2.3214395271294266,
+      "learning_rate": 7.351968563978259e-05,
+      "loss": 2.1873,
+      "step": 1560
+    },
+    {
+      "epoch": 87.22222222222223,
+      "grad_norm": 2.211596387404363,
+      "learning_rate": 7.343617367721179e-05,
+      "loss": 2.2148,
+      "step": 1570
+    },
+    {
+      "epoch": 87.77777777777777,
+      "grad_norm": 2.2170170208351148,
+      "learning_rate": 7.335217862077831e-05,
+      "loss": 2.1891,
+      "step": 1580
+    },
+    {
+      "epoch": 88.33333333333333,
+      "grad_norm": 2.2711732475820128,
+      "learning_rate": 7.326770178213425e-05,
+      "loss": 2.1803,
+      "step": 1590
+    },
+    {
+      "epoch": 88.88888888888889,
+      "grad_norm": 2.4076774623663497,
+      "learning_rate": 7.31827444804551e-05,
+      "loss": 2.1614,
+      "step": 1600
+    },
+    {
+      "epoch": 89.44444444444444,
+      "grad_norm": 2.333760960312592,
+      "learning_rate": 7.309730804241918e-05,
+      "loss": 2.162,
+      "step": 1610
+    },
+    {
+      "epoch": 90.0,
+      "grad_norm": 2.2616895499181005,
+      "learning_rate": 7.301139380218691e-05,
+      "loss": 2.1571,
+      "step": 1620
+    },
+    {
+      "epoch": 90.55555555555556,
+      "grad_norm": 2.291582094423719,
+      "learning_rate": 7.292500310138003e-05,
+      "loss": 2.1331,
+      "step": 1630
+    },
+    {
+      "epoch": 91.11111111111111,
+      "grad_norm": 2.559504549243399,
+      "learning_rate": 7.283813728906054e-05,
+      "loss": 2.1497,
+      "step": 1640
+    },
+    {
+      "epoch": 91.66666666666667,
+      "grad_norm": 2.4641897842342844,
+      "learning_rate": 7.275079772170975e-05,
+      "loss": 2.1468,
+      "step": 1650
+    },
+    {
+      "epoch": 92.22222222222223,
+      "grad_norm": 2.332721346117361,
+      "learning_rate": 7.266298576320701e-05,
+      "loss": 2.1371,
+      "step": 1660
+    },
+    {
+      "epoch": 92.77777777777777,
+      "grad_norm": 2.332516004828426,
+      "learning_rate": 7.257470278480848e-05,
+      "loss": 2.1031,
+      "step": 1670
+    },
+    {
+      "epoch": 93.33333333333333,
+      "grad_norm": 2.2619055372729204,
+      "learning_rate": 7.248595016512566e-05,
+      "loss": 2.0881,
+      "step": 1680
+    },
+    {
+      "epoch": 93.88888888888889,
+      "grad_norm": 2.303038961182827,
+      "learning_rate": 7.239672929010391e-05,
+      "loss": 2.073,
+      "step": 1690
+    },
+    {
+      "epoch": 94.44444444444444,
+      "grad_norm": 2.275538152406292,
+      "learning_rate": 7.230704155300076e-05,
+      "loss": 2.0893,
+      "step": 1700
+    },
+    {
+      "epoch": 95.0,
+      "grad_norm": 2.79339024491097,
+      "learning_rate": 7.221688835436417e-05,
+      "loss": 2.0862,
+      "step": 1710
+    },
+    {
+      "epoch": 95.55555555555556,
+      "grad_norm": 2.357567808727419,
+      "learning_rate": 7.212627110201067e-05,
+      "loss": 2.0925,
+      "step": 1720
+    },
+    {
+      "epoch": 96.11111111111111,
+      "grad_norm": 2.438625955951333,
+      "learning_rate": 7.203519121100341e-05,
+      "loss": 2.0833,
+      "step": 1730
+    },
+    {
+      "epoch": 96.66666666666667,
+      "grad_norm": 2.268911884511815,
+      "learning_rate": 7.194365010362997e-05,
+      "loss": 2.072,
+      "step": 1740
+    },
+    {
+      "epoch": 97.22222222222223,
+      "grad_norm": 2.4071647589682863,
+      "learning_rate": 7.185164920938025e-05,
+      "loss": 2.0614,
+      "step": 1750
+    },
+    {
+      "epoch": 97.77777777777777,
+      "grad_norm": 2.4003159315067095,
+      "learning_rate": 7.175918996492408e-05,
+      "loss": 2.0327,
+      "step": 1760
+    },
+    {
+      "epoch": 98.33333333333333,
+      "grad_norm": 2.9641989105961817,
+      "learning_rate": 7.166627381408878e-05,
+      "loss": 2.029,
+      "step": 1770
+    },
+    {
+      "epoch": 98.88888888888889,
+      "grad_norm": 2.33150487738845,
+      "learning_rate": 7.157290220783669e-05,
+      "loss": 2.065,
+      "step": 1780
+    },
+    {
+      "epoch": 99.44444444444444,
+      "grad_norm": 2.389128272558943,
+      "learning_rate": 7.147907660424242e-05,
+      "loss": 2.0492,
+      "step": 1790
+    },
+    {
+      "epoch": 100.0,
+      "grad_norm": 2.340393307554397,
+      "learning_rate": 7.138479846847017e-05,
+      "loss": 2.0283,
+      "step": 1800
+    },
+    {
+      "epoch": 100.55555555555556,
+      "grad_norm": 2.41990473189297,
+      "learning_rate": 7.129006927275073e-05,
+      "loss": 2.0122,
+      "step": 1810
+    },
+    {
+      "epoch": 101.11111111111111,
+      "grad_norm": 2.3030306792905337,
+      "learning_rate": 7.119489049635866e-05,
+      "loss": 2.0249,
+      "step": 1820
+    },
+    {
+      "epoch": 101.66666666666667,
+      "grad_norm": 2.393855287313057,
+      "learning_rate": 7.109926362558899e-05,
+      "loss": 2.0025,
+      "step": 1830
+    },
+    {
+      "epoch": 102.22222222222223,
+      "grad_norm": 2.291151428100202,
+      "learning_rate": 7.10031901537342e-05,
+      "loss": 2.007,
+      "step": 1840
+    },
+    {
+      "epoch": 102.77777777777777,
+      "grad_norm": 2.3479514547330314,
+      "learning_rate": 7.090667158106076e-05,
+      "loss": 1.9633,
+      "step": 1850
+    },
+    {
+      "epoch": 103.33333333333333,
+      "grad_norm": 2.3204828578691443,
+      "learning_rate": 7.080970941478578e-05,
+      "loss": 1.9787,
+      "step": 1860
+    },
+    {
+      "epoch": 103.88888888888889,
+      "grad_norm": 2.250112212879735,
+      "learning_rate": 7.071230516905344e-05,
+      "loss": 1.9758,
+      "step": 1870
+    },
+    {
+      "epoch": 104.44444444444444,
+      "grad_norm": 2.1686786211123863,
+      "learning_rate": 7.061446036491134e-05,
+      "loss": 1.9518,
+      "step": 1880
+    },
+    {
+      "epoch": 105.0,
+      "grad_norm": 2.3881235436857127,
+      "learning_rate": 7.051617653028682e-05,
+      "loss": 1.9398,
+      "step": 1890
+    },
+    {
+      "epoch": 105.55555555555556,
+      "grad_norm": 2.6739034704790106,
+      "learning_rate": 7.041745519996298e-05,
+      "loss": 1.9224,
+      "step": 1900
+    },
+    {
+      "epoch": 106.11111111111111,
+      "grad_norm": 2.280685877616025,
+      "learning_rate": 7.031829791555481e-05,
+      "loss": 1.9453,
+      "step": 1910
+    },
+    {
+      "epoch": 106.66666666666667,
+      "grad_norm": 2.3490636989422473,
+      "learning_rate": 7.021870622548507e-05,
+      "loss": 1.8938,
+      "step": 1920
+    },
+    {
+      "epoch": 107.22222222222223,
+      "grad_norm": 2.4469573609971396,
+      "learning_rate": 7.011868168496013e-05,
+      "loss": 1.902,
+      "step": 1930
+    },
+    {
+      "epoch": 107.77777777777777,
+      "grad_norm": 2.3674563012929077,
+      "learning_rate": 7.001822585594565e-05,
+      "loss": 1.8738,
+      "step": 1940
+    },
+    {
+      "epoch": 108.33333333333333,
+      "grad_norm": 2.3803844651066752,
+      "learning_rate": 6.991734030714228e-05,
+      "loss": 1.889,
+      "step": 1950
+    },
+    {
+      "epoch": 108.88888888888889,
+      "grad_norm": 2.3923680551911373,
+      "learning_rate": 6.981602661396103e-05,
+      "loss": 1.873,
+      "step": 1960
+    },
+    {
+      "epoch": 109.44444444444444,
+      "grad_norm": 2.607392065899362,
+      "learning_rate": 6.971428635849877e-05,
+      "loss": 1.8968,
+      "step": 1970
+    },
+    {
+      "epoch": 110.0,
+      "grad_norm": 2.3080207126421874,
+      "learning_rate": 6.961212112951348e-05,
+      "loss": 1.8771,
+      "step": 1980
+    },
+    {
+      "epoch": 110.55555555555556,
+      "grad_norm": 2.3861143155336113,
+      "learning_rate": 6.950953252239947e-05,
+      "loss": 1.8945,
+      "step": 1990
+    },
+    {
+      "epoch": 111.11111111111111,
+      "grad_norm": 2.3536956338382558,
+      "learning_rate": 6.940652213916241e-05,
+      "loss": 1.8398,
+      "step": 2000
+    },
+    {
+      "epoch": 111.66666666666667,
+      "grad_norm": 2.3941386213650553,
+      "learning_rate": 6.930309158839443e-05,
+      "loss": 1.8181,
+      "step": 2010
+    },
+    {
+      "epoch": 112.22222222222223,
+      "grad_norm": 2.3682854753026072,
+      "learning_rate": 6.919924248524884e-05,
+      "loss": 1.828,
+      "step": 2020
+    },
+    {
+      "epoch": 112.77777777777777,
+      "grad_norm": 2.2868577149252176,
+      "learning_rate": 6.9094976451415e-05,
+      "loss": 1.8397,
+      "step": 2030
+    },
+    {
+      "epoch": 113.33333333333333,
+      "grad_norm": 2.322921981007152,
+      "learning_rate": 6.899029511509305e-05,
+      "loss": 1.8396,
+      "step": 2040
+    },
+    {
+      "epoch": 113.88888888888889,
+      "grad_norm": 2.40320612617936,
+      "learning_rate": 6.888520011096835e-05,
+      "loss": 1.8116,
+      "step": 2050
+    },
+    {
+      "epoch": 114.44444444444444,
+      "grad_norm": 2.352330687507841,
+      "learning_rate": 6.877969308018608e-05,
+      "loss": 1.8193,
+      "step": 2060
+    },
+    {
+      "epoch": 115.0,
+      "grad_norm": 2.330749466166547,
+      "learning_rate": 6.867377567032551e-05,
+      "loss": 1.7992,
+      "step": 2070
+    },
+    {
+      "epoch": 115.55555555555556,
+      "grad_norm": 2.469285061954898,
+      "learning_rate": 6.856744953537433e-05,
+      "loss": 1.7758,
+      "step": 2080
+    },
+    {
+      "epoch": 116.11111111111111,
+      "grad_norm": 2.4215776814728094,
+      "learning_rate": 6.846071633570285e-05,
+      "loss": 1.7869,
+      "step": 2090
+    },
+    {
+      "epoch": 116.66666666666667,
+      "grad_norm": 3.052624252002459,
+      "learning_rate": 6.835357773803795e-05,
+      "loss": 1.7792,
+      "step": 2100
+    },
+    {
+      "epoch": 117.22222222222223,
+      "grad_norm": 2.3441541196195996,
+      "learning_rate": 6.824603541543722e-05,
+      "loss": 1.786,
+      "step": 2110
+    },
+    {
+      "epoch": 117.77777777777777,
+      "grad_norm": 2.425047046165691,
+      "learning_rate": 6.81380910472627e-05,
+      "loss": 1.7672,
+      "step": 2120
+    },
+    {
+      "epoch": 118.33333333333333,
+      "grad_norm": 2.302107475695016,
+      "learning_rate": 6.802974631915469e-05,
+      "loss": 1.7274,
+      "step": 2130
+    },
+    {
+      "epoch": 118.88888888888889,
+      "grad_norm": 2.552127033300655,
+      "learning_rate": 6.79210029230055e-05,
+      "loss": 1.7492,
+      "step": 2140
+    },
+    {
+      "epoch": 119.44444444444444,
+      "grad_norm": 2.3531536409577853,
+      "learning_rate": 6.781186255693291e-05,
+      "loss": 1.721,
+      "step": 2150
+    },
+    {
+      "epoch": 120.0,
+      "grad_norm": 2.348933883054798,
+      "learning_rate": 6.770232692525372e-05,
+      "loss": 1.7203,
+      "step": 2160
+    },
+    {
+      "epoch": 120.55555555555556,
+      "grad_norm": 2.4017389792313257,
+      "learning_rate": 6.759239773845719e-05,
+      "loss": 1.7064,
+      "step": 2170
+    },
+    {
+      "epoch": 121.11111111111111,
+      "grad_norm": 2.325792411390438,
+      "learning_rate": 6.74820767131782e-05,
+      "loss": 1.7103,
+      "step": 2180
+    },
+    {
+      "epoch": 121.66666666666667,
+      "grad_norm": 2.6218349812923805,
+      "learning_rate": 6.73713655721705e-05,
+      "loss": 1.7316,
+      "step": 2190
+    },
+    {
+      "epoch": 122.22222222222223,
+      "grad_norm": 2.3036782333943986,
+      "learning_rate": 6.726026604427991e-05,
+      "loss": 1.6787,
+      "step": 2200
+    },
+    {
+      "epoch": 122.77777777777777,
+      "grad_norm": 2.329702572482786,
+      "learning_rate": 6.714877986441712e-05,
+      "loss": 1.6726,
+      "step": 2210
+    },
+    {
+      "epoch": 123.33333333333333,
+      "grad_norm": 2.5525115192388785,
+      "learning_rate": 6.703690877353082e-05,
+      "loss": 1.6443,
+      "step": 2220
+    },
+    {
+      "epoch": 123.88888888888889,
+      "grad_norm": 2.2951760788400613,
+      "learning_rate": 6.69246545185803e-05,
+      "loss": 1.6462,
+      "step": 2230
+    },
+    {
+      "epoch": 124.44444444444444,
+      "grad_norm": 2.4461983243214034,
+      "learning_rate": 6.681201885250831e-05,
+      "loss": 1.6517,
+      "step": 2240
+    },
+    {
+      "epoch": 125.0,
+      "grad_norm": 2.2585277018593146,
+      "learning_rate": 6.66990035342137e-05,
+      "loss": 1.6271,
+      "step": 2250
+    },
+    {
+      "epoch": 125.55555555555556,
+      "grad_norm": 2.382648299766384,
+      "learning_rate": 6.658561032852383e-05,
+      "loss": 1.6182,
+      "step": 2260
+    },
+    {
+      "epoch": 126.11111111111111,
+      "grad_norm": 2.563355186858692,
+      "learning_rate": 6.647184100616713e-05,
+      "loss": 1.6122,
+      "step": 2270
+    },
+    {
+      "epoch": 126.66666666666667,
+      "grad_norm": 2.302901684973611,
+      "learning_rate": 6.635769734374537e-05,
+      "loss": 1.5887,
+      "step": 2280
+    },
+    {
+      "epoch": 127.22222222222223,
+      "grad_norm": 2.435675011195532,
+      "learning_rate": 6.6243181123706e-05,
+      "loss": 1.6541,
+      "step": 2290
+    },
+    {
+      "epoch": 127.77777777777777,
+      "grad_norm": 2.346454230124902,
+      "learning_rate": 6.612829413431417e-05,
+      "loss": 1.5832,
+      "step": 2300
+    },
+    {
+      "epoch": 128.33333333333334,
+      "grad_norm": 2.29318355873345,
+      "learning_rate": 6.6013038169625e-05,
+      "loss": 1.5964,
+      "step": 2310
+    },
+    {
+      "epoch": 128.88888888888889,
+      "grad_norm": 2.3274624284934404,
+      "learning_rate": 6.58974150294554e-05,
+      "loss": 1.5613,
+      "step": 2320
+    },
+    {
+      "epoch": 129.44444444444446,
+      "grad_norm": 2.52388775945014,
+      "learning_rate": 6.578142651935609e-05,
+      "loss": 1.5866,
+      "step": 2330
+    },
+    {
+      "epoch": 130.0,
+      "grad_norm": 2.4785625672761107,
+      "learning_rate": 6.566507445058325e-05,
+      "loss": 1.5655,
+      "step": 2340
+    },
+    {
+      "epoch": 130.55555555555554,
+      "grad_norm": 2.4038863029061512,
+      "learning_rate": 6.554836064007041e-05,
+      "loss": 1.5673,
+      "step": 2350
+    },
+    {
+      "epoch": 131.11111111111111,
+      "grad_norm": 2.3402915360536527,
+      "learning_rate": 6.543128691039996e-05,
+      "loss": 1.5908,
+      "step": 2360
+    },
+    {
+      "epoch": 131.66666666666666,
+      "grad_norm": 2.408138079356503,
+      "learning_rate": 6.531385508977472e-05,
+      "loss": 1.5588,
+      "step": 2370
+    },
+    {
+      "epoch": 132.22222222222223,
+      "grad_norm": 2.5211155841119517,
+      "learning_rate": 6.519606701198941e-05,
+      "loss": 1.574,
+      "step": 2380
+    },
+    {
+      "epoch": 132.77777777777777,
+      "grad_norm": 2.438480962307054,
+      "learning_rate": 6.507792451640202e-05,
+      "loss": 1.5404,
+      "step": 2390
+    },
+    {
+      "epoch": 133.33333333333334,
+      "grad_norm": 2.438770061119076,
+      "learning_rate": 6.495942944790501e-05,
+      "loss": 1.5225,
+      "step": 2400
+    },
+    {
+      "epoch": 133.88888888888889,
+      "grad_norm": 2.3877662070891272,
+      "learning_rate": 6.48405836568966e-05,
+      "loss": 1.5313,
+      "step": 2410
+    },
+    {
+      "epoch": 134.44444444444446,
+      "grad_norm": 2.3033377105332704,
+      "learning_rate": 6.472138899925184e-05,
+      "loss": 1.5241,
+      "step": 2420
+    },
+    {
+      "epoch": 135.0,
+      "grad_norm": 2.435449667109581,
+      "learning_rate": 6.460184733629356e-05,
+      "loss": 1.5235,
+      "step": 2430
+    },
+    {
+      "epoch": 135.55555555555554,
+      "grad_norm": 2.9958731558907035,
+      "learning_rate": 6.448196053476343e-05,
+      "loss": 1.492,
+      "step": 2440
+    },
+    {
+      "epoch": 136.11111111111111,
+      "grad_norm": 2.4029638467038312,
+      "learning_rate": 6.436173046679271e-05,
+      "loss": 1.4927,
+      "step": 2450
+    },
+    {
+      "epoch": 136.66666666666666,
+      "grad_norm": 2.4764672871767934,
+      "learning_rate": 6.424115900987306e-05,
+      "loss": 1.4987,
+      "step": 2460
+    },
+    {
+      "epoch": 137.22222222222223,
+      "grad_norm": 2.336475221116805,
+      "learning_rate": 6.41202480468272e-05,
+      "loss": 1.4737,
+      "step": 2470
+    },
+    {
+      "epoch": 137.77777777777777,
+      "grad_norm": 2.4602886979569942,
+      "learning_rate": 6.399899946577953e-05,
+      "loss": 1.4591,
+      "step": 2480
+    },
+    {
+      "epoch": 138.33333333333334,
+      "grad_norm": 2.4310178395805,
+      "learning_rate": 6.387741516012663e-05,
+      "loss": 1.4269,
+      "step": 2490
+    },
+    {
+      "epoch": 138.88888888888889,
+      "grad_norm": 2.3304162751344184,
+      "learning_rate": 6.375549702850768e-05,
+      "loss": 1.4381,
+      "step": 2500
+    },
+    {
+      "epoch": 139.44444444444446,
+      "grad_norm": 2.366762028595186,
+      "learning_rate": 6.363324697477485e-05,
+      "loss": 1.4098,
+      "step": 2510
+    },
+    {
+      "epoch": 140.0,
+      "grad_norm": 2.3863034550398052,
+      "learning_rate": 6.351066690796354e-05,
+      "loss": 1.4506,
+      "step": 2520
+    },
+    {
+      "epoch": 140.55555555555554,
+      "grad_norm": 2.3935560820196877,
+      "learning_rate": 6.338775874226258e-05,
+      "loss": 1.4233,
+      "step": 2530
+    },
+    {
+      "epoch": 141.11111111111111,
+      "grad_norm": 2.2746052577698657,
+      "learning_rate": 6.326452439698431e-05,
+      "loss": 1.4098,
+      "step": 2540
+    },
+    {
+      "epoch": 141.66666666666666,
+      "grad_norm": 2.3369742551927253,
+      "learning_rate": 6.314096579653469e-05,
+      "loss": 1.423,
+      "step": 2550
+    },
+    {
+      "epoch": 142.22222222222223,
+      "grad_norm": 2.3797598874760024,
+      "learning_rate": 6.301708487038312e-05,
+      "loss": 1.4193,
+      "step": 2560
+    },
+    {
+      "epoch": 142.77777777777777,
+      "grad_norm": 2.7107745294753123,
+      "learning_rate": 6.289288355303245e-05,
+      "loss": 1.4147,
+      "step": 2570
+    },
+    {
+      "epoch": 143.33333333333334,
+      "grad_norm": 2.2650678278385996,
+      "learning_rate": 6.276836378398864e-05,
+      "loss": 1.4249,
+      "step": 2580
+    },
+    {
+      "epoch": 143.88888888888889,
+      "grad_norm": 2.336087531435586,
+      "learning_rate": 6.264352750773059e-05,
+      "loss": 1.3956,
+      "step": 2590
+    },
+    {
+      "epoch": 144.44444444444446,
+      "grad_norm": 2.4428490853714533,
+      "learning_rate": 6.251837667367967e-05,
+      "loss": 1.3683,
+      "step": 2600
+    },
+    {
+      "epoch": 145.0,
+      "grad_norm": 2.204887929304398,
+      "learning_rate": 6.239291323616937e-05,
+      "loss": 1.3842,
+      "step": 2610
+    },
+    {
+      "epoch": 145.55555555555554,
+      "grad_norm": 2.33449609485789,
+      "learning_rate": 6.226713915441472e-05,
+      "loss": 1.3738,
+      "step": 2620
+    },
+    {
+      "epoch": 146.11111111111111,
+      "grad_norm": 2.3515529695742297,
+      "learning_rate": 6.214105639248174e-05,
+      "loss": 1.3343,
+      "step": 2630
+    },
+    {
+      "epoch": 146.66666666666666,
+      "grad_norm": 2.2602560686524273,
+      "learning_rate": 6.20146669192567e-05,
+      "loss": 1.3278,
+      "step": 2640
+    },
+    {
+      "epoch": 147.22222222222223,
+      "grad_norm": 2.343799234508914,
+      "learning_rate": 6.188797270841548e-05,
+      "loss": 1.3517,
+      "step": 2650
+    },
+    {
+      "epoch": 147.77777777777777,
+      "grad_norm": 2.416660626721782,
+      "learning_rate": 6.176097573839264e-05,
+      "loss": 1.3303,
+      "step": 2660
+    },
+    {
+      "epoch": 148.33333333333334,
+      "grad_norm": 2.4520797932571443,
+      "learning_rate": 6.16336779923506e-05,
+      "loss": 1.3189,
+      "step": 2670
+    },
+    {
+      "epoch": 148.88888888888889,
+      "grad_norm": 2.4168299148493655,
+      "learning_rate": 6.150608145814863e-05,
+      "loss": 1.3244,
+      "step": 2680
+    },
+    {
+      "epoch": 149.44444444444446,
+      "grad_norm": 2.4400645727462096,
+      "learning_rate": 6.137818812831183e-05,
+      "loss": 1.317,
+      "step": 2690
+    },
+    {
+      "epoch": 150.0,
+      "grad_norm": 2.3507755014508374,
+      "learning_rate": 6.125000000000001e-05,
+      "loss": 1.318,
+      "step": 2700
+    },
+    {
+      "epoch": 150.55555555555554,
+      "grad_norm": 2.5407585230951533,
+      "learning_rate": 6.11215190749765e-05,
+      "loss": 1.3221,
+      "step": 2710
+    },
+    {
+      "epoch": 151.11111111111111,
+      "grad_norm": 2.287937864493581,
+      "learning_rate": 6.0992747359576896e-05,
+      "loss": 1.2733,
+      "step": 2720
+    },
+    {
+      "epoch": 151.66666666666666,
+      "grad_norm": 2.365413391243613,
+      "learning_rate": 6.0863686864677734e-05,
+      "loss": 1.2991,
+      "step": 2730
+    },
+    {
+      "epoch": 152.22222222222223,
+      "grad_norm": 2.495331792691151,
+      "learning_rate": 6.0734339605665026e-05,
+      "loss": 1.3122,
+      "step": 2740
+    },
+    {
+      "epoch": 152.77777777777777,
+      "grad_norm": 2.2367825641915746,
+      "learning_rate": 6.060470760240294e-05,
+      "loss": 1.2901,
+      "step": 2750
+    },
+    {
+      "epoch": 153.33333333333334,
+      "grad_norm": 2.361473009635731,
+      "learning_rate": 6.047479287920208e-05,
+      "loss": 1.2832,
+      "step": 2760
+    },
+    {
+      "epoch": 153.88888888888889,
+      "grad_norm": 2.4837243529686504,
+      "learning_rate": 6.0344597464787954e-05,
+      "loss": 1.2679,
+      "step": 2770
+    },
+    {
+      "epoch": 154.44444444444446,
+      "grad_norm": 2.3330651424321154,
+      "learning_rate": 6.021412339226936e-05,
+      "loss": 1.2542,
+      "step": 2780
+    },
+    {
+      "epoch": 155.0,
+      "grad_norm": 2.188237529128572,
+      "learning_rate": 6.008337269910651e-05,
+      "loss": 1.2527,
+      "step": 2790
+    },
+    {
+      "epoch": 155.55555555555554,
+      "grad_norm": 2.366065738393517,
+      "learning_rate": 5.9952347427079305e-05,
+      "loss": 1.2506,
+      "step": 2800
+    },
+    {
+      "epoch": 156.11111111111111,
+      "grad_norm": 2.324750259276332,
+      "learning_rate": 5.9821049622255403e-05,
+      "loss": 1.2748,
+      "step": 2810
+    },
+    {
+      "epoch": 156.66666666666666,
+      "grad_norm": 2.3360207839459273,
+      "learning_rate": 5.968948133495832e-05,
+      "loss": 1.2142,
+      "step": 2820
+    },
+    {
+      "epoch": 157.22222222222223,
+      "grad_norm": 2.5058026682740007,
+      "learning_rate": 5.955764461973535e-05,
+      "loss": 1.2548,
+      "step": 2830
+    },
+    {
+      "epoch": 157.77777777777777,
+      "grad_norm": 2.3373577176613938,
+      "learning_rate": 5.9425541535325495e-05,
+      "loss": 1.2256,
+      "step": 2840
+    },
+    {
+      "epoch": 158.33333333333334,
+      "grad_norm": 2.270085331157418,
+      "learning_rate": 5.929317414462738e-05,
+      "loss": 1.2316,
+      "step": 2850
+    },
+    {
+      "epoch": 158.88888888888889,
+      "grad_norm": 2.342955391016746,
+      "learning_rate": 5.916054451466694e-05,
+      "loss": 1.1791,
+      "step": 2860
+    },
+    {
+      "epoch": 159.44444444444446,
+      "grad_norm": 2.3399377972739113,
+      "learning_rate": 5.9027654716565247e-05,
+      "loss": 1.178,
+      "step": 2870
+    },
+    {
+      "epoch": 160.0,
+      "grad_norm": 2.5947773748611946,
+      "learning_rate": 5.889450682550605e-05,
+      "loss": 1.2342,
+      "step": 2880
+    },
+    {
+      "epoch": 160.55555555555554,
+      "grad_norm": 2.5886440346404975,
+      "learning_rate": 5.8761102920703476e-05,
+      "loss": 1.1762,
+      "step": 2890
+    },
+    {
+      "epoch": 161.11111111111111,
+      "grad_norm": 2.4699068859634346,
+      "learning_rate": 5.862744508536953e-05,
+      "loss": 1.2057,
+      "step": 2900
+    },
+    {
+      "epoch": 161.66666666666666,
+      "grad_norm": 2.2415259046451075,
+      "learning_rate": 5.84935354066815e-05,
+      "loss": 1.1675,
+      "step": 2910
+    },
+    {
+      "epoch": 162.22222222222223,
+      "grad_norm": 2.2555072988579346,
+      "learning_rate": 5.835937597574944e-05,
+      "loss": 1.1841,
+      "step": 2920
+    },
+    {
+      "epoch": 162.77777777777777,
+      "grad_norm": 2.379456304589441,
+      "learning_rate": 5.822496888758352e-05,
+      "loss": 1.1772,
+      "step": 2930
+    },
+    {
+      "epoch": 163.33333333333334,
+      "grad_norm": 2.288367156047133,
+      "learning_rate": 5.8090316241061225e-05,
+      "loss": 1.1734,
+      "step": 2940
+    },
+    {
+      "epoch": 163.88888888888889,
+      "grad_norm": 2.4845313256941868,
+      "learning_rate": 5.7955420138894706e-05,
+      "loss": 1.1809,
+      "step": 2950
+    },
+    {
+      "epoch": 164.44444444444446,
+      "grad_norm": 2.4047479276147494,
+      "learning_rate": 5.7820282687597816e-05,
+      "loss": 1.1749,
+      "step": 2960
+    },
+    {
+      "epoch": 165.0,
+      "grad_norm": 2.3019825724879164,
+      "learning_rate": 5.7684905997453294e-05,
+      "loss": 1.1622,
+      "step": 2970
+    },
+    {
+      "epoch": 165.55555555555554,
+      "grad_norm": 2.3499977761116058,
+      "learning_rate": 5.7549292182479784e-05,
+      "loss": 1.1502,
+      "step": 2980
+    },
+    {
+      "epoch": 166.11111111111111,
+      "grad_norm": 2.446050563005803,
+      "learning_rate": 5.741344336039887e-05,
+      "loss": 1.1055,
+      "step": 2990
+    },
+    {
+      "epoch": 166.66666666666666,
+      "grad_norm": 2.4719435883662357,
+      "learning_rate": 5.727736165260188e-05,
+      "loss": 1.1167,
+      "step": 3000
+    },
+    {
+      "epoch": 167.22222222222223,
+      "grad_norm": 2.27827897669987,
+      "learning_rate": 5.714104918411696e-05,
+      "loss": 1.1244,
+      "step": 3010
+    },
+    {
+      "epoch": 167.77777777777777,
+      "grad_norm": 2.240720474996895,
+      "learning_rate": 5.7004508083575645e-05,
+      "loss": 1.1132,
+      "step": 3020
+    },
+    {
+      "epoch": 168.33333333333334,
+      "grad_norm": 2.389602242893562,
+      "learning_rate": 5.686774048317986e-05,
+      "loss": 1.1306,
+      "step": 3030
+    },
+    {
+      "epoch": 168.88888888888889,
+      "grad_norm": 2.2844622218313995,
+      "learning_rate": 5.6730748518668455e-05,
+      "loss": 1.0918,
+      "step": 3040
+    },
+    {
+      "epoch": 169.44444444444446,
+      "grad_norm": 2.2515737539906224,
+      "learning_rate": 5.6593534329283935e-05,
+      "loss": 1.1458,
+      "step": 3050
+    },
+    {
+      "epoch": 170.0,
+      "grad_norm": 2.2338284977898017,
+      "learning_rate": 5.645610005773901e-05,
+      "loss": 1.0875,
+      "step": 3060
+    },
+    {
+      "epoch": 170.55555555555554,
+      "grad_norm": 2.2800559485413956,
+      "learning_rate": 5.631844785018314e-05,
+      "loss": 1.0782,
+      "step": 3070
+    },
+    {
+      "epoch": 171.11111111111111,
+      "grad_norm": 2.2966384246854044,
+      "learning_rate": 5.6180579856169076e-05,
+      "loss": 1.0576,
+      "step": 3080
+    },
+    {
+      "epoch": 171.66666666666666,
+      "grad_norm": 2.379360813251885,
+      "learning_rate": 5.604249822861923e-05,
+      "loss": 1.1166,
+      "step": 3090
+    },
+    {
+      "epoch": 172.22222222222223,
+      "grad_norm": 2.2538050161069085,
+      "learning_rate": 5.5904205123792096e-05,
+      "loss": 1.065,
+      "step": 3100
+    },
+    {
+      "epoch": 172.77777777777777,
+      "grad_norm": 2.3419993665386487,
+      "learning_rate": 5.576570270124853e-05,
+      "loss": 1.0603,
+      "step": 3110
+    },
+    {
+      "epoch": 173.33333333333334,
+      "grad_norm": 2.2852459751348686,
+      "learning_rate": 5.562699312381807e-05,
+      "loss": 1.0839,
+      "step": 3120
+    },
+    {
+      "epoch": 173.88888888888889,
+      "grad_norm": 2.3920048739278377,
+      "learning_rate": 5.548807855756516e-05,
+      "loss": 1.0482,
+      "step": 3130
+    },
+    {
+      "epoch": 174.44444444444446,
+      "grad_norm": 2.183189286389313,
+      "learning_rate": 5.53489611717553e-05,
+      "loss": 1.0589,
+      "step": 3140
+    },
+    {
+      "epoch": 175.0,
+      "grad_norm": 2.317519951068878,
+      "learning_rate": 5.5209643138821204e-05,
+      "loss": 1.0508,
+      "step": 3150
+    },
+    {
+      "epoch": 175.55555555555554,
+      "grad_norm": 2.1790092260647707,
+      "learning_rate": 5.50701266343288e-05,
+      "loss": 1.032,
+      "step": 3160
+    },
+    {
+      "epoch": 176.11111111111111,
+      "grad_norm": 2.2551249311734805,
+      "learning_rate": 5.4930413836943396e-05,
+      "loss": 1.0354,
+      "step": 3170
+    },
+    {
+      "epoch": 176.66666666666666,
+      "grad_norm": 2.19739474447012,
+      "learning_rate": 5.47905069283955e-05,
+      "loss": 1.039,
+      "step": 3180
+    },
+    {
+      "epoch": 177.22222222222223,
+      "grad_norm": 2.179176735596404,
+      "learning_rate": 5.465040809344691e-05,
+      "loss": 1.0675,
+      "step": 3190
+    },
+    {
+      "epoch": 177.77777777777777,
+      "grad_norm": 2.266135848758427,
+      "learning_rate": 5.451011951985641e-05,
+      "loss": 0.9877,
+      "step": 3200
+    },
+    {
+      "epoch": 178.33333333333334,
+      "grad_norm": 2.2904217430790847,
+      "learning_rate": 5.436964339834579e-05,
+      "loss": 1.001,
+      "step": 3210
+    },
+    {
+      "epoch": 178.88888888888889,
+      "grad_norm": 2.2529184699140976,
+      "learning_rate": 5.4228981922565525e-05,
+      "loss": 1.0472,
+      "step": 3220
+    },
+    {
+      "epoch": 179.44444444444446,
+      "grad_norm": 2.314495874967791,
+      "learning_rate": 5.408813728906053e-05,
+      "loss": 1.0475,
+      "step": 3230
+    },
+    {
+      "epoch": 180.0,
+      "grad_norm": 2.1559484519741146,
+      "learning_rate": 5.394711169723593e-05,
+      "loss": 1.0118,
+      "step": 3240
+    },
+    {
+      "epoch": 180.55555555555554,
+      "grad_norm": 2.203835014482057,
+      "learning_rate": 5.380590734932262e-05,
+      "loss": 1.0282,
+      "step": 3250
+    },
+    {
+      "epoch": 181.11111111111111,
+      "grad_norm": 2.259884105006339,
+      "learning_rate": 5.366452645034293e-05,
+      "loss": 1.0045,
+      "step": 3260
+    },
+    {
+      "epoch": 181.66666666666666,
+      "grad_norm": 2.369344301580839,
+      "learning_rate": 5.352297120807619e-05,
+      "loss": 1.0005,
+      "step": 3270
+    },
+    {
+      "epoch": 182.22222222222223,
+      "grad_norm": 2.1281099562596375,
+      "learning_rate": 5.3381243833024236e-05,
+      "loss": 0.99,
+      "step": 3280
+    },
+    {
+      "epoch": 182.77777777777777,
+      "grad_norm": 2.1887871361440916,
+      "learning_rate": 5.3239346538376895e-05,
+      "loss": 0.9823,
+      "step": 3290
+    },
+    {
+      "epoch": 183.33333333333334,
+      "grad_norm": 2.2268847851320106,
+      "learning_rate": 5.3097281539977465e-05,
+      "loss": 0.9622,
+      "step": 3300
+    },
+    {
+      "epoch": 183.88888888888889,
+      "grad_norm": 2.1096761982948027,
+      "learning_rate": 5.2955051056288006e-05,
+      "loss": 0.9967,
+      "step": 3310
+    },
+    {
+      "epoch": 184.44444444444446,
+      "grad_norm": 2.1515698711903393,
+      "learning_rate": 5.281265730835483e-05,
+      "loss": 0.951,
+      "step": 3320
+    },
+    {
+      "epoch": 185.0,
+      "grad_norm": 2.1728798716554736,
+      "learning_rate": 5.267010251977371e-05,
+      "loss": 0.9697,
+      "step": 3330
+    },
+    {
+      "epoch": 185.55555555555554,
+      "grad_norm": 2.2526544172088094,
+      "learning_rate": 5.252738891665525e-05,
+      "loss": 0.9655,
+      "step": 3340
+    },
+    {
+      "epoch": 186.11111111111111,
+      "grad_norm": 2.309482048557876,
+      "learning_rate": 5.2384518727590035e-05,
+      "loss": 0.957,
+      "step": 3350
+    },
+    {
+      "epoch": 186.66666666666666,
+      "grad_norm": 3.214573344727921,
+      "learning_rate": 5.224149418361388e-05,
+      "loss": 0.9699,
+      "step": 3360
+    },
+    {
+      "epoch": 187.22222222222223,
+      "grad_norm": 2.274244866439059,
+      "learning_rate": 5.2098317518173004e-05,
+      "loss": 0.9907,
+      "step": 3370
+    },
+    {
+      "epoch": 187.77777777777777,
+      "grad_norm": 2.1421898461349183,
+      "learning_rate": 5.195499096708908e-05,
+      "loss": 0.9457,
+      "step": 3380
+    },
+    {
+      "epoch": 188.33333333333334,
+      "grad_norm": 2.3156410741827282,
+      "learning_rate": 5.1811516768524385e-05,
+      "loss": 1.0787,
+      "step": 3390
+    },
+    {
+      "epoch": 188.88888888888889,
+      "grad_norm": 2.3962526590236024,
+      "learning_rate": 5.166789716294686e-05,
+      "loss": 0.9259,
+      "step": 3400
+    },
+    {
+      "epoch": 189.44444444444446,
+      "grad_norm": 2.261476283014636,
+      "learning_rate": 5.152413439309508e-05,
+      "loss": 0.9731,
+      "step": 3410
+    },
+    {
+      "epoch": 190.0,
+      "grad_norm": 2.141578072324719,
+      "learning_rate": 5.1380230703943226e-05,
+      "loss": 0.9291,
+      "step": 3420
+    },
+    {
+      "epoch": 190.55555555555554,
+      "grad_norm": 2.1517757490161364,
+      "learning_rate": 5.1236188342666084e-05,
+      "loss": 0.9053,
+      "step": 3430
+    },
+    {
+      "epoch": 191.11111111111111,
+      "grad_norm": 2.261919556237122,
+      "learning_rate": 5.109200955860392e-05,
+      "loss": 0.944,
+      "step": 3440
+    },
+    {
+      "epoch": 191.66666666666666,
+      "grad_norm": 2.1587716436930138,
+      "learning_rate": 5.0947696603227316e-05,
+      "loss": 0.8758,
+      "step": 3450
+    },
+    {
+      "epoch": 192.22222222222223,
+      "grad_norm": 2.3207229610841686,
+      "learning_rate": 5.0803251730102097e-05,
+      "loss": 0.9159,
+      "step": 3460
+    },
+    {
+      "epoch": 192.77777777777777,
+      "grad_norm": 2.0660371216340647,
+      "learning_rate": 5.065867719485405e-05,
+      "loss": 0.881,
+      "step": 3470
+    },
+    {
+      "epoch": 193.33333333333334,
+      "grad_norm": 2.1833927287207886,
+      "learning_rate": 5.051397525513377e-05,
+      "loss": 0.8856,
+      "step": 3480
+    },
+    {
+      "epoch": 193.88888888888889,
+      "grad_norm": 2.132586121335054,
+      "learning_rate": 5.036914817058136e-05,
+      "loss": 0.9095,
+      "step": 3490
+    },
+    {
+      "epoch": 194.44444444444446,
+      "grad_norm": 2.1648577939875118,
+      "learning_rate": 5.022419820279119e-05,
+      "loss": 0.8966,
+      "step": 3500
+    },
+    {
+      "epoch": 195.0,
+      "grad_norm": 2.1854436746049086,
+      "learning_rate": 5.0079127615276534e-05,
+      "loss": 0.8733,
+      "step": 3510
+    },
+    {
+      "epoch": 195.55555555555554,
+      "grad_norm": 2.2467295932541274,
+      "learning_rate": 4.993393867343423e-05,
+      "loss": 0.9017,
+      "step": 3520
+    },
+    {
+      "epoch": 196.11111111111111,
+      "grad_norm": 2.2855930530379767,
+      "learning_rate": 4.978863364450935e-05,
+      "loss": 0.8868,
+      "step": 3530
+    },
+    {
+      "epoch": 196.66666666666666,
+      "grad_norm": 2.2161862242290575,
+      "learning_rate": 4.964321479755974e-05,
+      "loss": 0.8775,
+      "step": 3540
+    },
+    {
+      "epoch": 197.22222222222223,
+      "grad_norm": 2.1447136806583704,
+      "learning_rate": 4.949768440342059e-05,
+      "loss": 0.8834,
+      "step": 3550
+    },
+    {
+      "epoch": 197.77777777777777,
+      "grad_norm": 2.160335043600958,
+      "learning_rate": 4.935204473466903e-05,
+      "loss": 0.8751,
+      "step": 3560
+    },
+    {
+      "epoch": 198.33333333333334,
+      "grad_norm": 2.226851916326021,
+      "learning_rate": 4.920629806558859e-05,
+      "loss": 0.8594,
+      "step": 3570
+    },
+    {
+      "epoch": 198.88888888888889,
+      "grad_norm": 2.1013674734161323,
+      "learning_rate": 4.906044667213368e-05,
+      "loss": 0.8419,
+      "step": 3580
+    },
+    {
+      "epoch": 199.44444444444446,
+      "grad_norm": 2.121429023449048,
+      "learning_rate": 4.891449283189408e-05,
+      "loss": 0.8326,
+      "step": 3590
+    },
+    {
+      "epoch": 200.0,
+      "grad_norm": 2.1562036702804104,
+      "learning_rate": 4.876843882405937e-05,
+      "loss": 0.8334,
+      "step": 3600
+    },
+    {
+      "epoch": 200.55555555555554,
+      "grad_norm": 2.2420077899862183,
+      "learning_rate": 4.862228692938334e-05,
+      "loss": 0.8048,
+      "step": 3610
+    },
+    {
+      "epoch": 201.11111111111111,
+      "grad_norm": 2.134219421703876,
+      "learning_rate": 4.8476039430148316e-05,
+      "loss": 0.8532,
+      "step": 3620
+    },
+    {
+      "epoch": 201.66666666666666,
+      "grad_norm": 2.0203886296561238,
+      "learning_rate": 4.83296986101296e-05,
+      "loss": 0.8103,
+      "step": 3630
+    },
+    {
+      "epoch": 202.22222222222223,
+      "grad_norm": 2.0926855997154927,
+      "learning_rate": 4.818326675455978e-05,
+      "loss": 0.8455,
+      "step": 3640
+    },
+    {
+      "epoch": 202.77777777777777,
+      "grad_norm": 2.0798054261166237,
+      "learning_rate": 4.803674615009305e-05,
+      "loss": 0.8383,
+      "step": 3650
+    },
+    {
+      "epoch": 203.33333333333334,
+      "grad_norm": 2.30133464485579,
+      "learning_rate": 4.789013908476944e-05,
+      "loss": 0.8017,
+      "step": 3660
+    },
+    {
+      "epoch": 203.88888888888889,
+      "grad_norm": 2.224684930846259,
+      "learning_rate": 4.7743447847979194e-05,
+      "loss": 0.8314,
+      "step": 3670
+    },
+    {
+      "epoch": 204.44444444444446,
+      "grad_norm": 2.1403076604921263,
+      "learning_rate": 4.7596674730426936e-05,
+      "loss": 0.8089,
+      "step": 3680
+    },
+    {
+      "epoch": 205.0,
+      "grad_norm": 2.153198626273998,
+      "learning_rate": 4.744982202409592e-05,
+      "loss": 0.8134,
+      "step": 3690
+    },
+    {
+      "epoch": 205.55555555555554,
+      "grad_norm": 2.1002788948874898,
+      "learning_rate": 4.730289202221226e-05,
+      "loss": 0.7876,
+      "step": 3700
+    },
+    {
+      "epoch": 206.11111111111111,
+      "grad_norm": 2.1449534516519044,
+      "learning_rate": 4.7155887019209106e-05,
+      "loss": 0.8164,
+      "step": 3710
+    },
+    {
+      "epoch": 206.66666666666666,
+      "grad_norm": 2.061252679215836,
+      "learning_rate": 4.700880931069078e-05,
+      "loss": 0.824,
+      "step": 3720
+    },
+    {
+      "epoch": 207.22222222222223,
+      "grad_norm": 2.0174689323368367,
+      "learning_rate": 4.6861661193397e-05,
+      "loss": 0.7971,
+      "step": 3730
+    },
+    {
+      "epoch": 207.77777777777777,
+      "grad_norm": 2.148462468782465,
+      "learning_rate": 4.671444496516696e-05,
+      "loss": 0.7856,
+      "step": 3740
+    },
+    {
+      "epoch": 208.33333333333334,
+      "grad_norm": 2.2281360085577893,
+      "learning_rate": 4.6567162924903464e-05,
+      "loss": 0.8278,
+      "step": 3750
+    },
+    {
+      "epoch": 208.88888888888889,
+      "grad_norm": 2.144614073885888,
+      "learning_rate": 4.6419817372537015e-05,
+      "loss": 0.8069,
+      "step": 3760
+    },
+    {
+      "epoch": 209.44444444444446,
+      "grad_norm": 2.113686908703004,
+      "learning_rate": 4.627241060898991e-05,
+      "loss": 0.7792,
+      "step": 3770
+    },
+    {
+      "epoch": 210.0,
+      "grad_norm": 2.1014203445400974,
+      "learning_rate": 4.612494493614034e-05,
+      "loss": 0.7559,
+      "step": 3780
+    },
+    {
+      "epoch": 210.55555555555554,
+      "grad_norm": 2.1732960175702187,
+      "learning_rate": 4.597742265678638e-05,
+      "loss": 0.791,
+      "step": 3790
+    },
+    {
+      "epoch": 211.11111111111111,
+      "grad_norm": 2.0750735258381274,
+      "learning_rate": 4.582984607461006e-05,
+      "loss": 0.7557,
+      "step": 3800
+    },
+    {
+      "epoch": 211.66666666666666,
+      "grad_norm": 2.1478459324060064,
+      "learning_rate": 4.568221749414142e-05,
+      "loss": 0.752,
+      "step": 3810
+    },
+    {
+      "epoch": 212.22222222222223,
+      "grad_norm": 2.2279634050620434,
+      "learning_rate": 4.553453922072248e-05,
+      "loss": 0.7733,
+      "step": 3820
+    },
+    {
+      "epoch": 212.77777777777777,
+      "grad_norm": 1.975072547932186,
+      "learning_rate": 4.538681356047126e-05,
+      "loss": 0.7524,
+      "step": 3830
+    },
+    {
+      "epoch": 213.33333333333334,
+      "grad_norm": 1.9838406301370255,
+      "learning_rate": 4.523904282024575e-05,
+      "loss": 0.7482,
+      "step": 3840
+    },
+    {
+      "epoch": 213.88888888888889,
+      "grad_norm": 1.9800639143174406,
+      "learning_rate": 4.509122930760793e-05,
+      "loss": 0.7328,
+      "step": 3850
+    },
+    {
+      "epoch": 214.44444444444446,
+      "grad_norm": 2.0267730674085,
+      "learning_rate": 4.494337533078768e-05,
+      "loss": 0.7323,
+      "step": 3860
+    },
+    {
+      "epoch": 215.0,
+      "grad_norm": 2.3714002131540903,
+      "learning_rate": 4.479548319864678e-05,
+      "loss": 0.7521,
+      "step": 3870
+    },
+    {
+      "epoch": 215.55555555555554,
+      "grad_norm": 2.0361877287633274,
+      "learning_rate": 4.464755522064282e-05,
+      "loss": 0.7201,
+      "step": 3880
+    },
+    {
+      "epoch": 216.11111111111111,
+      "grad_norm": 2.10056516671652,
+      "learning_rate": 4.449959370679317e-05,
+      "loss": 0.7781,
+      "step": 3890
+    },
+    {
+      "epoch": 216.66666666666666,
+      "grad_norm": 2.0108587408166687,
+      "learning_rate": 4.435160096763887e-05,
+      "loss": 0.7205,
+      "step": 3900
+    },
+    {
+      "epoch": 217.22222222222223,
+      "grad_norm": 2.163597304819988,
+      "learning_rate": 4.4203579314208594e-05,
+      "loss": 0.7296,
+      "step": 3910
+    },
+    {
+      "epoch": 217.77777777777777,
+      "grad_norm": 2.0416109375905838,
+      "learning_rate": 4.40555310579825e-05,
+      "loss": 0.7289,
+      "step": 3920
+    },
+    {
+      "epoch": 218.33333333333334,
+      "grad_norm": 2.1968095229743323,
+      "learning_rate": 4.390745851085622e-05,
+      "loss": 0.7131,
+      "step": 3930
+    },
+    {
+      "epoch": 218.88888888888889,
+      "grad_norm": 2.1012541250430075,
+      "learning_rate": 4.375936398510467e-05,
+      "loss": 0.7234,
+      "step": 3940
+    },
+    {
+      "epoch": 219.44444444444446,
+      "grad_norm": 2.206656257032445,
+      "learning_rate": 4.3611249793346e-05,
+      "loss": 0.7166,
+      "step": 3950
+    },
+    {
+      "epoch": 220.0,
+      "grad_norm": 2.057058960191551,
+      "learning_rate": 4.3463118248505436e-05,
+      "loss": 0.735,
+      "step": 3960
+    },
+    {
+      "epoch": 220.55555555555554,
+      "grad_norm": 2.1798208325805355,
+      "learning_rate": 4.3314971663779214e-05,
+      "loss": 0.7026,
+      "step": 3970
+    },
+    {
+      "epoch": 221.11111111111111,
+      "grad_norm": 2.0952505172220897,
+      "learning_rate": 4.316681235259841e-05,
+      "loss": 0.7288,
+      "step": 3980
+    },
+    {
+      "epoch": 221.66666666666666,
+      "grad_norm": 2.0147090279606608,
+      "learning_rate": 4.301864262859285e-05,
+      "loss": 0.6801,
+      "step": 3990
+    },
+    {
+      "epoch": 222.22222222222223,
+      "grad_norm": 1.9135849960424693,
+      "learning_rate": 4.287046480555496e-05,
+      "loss": 0.6849,
+      "step": 4000
+    },
+    {
+      "epoch": 222.77777777777777,
+      "grad_norm": 2.116502827088009,
+      "learning_rate": 4.272228119740365e-05,
+      "loss": 0.7298,
+      "step": 4010
+    },
+    {
+      "epoch": 223.33333333333334,
+      "grad_norm": 2.055602359525266,
+      "learning_rate": 4.257409411814814e-05,
+      "loss": 0.6563,
+      "step": 4020
+    },
+    {
+      "epoch": 223.88888888888889,
+      "grad_norm": 2.0815240124012218,
+      "learning_rate": 4.242590588185188e-05,
+      "loss": 0.6949,
+      "step": 4030
+    },
+    {
+      "epoch": 224.44444444444446,
+      "grad_norm": 1.9879271785938155,
+      "learning_rate": 4.227771880259636e-05,
+      "loss": 0.6803,
+      "step": 4040
+    },
+    {
+      "epoch": 225.0,
+      "grad_norm": 2.156109902073769,
+      "learning_rate": 4.2129535194445045e-05,
+      "loss": 0.7001,
+      "step": 4050
+    },
+    {
+      "epoch": 225.55555555555554,
+      "grad_norm": 2.0787850385822244,
+      "learning_rate": 4.1981357371407156e-05,
+      "loss": 0.6809,
+      "step": 4060
+    },
+    {
+      "epoch": 226.11111111111111,
+      "grad_norm": 1.8242856652609434,
+      "learning_rate": 4.18331876474016e-05,
+      "loss": 0.6738,
+      "step": 4070
+    },
+    {
+      "epoch": 226.66666666666666,
+      "grad_norm": 2.1111516181785106,
+      "learning_rate": 4.1685028336220806e-05,
+      "loss": 0.6729,
+      "step": 4080
+    },
+    {
+      "epoch": 227.22222222222223,
+      "grad_norm": 2.01080858696219,
+      "learning_rate": 4.1536881751494577e-05,
+      "loss": 0.6791,
+      "step": 4090
+    },
+    {
+      "epoch": 227.77777777777777,
+      "grad_norm": 2.098654284041219,
+      "learning_rate": 4.138875020665401e-05,
+      "loss": 0.6363,
+      "step": 4100
+    },
+    {
+      "epoch": 228.33333333333334,
+      "grad_norm": 2.078208175945013,
+      "learning_rate": 4.1240636014895337e-05,
+      "loss": 0.6833,
+      "step": 4110
+    },
+    {
+      "epoch": 228.88888888888889,
+      "grad_norm": 1.9714231961929303,
+      "learning_rate": 4.109254148914379e-05,
+      "loss": 0.6609,
+      "step": 4120
+    },
+    {
+      "epoch": 229.44444444444446,
+      "grad_norm": 2.0230908188924235,
+      "learning_rate": 4.094446894201751e-05,
+      "loss": 0.6646,
+      "step": 4130
+    },
+    {
+      "epoch": 230.0,
+      "grad_norm": 2.011414025199144,
+      "learning_rate": 4.0796420685791426e-05,
+      "loss": 0.6537,
+      "step": 4140
+    },
+    {
+      "epoch": 230.55555555555554,
+      "grad_norm": 1.964161081199597,
+      "learning_rate": 4.064839903236115e-05,
+      "loss": 0.6477,
+      "step": 4150
+    },
+    {
+      "epoch": 231.11111111111111,
+      "grad_norm": 2.0068768053928787,
+      "learning_rate": 4.050040629320684e-05,
+      "loss": 0.6579,
+      "step": 4160
+    },
+    {
+      "epoch": 231.66666666666666,
+      "grad_norm": 1.9675635138297263,
+      "learning_rate": 4.035244477935719e-05,
+      "loss": 0.6646,
+      "step": 4170
+    },
+    {
+      "epoch": 232.22222222222223,
+      "grad_norm": 1.9567251975009563,
+      "learning_rate": 4.020451680135323e-05,
+      "loss": 0.646,
+      "step": 4180
+    },
+    {
+      "epoch": 232.77777777777777,
+      "grad_norm": 1.9473363926615457,
+      "learning_rate": 4.005662466921234e-05,
+      "loss": 0.6543,
+      "step": 4190
+    },
+    {
+      "epoch": 233.33333333333334,
+      "grad_norm": 1.9586815490641993,
+      "learning_rate": 3.9908770692392085e-05,
+      "loss": 0.6454,
+      "step": 4200
+    },
+    {
+      "epoch": 233.88888888888889,
+      "grad_norm": 2.035335598765784,
+      "learning_rate": 3.976095717975426e-05,
+      "loss": 0.6261,
+      "step": 4210
+    },
+    {
+      "epoch": 234.44444444444446,
+      "grad_norm": 1.9015379654140836,
+      "learning_rate": 3.9613186439528754e-05,
+      "loss": 0.589,
+      "step": 4220
+    },
+    {
+      "epoch": 235.0,
+      "grad_norm": 2.0713819512631653,
+      "learning_rate": 3.946546077927753e-05,
+      "loss": 0.6337,
+      "step": 4230
+    },
+    {
+      "epoch": 235.55555555555554,
+      "grad_norm": 1.9767983638877071,
+      "learning_rate": 3.931778250585859e-05,
+      "loss": 0.5989,
+      "step": 4240
+    },
+    {
+      "epoch": 236.11111111111111,
+      "grad_norm": 2.0243552935949287,
+      "learning_rate": 3.917015392538995e-05,
+      "loss": 0.6198,
+      "step": 4250
+    },
+    {
+      "epoch": 236.66666666666666,
+      "grad_norm": 2.0494900371516893,
+      "learning_rate": 3.9022577343213635e-05,
+      "loss": 0.6053,
+      "step": 4260
+    },
+    {
+      "epoch": 237.22222222222223,
+      "grad_norm": 1.8584984388121666,
+      "learning_rate": 3.887505506385967e-05,
+      "loss": 0.6232,
+      "step": 4270
+    },
+    {
+      "epoch": 237.77777777777777,
+      "grad_norm": 1.917128804372393,
+      "learning_rate": 3.87275893910101e-05,
+      "loss": 0.5993,
+      "step": 4280
+    },
+    {
+      "epoch": 238.33333333333334,
+      "grad_norm": 2.015729324834695,
+      "learning_rate": 3.8580182627463004e-05,
+      "loss": 0.6149,
+      "step": 4290
+    },
+    {
+      "epoch": 238.88888888888889,
+      "grad_norm": 2.0064773573350005,
+      "learning_rate": 3.843283707509655e-05,
+      "loss": 0.6193,
+      "step": 4300
+    },
+    {
+      "epoch": 239.44444444444446,
+      "grad_norm": 2.0107558232827474,
+      "learning_rate": 3.8285555034833046e-05,
+      "loss": 0.5883,
+      "step": 4310
+    },
+    {
+      "epoch": 240.0,
+      "grad_norm": 1.99926762284486,
+      "learning_rate": 3.8138338806603005e-05,
+      "loss": 0.588,
+      "step": 4320
+    },
+    {
+      "epoch": 240.55555555555554,
+      "grad_norm": 1.9325627207971967,
+      "learning_rate": 3.799119068930923e-05,
+      "loss": 0.5743,
+      "step": 4330
+    },
+    {
+      "epoch": 241.11111111111111,
+      "grad_norm": 2.023160819796288,
+      "learning_rate": 3.784411298079091e-05,
+      "loss": 0.5913,
+      "step": 4340
+    },
+    {
+      "epoch": 241.66666666666666,
+      "grad_norm": 1.85426950348123,
+      "learning_rate": 3.769710797778775e-05,
+      "loss": 0.5811,
+      "step": 4350
+    },
+    {
+      "epoch": 242.22222222222223,
+      "grad_norm": 1.937002918097445,
+      "learning_rate": 3.7550177975904086e-05,
+      "loss": 0.6042,
+      "step": 4360
+    },
+    {
+      "epoch": 242.77777777777777,
+      "grad_norm": 1.8868097318901735,
+      "learning_rate": 3.740332526957308e-05,
+      "loss": 0.5736,
+      "step": 4370
+    },
+    {
+      "epoch": 243.33333333333334,
+      "grad_norm": 1.9052987148342444,
+      "learning_rate": 3.725655215202081e-05,
+      "loss": 0.5604,
+      "step": 4380
+    },
+    {
+      "epoch": 243.88888888888889,
+      "grad_norm": 1.874160896734524,
+      "learning_rate": 3.7109860915230564e-05,
+      "loss": 0.586,
+      "step": 4390
+    },
+    {
+      "epoch": 244.44444444444446,
+      "grad_norm": 1.917309121251374,
+      "learning_rate": 3.696325384990696e-05,
+      "loss": 0.5784,
+      "step": 4400
+    },
+    {
+      "epoch": 245.0,
+      "grad_norm": 1.8395452236952865,
+      "learning_rate": 3.681673324544022e-05,
+      "loss": 0.5701,
+      "step": 4410
+    },
+    {
+      "epoch": 245.55555555555554,
+      "grad_norm": 1.9798284394662176,
+      "learning_rate": 3.6670301389870404e-05,
+      "loss": 0.561,
+      "step": 4420
+    },
+    {
+      "epoch": 246.11111111111111,
+      "grad_norm": 1.9701177901668616,
+      "learning_rate": 3.65239605698517e-05,
+      "loss": 0.5813,
+      "step": 4430
+    },
+    {
+      "epoch": 246.66666666666666,
+      "grad_norm": 1.8602570437152004,
+      "learning_rate": 3.637771307061668e-05,
+      "loss": 0.5795,
+      "step": 4440
+    },
+    {
+      "epoch": 247.22222222222223,
+      "grad_norm": 1.8355552113001339,
+      "learning_rate": 3.6231561175940636e-05,
+      "loss": 0.5775,
+      "step": 4450
+    },
+    {
+      "epoch": 247.77777777777777,
+      "grad_norm": 1.839014664305757,
+      "learning_rate": 3.608550716810593e-05,
+      "loss": 0.5467,
+      "step": 4460
+    },
+    {
+      "epoch": 248.33333333333334,
+      "grad_norm": 1.844967004020913,
+      "learning_rate": 3.593955332786634e-05,
+      "loss": 0.5478,
+      "step": 4470
+    },
+    {
+      "epoch": 248.88888888888889,
+      "grad_norm": 1.8923740335970123,
+      "learning_rate": 3.579370193441142e-05,
+      "loss": 0.5616,
+      "step": 4480
+    },
+    {
+      "epoch": 249.44444444444446,
+      "grad_norm": 1.8906046338206066,
+      "learning_rate": 3.5647955265330975e-05,
+      "loss": 0.5473,
+      "step": 4490
+    },
+    {
+      "epoch": 250.0,
+      "grad_norm": 1.999332316528211,
+      "learning_rate": 3.5502315596579415e-05,
+      "loss": 0.5316,
+      "step": 4500
+    },
+    {
+      "epoch": 250.55555555555554,
+      "grad_norm": 1.8153281016704603,
+      "learning_rate": 3.535678520244027e-05,
+      "loss": 0.5302,
+      "step": 4510
+    },
+    {
+      "epoch": 251.11111111111111,
+      "grad_norm": 1.9211186106819624,
+      "learning_rate": 3.521136635549066e-05,
+      "loss": 0.5254,
+      "step": 4520
+    },
+    {
+      "epoch": 251.66666666666666,
+      "grad_norm": 1.9204629812727358,
+      "learning_rate": 3.506606132656578e-05,
+      "loss": 0.5462,
+      "step": 4530
+    },
+    {
+      "epoch": 252.22222222222223,
+      "grad_norm": 2.0026110056614432,
+      "learning_rate": 3.492087238472348e-05,
+      "loss": 0.5588,
+      "step": 4540
+    },
+    {
+      "epoch": 252.77777777777777,
+      "grad_norm": 1.99163080552934,
+      "learning_rate": 3.477580179720882e-05,
+      "loss": 0.5313,
+      "step": 4550
+    },
+    {
+      "epoch": 253.33333333333334,
+      "grad_norm": 1.8248632954647312,
+      "learning_rate": 3.4630851829418646e-05,
+      "loss": 0.5397,
+      "step": 4560
+    },
+    {
+      "epoch": 253.88888888888889,
+      "grad_norm": 1.8083957374223822,
+      "learning_rate": 3.4486024744866246e-05,
+      "loss": 0.5151,
+      "step": 4570
+    },
+    {
+      "epoch": 254.44444444444446,
+      "grad_norm": 2.0930925162509406,
+      "learning_rate": 3.4341322805145965e-05,
+      "loss": 0.5404,
+      "step": 4580
+    },
+    {
+      "epoch": 255.0,
+      "grad_norm": 1.9771530415097844,
+      "learning_rate": 3.4196748269897916e-05,
+      "loss": 0.5253,
+      "step": 4590
+    },
+    {
+      "epoch": 255.55555555555554,
+      "grad_norm": 1.7955135329184546,
+      "learning_rate": 3.405230339677269e-05,
+      "loss": 0.5165,
+      "step": 4600
+    },
+    {
+      "epoch": 256.1111111111111,
+      "grad_norm": 1.825663939638291,
+      "learning_rate": 3.390799044139609e-05,
+      "loss": 0.5142,
+      "step": 4610
+    },
+    {
+      "epoch": 256.6666666666667,
+      "grad_norm": 1.9615108670009747,
+      "learning_rate": 3.376381165733392e-05,
+      "loss": 0.5251,
+      "step": 4620
+    },
+    {
+      "epoch": 257.22222222222223,
+      "grad_norm": 1.8191986605524384,
+      "learning_rate": 3.3619769296056794e-05,
+      "loss": 0.4955,
+      "step": 4630
+    },
+    {
+      "epoch": 257.77777777777777,
+      "grad_norm": 1.9375286715447235,
+      "learning_rate": 3.347586560690495e-05,
+      "loss": 0.5081,
+      "step": 4640
+    },
+    {
+      "epoch": 258.3333333333333,
+      "grad_norm": 1.892291256920902,
+      "learning_rate": 3.3332102837053155e-05,
+      "loss": 0.5325,
+      "step": 4650
+    },
+    {
+      "epoch": 258.8888888888889,
+      "grad_norm": 1.7716017981729621,
+      "learning_rate": 3.318848323147563e-05,
+      "loss": 0.5268,
+      "step": 4660
+    },
+    {
+      "epoch": 259.44444444444446,
+      "grad_norm": 2.0033654269516155,
+      "learning_rate": 3.3045009032910934e-05,
+      "loss": 0.4809,
+      "step": 4670
+    },
+    {
+      "epoch": 260.0,
+      "grad_norm": 1.8942110715271658,
+      "learning_rate": 3.2901682481827016e-05,
+      "loss": 0.4965,
+      "step": 4680
+    },
+    {
+      "epoch": 260.55555555555554,
+      "grad_norm": 1.92512008738295,
+      "learning_rate": 3.2758505816386125e-05,
+      "loss": 0.5098,
+      "step": 4690
+    },
+    {
+      "epoch": 261.1111111111111,
+      "grad_norm": 1.7377518094047486,
+      "learning_rate": 3.261548127240998e-05,
+      "loss": 0.4905,
+      "step": 4700
+    },
+    {
+      "epoch": 261.6666666666667,
+      "grad_norm": 1.7339796698279328,
+      "learning_rate": 3.247261108334476e-05,
+      "loss": 0.4969,
+      "step": 4710
+    },
+    {
+      "epoch": 262.22222222222223,
+      "grad_norm": 1.9026521593319154,
+      "learning_rate": 3.23298974802263e-05,
+      "loss": 0.5054,
+      "step": 4720
+    },
+    {
+      "epoch": 262.77777777777777,
+      "grad_norm": 1.789087420294166,
+      "learning_rate": 3.218734269164519e-05,
+      "loss": 0.4775,
+      "step": 4730
+    },
+    {
+      "epoch": 263.3333333333333,
+      "grad_norm": 1.9585802720667096,
+      "learning_rate": 3.204494894371201e-05,
+      "loss": 0.4713,
+      "step": 4740
+    },
+    {
+      "epoch": 263.8888888888889,
+      "grad_norm": 1.712930293912446,
+      "learning_rate": 3.190271846002255e-05,
+      "loss": 0.4922,
+      "step": 4750
+    },
+    {
+      "epoch": 264.44444444444446,
+      "grad_norm": 1.8520323724533305,
+      "learning_rate": 3.1760653461623104e-05,
+      "loss": 0.4702,
+      "step": 4760
+    },
+    {
+      "epoch": 265.0,
+      "grad_norm": 1.8573901698217419,
+      "learning_rate": 3.161875616697578e-05,
+      "loss": 0.4723,
+      "step": 4770
+    },
+    {
+      "epoch": 265.55555555555554,
+      "grad_norm": 1.8126847403199202,
+      "learning_rate": 3.147702879192382e-05,
+      "loss": 0.4613,
+      "step": 4780
+    },
+    {
+      "epoch": 266.1111111111111,
+      "grad_norm": 1.8133544223391205,
+      "learning_rate": 3.1335473549657086e-05,
+      "loss": 0.492,
+      "step": 4790
+    },
+    {
+      "epoch": 266.6666666666667,
+      "grad_norm": 1.8368638319238662,
+      "learning_rate": 3.11940926506774e-05,
+      "loss": 0.4765,
+      "step": 4800
+    },
+    {
+      "epoch": 267.22222222222223,
+      "grad_norm": 1.877331365130939,
+      "learning_rate": 3.105288830276408e-05,
+      "loss": 0.4825,
+      "step": 4810
+    },
+    {
+      "epoch": 267.77777777777777,
+      "grad_norm": 1.731494883490903,
+      "learning_rate": 3.091186271093947e-05,
+      "loss": 0.4627,
+      "step": 4820
+    },
+    {
+      "epoch": 268.3333333333333,
+      "grad_norm": 1.8050534756629932,
+      "learning_rate": 3.077101807743449e-05,
+      "loss": 0.4557,
+      "step": 4830
+    },
+    {
+      "epoch": 268.8888888888889,
+      "grad_norm": 1.9677512042207053,
+      "learning_rate": 3.063035660165422e-05,
+      "loss": 0.4832,
+      "step": 4840
+    },
+    {
+      "epoch": 269.44444444444446,
+      "grad_norm": 1.5978593394089409,
+      "learning_rate": 3.04898804801436e-05,
+      "loss": 0.4812,
+      "step": 4850
+    },
+    {
+      "epoch": 270.0,
+      "grad_norm": 1.7450078877705246,
+      "learning_rate": 3.0349591906553102e-05,
+      "loss": 0.4495,
+      "step": 4860
+    },
+    {
+      "epoch": 270.55555555555554,
+      "grad_norm": 1.7703660086535629,
+      "learning_rate": 3.0209493071604496e-05,
+      "loss": 0.4582,
+      "step": 4870
+    },
+    {
+      "epoch": 271.1111111111111,
+      "grad_norm": 1.709400758427168,
+      "learning_rate": 3.006958616305662e-05,
+      "loss": 0.4367,
+      "step": 4880
+    },
+    {
+      "epoch": 271.6666666666667,
+      "grad_norm": 1.6826992269520449,
+      "learning_rate": 2.9929873365671204e-05,
+      "loss": 0.4293,
+      "step": 4890
+    },
+    {
+      "epoch": 272.22222222222223,
+      "grad_norm": 1.7363741925439322,
+      "learning_rate": 2.979035686117882e-05,
+      "loss": 0.4469,
+      "step": 4900
+    },
+    {
+      "epoch": 272.77777777777777,
+      "grad_norm": 1.9413754865638262,
+      "learning_rate": 2.96510388282447e-05,
+      "loss": 0.4595,
+      "step": 4910
+    },
+    {
+      "epoch": 273.3333333333333,
+      "grad_norm": 1.9721440925106755,
+      "learning_rate": 2.951192144243485e-05,
+      "loss": 0.448,
+      "step": 4920
+    },
+    {
+      "epoch": 273.8888888888889,
+      "grad_norm": 1.9006346245032462,
+      "learning_rate": 2.937300687618194e-05,
+      "loss": 0.443,
+      "step": 4930
+    },
+    {
+      "epoch": 274.44444444444446,
+      "grad_norm": 1.9908790629859134,
+      "learning_rate": 2.9234297298751482e-05,
+      "loss": 0.4605,
+      "step": 4940
+    },
+    {
+      "epoch": 275.0,
+      "grad_norm": 1.7964383465501768,
+      "learning_rate": 2.909579487620792e-05,
+      "loss": 0.442,
+      "step": 4950
+    },
+    {
+      "epoch": 275.55555555555554,
+      "grad_norm": 1.8278326469168034,
+      "learning_rate": 2.8957501771380773e-05,
+      "loss": 0.4424,
+      "step": 4960
+    },
+    {
+      "epoch": 276.1111111111111,
+      "grad_norm": 1.8487994989935057,
+      "learning_rate": 2.8819420143830936e-05,
+      "loss": 0.4612,
+      "step": 4970
+    },
+    {
+      "epoch": 276.6666666666667,
+      "grad_norm": 1.7747553831438767,
+      "learning_rate": 2.8681552149816874e-05,
+      "loss": 0.4506,
+      "step": 4980
+    },
+    {
+      "epoch": 277.22222222222223,
+      "grad_norm": 1.657551236288529,
+      "learning_rate": 2.854389994226101e-05,
+      "loss": 0.4307,
+      "step": 4990
+    },
+    {
+      "epoch": 277.77777777777777,
+      "grad_norm": 1.6719582795409074,
+      "learning_rate": 2.840646567071608e-05,
+      "loss": 0.4188,
+      "step": 5000
+    },
+    {
+      "epoch": 278.3333333333333,
+      "grad_norm": 1.7130353077035052,
+      "learning_rate": 2.8269251481331554e-05,
+      "loss": 0.4305,
+      "step": 5010
+    },
+    {
+      "epoch": 278.8888888888889,
+      "grad_norm": 1.8767863347377998,
+      "learning_rate": 2.8132259516820147e-05,
+      "loss": 0.4219,
+      "step": 5020
+    },
+    {
+      "epoch": 279.44444444444446,
+      "grad_norm": 1.6732352596712614,
+      "learning_rate": 2.7995491916424372e-05,
+      "loss": 0.4039,
+      "step": 5030
+    },
+    {
+      "epoch": 280.0,
+      "grad_norm": 1.726188170301529,
+      "learning_rate": 2.785895081588307e-05,
+      "loss": 0.4183,
+      "step": 5040
+    },
+    {
+      "epoch": 280.55555555555554,
+      "grad_norm": 1.919993029621505,
+      "learning_rate": 2.7722638347398127e-05,
+      "loss": 0.4299,
+      "step": 5050
+    },
+    {
+      "epoch": 281.1111111111111,
+      "grad_norm": 1.7738952718096157,
+      "learning_rate": 2.7586556639601152e-05,
+      "loss": 0.4131,
+      "step": 5060
+    },
+    {
+      "epoch": 281.6666666666667,
+      "grad_norm": 1.8985495828579106,
+      "learning_rate": 2.7450707817520226e-05,
+      "loss": 0.413,
+      "step": 5070
+    },
+    {
+      "epoch": 282.22222222222223,
+      "grad_norm": 1.817220460259881,
+      "learning_rate": 2.7315094002546726e-05,
+      "loss": 0.4162,
+      "step": 5080
+    },
+    {
+      "epoch": 282.77777777777777,
+      "grad_norm": 1.6728573346414208,
+      "learning_rate": 2.71797173124022e-05,
+      "loss": 0.3956,
+      "step": 5090
+    },
+    {
+      "epoch": 283.3333333333333,
+      "grad_norm": 1.869584400586761,
+      "learning_rate": 2.7044579861105303e-05,
+      "loss": 0.4224,
+      "step": 5100
+    },
+    {
+      "epoch": 283.8888888888889,
+      "grad_norm": 1.748016118466329,
+      "learning_rate": 2.6909683758938784e-05,
+      "loss": 0.4047,
+      "step": 5110
+    },
+    {
+      "epoch": 284.44444444444446,
+      "grad_norm": 1.6965537032248714,
+      "learning_rate": 2.6775031112416504e-05,
+      "loss": 0.3919,
+      "step": 5120
+    },
+    {
+      "epoch": 285.0,
+      "grad_norm": 1.730747725687347,
+      "learning_rate": 2.6640624024250583e-05,
+      "loss": 0.3944,
+      "step": 5130
+    },
+    {
+      "epoch": 285.55555555555554,
+      "grad_norm": 1.6681225457857671,
+      "learning_rate": 2.6506464593318518e-05,
+      "loss": 0.3903,
+      "step": 5140
+    },
+    {
+      "epoch": 286.1111111111111,
+      "grad_norm": 1.7254797530952586,
+      "learning_rate": 2.6372554914630496e-05,
+      "loss": 0.4104,
+      "step": 5150
+    },
+    {
+      "epoch": 286.6666666666667,
+      "grad_norm": 1.8283878365700639,
+      "learning_rate": 2.623889707929653e-05,
+      "loss": 0.3997,
+      "step": 5160
+    },
+    {
+      "epoch": 287.22222222222223,
+      "grad_norm": 1.6117041316735492,
+      "learning_rate": 2.610549317449397e-05,
+      "loss": 0.3951,
+      "step": 5170
+    },
+    {
+      "epoch": 287.77777777777777,
+      "grad_norm": 1.635364709826563,
+      "learning_rate": 2.5972345283434773e-05,
+      "loss": 0.3939,
+      "step": 5180
+    },
+    {
+      "epoch": 288.3333333333333,
+      "grad_norm": 1.704293226596198,
+      "learning_rate": 2.5839455485333073e-05,
+      "loss": 0.3921,
+      "step": 5190
+    },
+    {
+      "epoch": 288.8888888888889,
+      "grad_norm": 1.5439506400376537,
+      "learning_rate": 2.5706825855372648e-05,
+      "loss": 0.3917,
+      "step": 5200
+    },
+    {
+      "epoch": 289.44444444444446,
+      "grad_norm": 1.6389494559667992,
+      "learning_rate": 2.557445846467452e-05,
+      "loss": 0.3856,
+      "step": 5210
+    },
+    {
+      "epoch": 290.0,
+      "grad_norm": 1.757434990929955,
+      "learning_rate": 2.5442355380264677e-05,
+      "loss": 0.4137,
+      "step": 5220
+    },
+    {
+      "epoch": 290.55555555555554,
+      "grad_norm": 1.8009151648917516,
+      "learning_rate": 2.531051866504169e-05,
+      "loss": 0.3669,
+      "step": 5230
+    },
+    {
+      "epoch": 291.1111111111111,
+      "grad_norm": 1.6171910382660568,
+      "learning_rate": 2.517895037774461e-05,
+      "loss": 0.349,
+      "step": 5240
+    },
+    {
+      "epoch": 291.6666666666667,
+      "grad_norm": 1.7828004346835449,
+      "learning_rate": 2.5047652572920708e-05,
+      "loss": 0.3738,
+      "step": 5250
+    },
+    {
+      "epoch": 292.22222222222223,
+      "grad_norm": 1.681208999304019,
+      "learning_rate": 2.4916627300893508e-05,
+      "loss": 0.4058,
+      "step": 5260
+    },
+    {
+      "epoch": 292.77777777777777,
+      "grad_norm": 1.5471473222946628,
+      "learning_rate": 2.478587660773065e-05,
+      "loss": 0.3575,
+      "step": 5270
+    },
+    {
+      "epoch": 293.3333333333333,
+      "grad_norm": 1.7092575116552515,
+      "learning_rate": 2.465540253521206e-05,
+      "loss": 0.3633,
+      "step": 5280
+    },
+    {
+      "epoch": 293.8888888888889,
+      "grad_norm": 1.7116084068038393,
+      "learning_rate": 2.452520712079795e-05,
+      "loss": 0.3909,
+      "step": 5290
+    },
+    {
+      "epoch": 294.44444444444446,
+      "grad_norm": 1.7509317642327664,
+      "learning_rate": 2.4395292397597077e-05,
+      "loss": 0.3629,
+      "step": 5300
+    },
+    {
+      "epoch": 295.0,
+      "grad_norm": 1.6378365935118524,
+      "learning_rate": 2.4265660394334987e-05,
+      "loss": 0.3688,
+      "step": 5310
+    },
+    {
+      "epoch": 295.55555555555554,
+      "grad_norm": 1.58737500817551,
+      "learning_rate": 2.413631313532229e-05,
+      "loss": 0.3447,
+      "step": 5320
+    },
+    {
+      "epoch": 296.1111111111111,
+      "grad_norm": 1.573111776751783,
+      "learning_rate": 2.4007252640423117e-05,
+      "loss": 0.3689,
+      "step": 5330
+    },
+    {
+      "epoch": 296.6666666666667,
+      "grad_norm": 1.714592917482358,
+      "learning_rate": 2.387848092502351e-05,
+      "loss": 0.3653,
+      "step": 5340
+    },
+    {
+      "epoch": 297.22222222222223,
+      "grad_norm": 1.830269232576071,
+      "learning_rate": 2.375000000000001e-05,
+      "loss": 0.3468,
+      "step": 5350
+    },
+    {
+      "epoch": 297.77777777777777,
+      "grad_norm": 1.6861487736162484,
+      "learning_rate": 2.3621811871688184e-05,
+      "loss": 0.3744,
+      "step": 5360
+    },
+    {
+      "epoch": 298.3333333333333,
+      "grad_norm": 1.6518538810774221,
+      "learning_rate": 2.3493918541851386e-05,
+      "loss": 0.3489,
+      "step": 5370
+    },
+    {
+      "epoch": 298.8888888888889,
+      "grad_norm": 1.702771981426342,
+      "learning_rate": 2.336632200764942e-05,
+      "loss": 0.3672,
+      "step": 5380
+    },
+    {
+      "epoch": 299.44444444444446,
+      "grad_norm": 1.7194161251340088,
+      "learning_rate": 2.323902426160737e-05,
+      "loss": 0.3566,
+      "step": 5390
+    },
+    {
+      "epoch": 300.0,
+      "grad_norm": 1.7696534508284532,
+      "learning_rate": 2.311202729158454e-05,
+      "loss": 0.3739,
+      "step": 5400
+    },
+    {
+      "epoch": 300.55555555555554,
+      "grad_norm": 1.575575932151432,
+      "learning_rate": 2.298533308074331e-05,
+      "loss": 0.3443,
+      "step": 5410
+    },
+    {
+      "epoch": 301.1111111111111,
+      "grad_norm": 1.7819039332336521,
+      "learning_rate": 2.2858943607518288e-05,
+      "loss": 0.3542,
+      "step": 5420
+    },
+    {
+      "epoch": 301.6666666666667,
+      "grad_norm": 1.628726647444545,
+      "learning_rate": 2.2732860845585298e-05,
+      "loss": 0.353,
+      "step": 5430
+    },
+    {
+      "epoch": 302.22222222222223,
+      "grad_norm": 1.7068181723287703,
+      "learning_rate": 2.2607086763830653e-05,
+      "loss": 0.351,
+      "step": 5440
+    },
+    {
+      "epoch": 302.77777777777777,
+      "grad_norm": 1.663300102273973,
+      "learning_rate": 2.2481623326320366e-05,
+      "loss": 0.3323,
+      "step": 5450
+    },
+    {
+      "epoch": 303.3333333333333,
+      "grad_norm": 1.7027428574513783,
+      "learning_rate": 2.235647249226944e-05,
+      "loss": 0.3327,
+      "step": 5460
+    },
+    {
+      "epoch": 303.8888888888889,
+      "grad_norm": 1.6561102178355718,
+      "learning_rate": 2.223163621601138e-05,
+      "loss": 0.33,
+      "step": 5470
+    },
+    {
+      "epoch": 304.44444444444446,
+      "grad_norm": 1.588379656758968,
+      "learning_rate": 2.210711644696756e-05,
+      "loss": 0.3364,
+      "step": 5480
+    },
+    {
+      "epoch": 305.0,
+      "grad_norm": 1.6991622740031,
+      "learning_rate": 2.1982915129616885e-05,
+      "loss": 0.318,
+      "step": 5490
+    },
+    {
+      "epoch": 305.55555555555554,
+      "grad_norm": 1.5442127476318575,
+      "learning_rate": 2.185903420346532e-05,
+      "loss": 0.3435,
+      "step": 5500
+    },
+    {
+      "epoch": 306.1111111111111,
+      "grad_norm": 1.545138376452567,
+      "learning_rate": 2.1735475603015695e-05,
+      "loss": 0.3075,
+      "step": 5510
+    },
+    {
+      "epoch": 306.6666666666667,
+      "grad_norm": 1.7974889618557228,
+      "learning_rate": 2.161224125773743e-05,
+      "loss": 0.3356,
+      "step": 5520
+    },
+    {
+      "epoch": 307.22222222222223,
+      "grad_norm": 1.6777715356905751,
+      "learning_rate": 2.1489333092036477e-05,
+      "loss": 0.3244,
+      "step": 5530
+    },
+    {
+      "epoch": 307.77777777777777,
+      "grad_norm": 1.638873082244568,
+      "learning_rate": 2.1366753025225178e-05,
+      "loss": 0.3384,
+      "step": 5540
+    },
+    {
+      "epoch": 308.3333333333333,
+      "grad_norm": 1.6234012955859096,
+      "learning_rate": 2.1244502971492338e-05,
+      "loss": 0.3427,
+      "step": 5550
+    },
+    {
+      "epoch": 308.8888888888889,
+      "grad_norm": 1.7315578779108292,
+      "learning_rate": 2.1122584839873397e-05,
+      "loss": 0.3197,
+      "step": 5560
+    },
+    {
+      "epoch": 309.44444444444446,
+      "grad_norm": 1.7133770972845965,
+      "learning_rate": 2.1001000534220486e-05,
+      "loss": 0.3181,
+      "step": 5570
+    },
+    {
+      "epoch": 310.0,
+      "grad_norm": 1.6042417120616157,
+      "learning_rate": 2.087975195317282e-05,
+      "loss": 0.3136,
+      "step": 5580
+    },
+    {
+      "epoch": 310.55555555555554,
+      "grad_norm": 1.4976318261866803,
+      "learning_rate": 2.0758840990126954e-05,
+      "loss": 0.3081,
+      "step": 5590
+    },
+    {
+      "epoch": 311.1111111111111,
+      "grad_norm": 1.710249252892409,
+      "learning_rate": 2.063826953320731e-05,
+      "loss": 0.3174,
+      "step": 5600
+    },
+    {
+      "epoch": 311.6666666666667,
+      "grad_norm": 1.6328364466321161,
+      "learning_rate": 2.0518039465236593e-05,
+      "loss": 0.3078,
+      "step": 5610
+    },
+    {
+      "epoch": 312.22222222222223,
+      "grad_norm": 1.6462596228571025,
+      "learning_rate": 2.0398152663706455e-05,
+      "loss": 0.3356,
+      "step": 5620
+    },
+    {
+      "epoch": 312.77777777777777,
+      "grad_norm": 1.717981548183592,
+      "learning_rate": 2.0278611000748184e-05,
+      "loss": 0.3037,
+      "step": 5630
+    },
+    {
+      "epoch": 313.3333333333333,
+      "grad_norm": 1.6836100354037618,
+      "learning_rate": 2.0159416343103403e-05,
+      "loss": 0.3171,
+      "step": 5640
+    },
+    {
+      "epoch": 313.8888888888889,
+      "grad_norm": 1.6817233500585649,
+      "learning_rate": 2.0040570552094998e-05,
+      "loss": 0.3093,
+      "step": 5650
+    },
+    {
+      "epoch": 314.44444444444446,
+      "grad_norm": 1.5979281989141088,
+      "learning_rate": 1.992207548359799e-05,
+      "loss": 0.3026,
+      "step": 5660
+    },
+    {
+      "epoch": 315.0,
+      "grad_norm": 1.5980868703555,
+      "learning_rate": 1.9803932988010596e-05,
+      "loss": 0.3078,
+      "step": 5670
+    },
+    {
+      "epoch": 315.55555555555554,
+      "grad_norm": 1.4747388770225092,
+      "learning_rate": 1.968614491022529e-05,
+      "loss": 0.3067,
+      "step": 5680
+    },
+    {
+      "epoch": 316.1111111111111,
+      "grad_norm": 1.7252002019390764,
+      "learning_rate": 1.956871308960006e-05,
+      "loss": 0.2932,
+      "step": 5690
+    },
+    {
+      "epoch": 316.6666666666667,
+      "grad_norm": 1.564866067438679,
+      "learning_rate": 1.945163935992961e-05,
+      "loss": 0.2886,
+      "step": 5700
+    },
+    {
+      "epoch": 317.22222222222223,
+      "grad_norm": 1.519521757356074,
+      "learning_rate": 1.933492554941676e-05,
+      "loss": 0.3101,
+      "step": 5710
+    },
+    {
+      "epoch": 317.77777777777777,
+      "grad_norm": 1.7706607809332802,
+      "learning_rate": 1.9218573480643932e-05,
+      "loss": 0.4689,
+      "step": 5720
+    },
+    {
+      "epoch": 318.3333333333333,
+      "grad_norm": 1.6041966059809718,
+      "learning_rate": 1.9102584970544597e-05,
+      "loss": 0.2891,
+      "step": 5730
+    },
+    {
+      "epoch": 318.8888888888889,
+      "grad_norm": 1.6247978818347146,
+      "learning_rate": 1.8986961830375014e-05,
+      "loss": 0.2869,
+      "step": 5740
+    },
+    {
+      "epoch": 319.44444444444446,
+      "grad_norm": 1.635980626278351,
+      "learning_rate": 1.8871705865685837e-05,
+      "loss": 0.3088,
+      "step": 5750
+    },
+    {
+      "epoch": 320.0,
+      "grad_norm": 1.4910524854933342,
+      "learning_rate": 1.875681887629402e-05,
+      "loss": 0.272,
+      "step": 5760
+    },
+    {
+      "epoch": 320.55555555555554,
+      "grad_norm": 1.5261655199569344,
+      "learning_rate": 1.864230265625464e-05,
+      "loss": 0.3051,
+      "step": 5770
+    },
+    {
+      "epoch": 321.1111111111111,
+      "grad_norm": 1.5770760250888072,
+      "learning_rate": 1.852815899383288e-05,
+      "loss": 0.2769,
+      "step": 5780
+    },
+    {
+      "epoch": 321.6666666666667,
+      "grad_norm": 1.5462083246879608,
+      "learning_rate": 1.8414389671476188e-05,
+      "loss": 0.2612,
+      "step": 5790
+    },
+    {
+      "epoch": 322.22222222222223,
+      "grad_norm": 1.4550014398836437,
+      "learning_rate": 1.8300996465786317e-05,
+      "loss": 0.2826,
+      "step": 5800
+    },
+    {
+      "epoch": 322.77777777777777,
+      "grad_norm": 1.5356390972678102,
+      "learning_rate": 1.8187981147491703e-05,
+      "loss": 0.2836,
+      "step": 5810
+    },
+    {
+      "epoch": 323.3333333333333,
+      "grad_norm": 1.5622364584875694,
+      "learning_rate": 1.8075345481419723e-05,
+      "loss": 0.273,
+      "step": 5820
+    },
+    {
+      "epoch": 323.8888888888889,
+      "grad_norm": 1.6070486495244867,
+      "learning_rate": 1.7963091226469206e-05,
+      "loss": 0.294,
+      "step": 5830
+    },
+    {
+      "epoch": 324.44444444444446,
+      "grad_norm": 1.514112760262724,
+      "learning_rate": 1.7851220135582877e-05,
+      "loss": 0.2678,
+      "step": 5840
+    },
+    {
+      "epoch": 325.0,
+      "grad_norm": 1.5991453958077055,
+      "learning_rate": 1.773973395572011e-05,
+      "loss": 0.3008,
+      "step": 5850
+    },
+    {
+      "epoch": 325.55555555555554,
+      "grad_norm": 1.6191307592148916,
+      "learning_rate": 1.7628634427829513e-05,
+      "loss": 0.2839,
+      "step": 5860
+    },
+    {
+      "epoch": 326.1111111111111,
+      "grad_norm": 1.6197113393346252,
+      "learning_rate": 1.751792328682183e-05,
+      "loss": 0.2795,
+      "step": 5870
+    },
+    {
+      "epoch": 326.6666666666667,
+      "grad_norm": 1.5467025198288682,
+      "learning_rate": 1.740760226154283e-05,
+      "loss": 0.2795,
+      "step": 5880
+    },
+    {
+      "epoch": 327.22222222222223,
+      "grad_norm": 1.515873524847954,
+      "learning_rate": 1.7297673074746285e-05,
+      "loss": 0.2707,
+      "step": 5890
+    },
+    {
+      "epoch": 327.77777777777777,
+      "grad_norm": 1.4903329239203882,
+      "learning_rate": 1.7188137443067116e-05,
+      "loss": 0.276,
+      "step": 5900
+    },
+    {
+      "epoch": 328.3333333333333,
+      "grad_norm": 1.521596333959397,
+      "learning_rate": 1.707899707699451e-05,
+      "loss": 0.2687,
+      "step": 5910
+    },
+    {
+      "epoch": 328.8888888888889,
+      "grad_norm": 1.4970754087078948,
+      "learning_rate": 1.697025368084532e-05,
+      "loss": 0.2607,
+      "step": 5920
+    },
+    {
+      "epoch": 329.44444444444446,
+      "grad_norm": 1.5994070295411453,
+      "learning_rate": 1.686190895273733e-05,
+      "loss": 0.2595,
+      "step": 5930
+    },
+    {
+      "epoch": 330.0,
+      "grad_norm": 1.6113486918668491,
+      "learning_rate": 1.6753964584562798e-05,
+      "loss": 0.2738,
+      "step": 5940
+    },
+    {
+      "epoch": 330.55555555555554,
+      "grad_norm": 1.6053156350439173,
+      "learning_rate": 1.6646422261962067e-05,
+      "loss": 0.2695,
+      "step": 5950
+    },
+    {
+      "epoch": 331.1111111111111,
+      "grad_norm": 1.4206331300314035,
+      "learning_rate": 1.653928366429717e-05,
+      "loss": 0.2584,
+      "step": 5960
+    },
+    {
+      "epoch": 331.6666666666667,
+      "grad_norm": 1.5944098060484366,
+      "learning_rate": 1.643255046462568e-05,
+      "loss": 0.2623,
+      "step": 5970
+    },
+    {
+      "epoch": 332.22222222222223,
+      "grad_norm": 1.4569183387676203,
+      "learning_rate": 1.6326224329674502e-05,
+      "loss": 0.2556,
+      "step": 5980
+    },
+    {
+      "epoch": 332.77777777777777,
+      "grad_norm": 1.3932152958354738,
+      "learning_rate": 1.6220306919813932e-05,
+      "loss": 0.2467,
+      "step": 5990
+    },
+    {
+      "epoch": 333.3333333333333,
+      "grad_norm": 1.66782966409092,
+      "learning_rate": 1.611479988903165e-05,
+      "loss": 0.2607,
+      "step": 6000
+    },
+    {
+      "epoch": 333.8888888888889,
+      "grad_norm": 1.527440607068153,
+      "learning_rate": 1.6009704884906963e-05,
+      "loss": 0.2742,
+      "step": 6010
+    },
+    {
+      "epoch": 334.44444444444446,
+      "grad_norm": 1.48071641284747,
+      "learning_rate": 1.5905023548585002e-05,
+      "loss": 0.25,
+      "step": 6020
+    },
+    {
+      "epoch": 335.0,
+      "grad_norm": 1.355614646073615,
+      "learning_rate": 1.580075751475117e-05,
+      "loss": 0.2502,
+      "step": 6030
+    },
+    {
+      "epoch": 335.55555555555554,
+      "grad_norm": 1.6116932588189543,
+      "learning_rate": 1.5696908411605575e-05,
+      "loss": 0.2524,
+      "step": 6040
+    },
+    {
+      "epoch": 336.1111111111111,
+      "grad_norm": 1.3676253135491818,
+      "learning_rate": 1.5593477860837575e-05,
+      "loss": 0.2362,
+      "step": 6050
+    },
+    {
+      "epoch": 336.6666666666667,
+      "grad_norm": 1.5130850361931774,
+      "learning_rate": 1.5490467477600537e-05,
+      "loss": 0.2424,
+      "step": 6060
+    },
+    {
+      "epoch": 337.22222222222223,
+      "grad_norm": 1.529778766344778,
+      "learning_rate": 1.5387878870486533e-05,
+      "loss": 0.2549,
+      "step": 6070
+    },
+    {
+      "epoch": 337.77777777777777,
+      "grad_norm": 1.6029099169553414,
+      "learning_rate": 1.5285713641501237e-05,
+      "loss": 0.2558,
+      "step": 6080
+    },
+    {
+      "epoch": 338.3333333333333,
+      "grad_norm": 1.4917039337333422,
+      "learning_rate": 1.518397338603898e-05,
+      "loss": 0.2499,
+      "step": 6090
+    },
+    {
+      "epoch": 338.8888888888889,
+      "grad_norm": 1.4472609299803654,
+      "learning_rate": 1.508265969285772e-05,
+      "loss": 0.2536,
+      "step": 6100
+    },
+    {
+      "epoch": 339.44444444444446,
+      "grad_norm": 1.478932330772614,
+      "learning_rate": 1.4981774144054346e-05,
+      "loss": 0.2399,
+      "step": 6110
+    },
+    {
+      "epoch": 340.0,
+      "grad_norm": 1.6033980109214936,
+      "learning_rate": 1.4881318315039875e-05,
+      "loss": 0.2445,
+      "step": 6120
+    },
+    {
+      "epoch": 340.55555555555554,
+      "grad_norm": 1.4854469845510283,
+      "learning_rate": 1.4781293774514936e-05,
+      "loss": 0.2601,
+      "step": 6130
+    },
+    {
+      "epoch": 341.1111111111111,
+      "grad_norm": 1.5523210467217587,
+      "learning_rate": 1.4681702084445201e-05,
+      "loss": 0.2439,
+      "step": 6140
+    },
+    {
+      "epoch": 341.6666666666667,
+      "grad_norm": 1.476830352224557,
+      "learning_rate": 1.458254480003702e-05,
+      "loss": 0.2443,
+      "step": 6150
+    },
+    {
+      "epoch": 342.22222222222223,
+      "grad_norm": 1.4979671372500152,
+      "learning_rate": 1.448382346971319e-05,
+      "loss": 0.2361,
+      "step": 6160
+    },
+    {
+      "epoch": 342.77777777777777,
+      "grad_norm": 1.4490401576962362,
+      "learning_rate": 1.438553963508866e-05,
+      "loss": 0.233,
+      "step": 6170
+    },
+    {
+      "epoch": 343.3333333333333,
+      "grad_norm": 1.3369227563033388,
+      "learning_rate": 1.428769483094658e-05,
+      "loss": 0.2326,
+      "step": 6180
+    },
+    {
+      "epoch": 343.8888888888889,
+      "grad_norm": 1.3843625965811364,
+      "learning_rate": 1.4190290585214231e-05,
+      "loss": 0.2333,
+      "step": 6190
+    },
+    {
+      "epoch": 344.44444444444446,
+      "grad_norm": 1.4021204183509928,
+      "learning_rate": 1.4093328418939246e-05,
+      "loss": 0.228,
+      "step": 6200
+    },
+    {
+      "epoch": 345.0,
+      "grad_norm": 1.4793020195439657,
+      "learning_rate": 1.3996809846265802e-05,
+      "loss": 0.2451,
+      "step": 6210
+    },
+    {
+      "epoch": 345.55555555555554,
+      "grad_norm": 1.5847783103925988,
+      "learning_rate": 1.390073637441101e-05,
+      "loss": 0.2188,
+      "step": 6220
+    },
+    {
+      "epoch": 346.1111111111111,
+      "grad_norm": 1.3926136927188861,
+      "learning_rate": 1.3805109503641356e-05,
+      "loss": 0.2212,
+      "step": 6230
+    },
+    {
+      "epoch": 346.6666666666667,
+      "grad_norm": 1.5906497496232672,
+      "learning_rate": 1.3709930727249264e-05,
+      "loss": 0.2261,
+      "step": 6240
+    },
+    {
+      "epoch": 347.22222222222223,
+      "grad_norm": 1.5124997036515884,
+      "learning_rate": 1.3615201531529845e-05,
+      "loss": 0.2331,
+      "step": 6250
+    },
+    {
+      "epoch": 347.77777777777777,
+      "grad_norm": 1.4372448902155415,
+      "learning_rate": 1.3520923395757577e-05,
+      "loss": 0.2114,
+      "step": 6260
+    },
+    {
+      "epoch": 348.3333333333333,
+      "grad_norm": 1.688693083983564,
+      "learning_rate": 1.342709779216332e-05,
+      "loss": 0.2217,
+      "step": 6270
+    },
+    {
+      "epoch": 348.8888888888889,
+      "grad_norm": 1.5288876586469855,
+      "learning_rate": 1.3333726185911221e-05,
+      "loss": 0.2133,
+      "step": 6280
+    },
+    {
+      "epoch": 349.44444444444446,
+      "grad_norm": 1.4846155875134777,
+      "learning_rate": 1.3240810035075934e-05,
+      "loss": 0.2237,
+      "step": 6290
+    },
+    {
+      "epoch": 350.0,
+      "grad_norm": 1.536022378743584,
+      "learning_rate": 1.3148350790619758e-05,
+      "loss": 0.2295,
+      "step": 6300
+    },
+    {
+      "epoch": 350.55555555555554,
+      "grad_norm": 1.5002242556465433,
+      "learning_rate": 1.3056349896370031e-05,
+      "loss": 0.2253,
+      "step": 6310
+    },
+    {
+      "epoch": 351.1111111111111,
+      "grad_norm": 1.3892155480361335,
+      "learning_rate": 1.2964808788996606e-05,
+      "loss": 0.1989,
+      "step": 6320
+    },
+    {
+      "epoch": 351.6666666666667,
+      "grad_norm": 1.5382358247899561,
+      "learning_rate": 1.2873728897989336e-05,
+      "loss": 0.2212,
+      "step": 6330
+    },
+    {
+      "epoch": 352.22222222222223,
+      "grad_norm": 1.5213566580802824,
+      "learning_rate": 1.2783111645635851e-05,
+      "loss": 0.215,
+      "step": 6340
+    },
+    {
+      "epoch": 352.77777777777777,
+      "grad_norm": 1.2976490606192654,
+      "learning_rate": 1.2692958446999248e-05,
+      "loss": 0.2043,
+      "step": 6350
+    },
+    {
+      "epoch": 353.3333333333333,
+      "grad_norm": 1.4688506396288514,
+      "learning_rate": 1.2603270709896092e-05,
+      "loss": 0.2181,
+      "step": 6360
+    },
+    {
+      "epoch": 353.8888888888889,
+      "grad_norm": 1.4596436018520036,
+      "learning_rate": 1.2514049834874331e-05,
+      "loss": 0.2196,
+      "step": 6370
+    },
+    {
+      "epoch": 354.44444444444446,
+      "grad_norm": 1.4639728856624477,
+      "learning_rate": 1.2425297215191518e-05,
+      "loss": 0.2065,
+      "step": 6380
+    },
+    {
+      "epoch": 355.0,
+      "grad_norm": 1.3436064532589198,
+      "learning_rate": 1.2337014236793001e-05,
+      "loss": 0.2056,
+      "step": 6390
+    },
+    {
+      "epoch": 355.55555555555554,
+      "grad_norm": 1.491233800526682,
+      "learning_rate": 1.2249202278290264e-05,
+      "loss": 0.2186,
+      "step": 6400
+    },
+    {
+      "epoch": 356.1111111111111,
+      "grad_norm": 1.4555700916958962,
+      "learning_rate": 1.2161862710939476e-05,
+      "loss": 0.2236,
+      "step": 6410
+    },
+    {
+      "epoch": 356.6666666666667,
+      "grad_norm": 1.4434458457637784,
+      "learning_rate": 1.2074996898619982e-05,
+      "loss": 0.2173,
+      "step": 6420
+    },
+    {
+      "epoch": 357.22222222222223,
+      "grad_norm": 1.4144644824490946,
+      "learning_rate": 1.1988606197813097e-05,
+      "loss": 0.2109,
+      "step": 6430
+    },
+    {
+      "epoch": 357.77777777777777,
+      "grad_norm": 1.4480494797351051,
+      "learning_rate": 1.1902691957580829e-05,
+      "loss": 0.2119,
+      "step": 6440
+    },
+    {
+      "epoch": 358.3333333333333,
+      "grad_norm": 1.3850541991084966,
+      "learning_rate": 1.181725551954491e-05,
+      "loss": 0.1865,
+      "step": 6450
+    },
+    {
+      "epoch": 358.8888888888889,
+      "grad_norm": 1.263911889231421,
+      "learning_rate": 1.1732298217865753e-05,
+      "loss": 0.195,
+      "step": 6460
+    },
+    {
+      "epoch": 359.44444444444446,
+      "grad_norm": 1.4199111287420036,
+      "learning_rate": 1.1647821379221695e-05,
+      "loss": 0.1975,
+      "step": 6470
+    },
+    {
+      "epoch": 360.0,
+      "grad_norm": 1.4564191335274619,
+      "learning_rate": 1.1563826322788229e-05,
+      "loss": 0.1969,
+      "step": 6480
+    },
+    {
+      "epoch": 360.55555555555554,
+      "grad_norm": 1.4200044001927503,
+      "learning_rate": 1.148031436021742e-05,
+      "loss": 0.2042,
+      "step": 6490
+    },
+    {
+      "epoch": 361.1111111111111,
+      "grad_norm": 1.4961774120394677,
+      "learning_rate": 1.1397286795617434e-05,
+      "loss": 0.1999,
+      "step": 6500
+    },
+    {
+      "epoch": 361.6666666666667,
+      "grad_norm": 1.2845174947061786,
+      "learning_rate": 1.1314744925532153e-05,
+      "loss": 0.2036,
+      "step": 6510
+    },
+    {
+      "epoch": 362.22222222222223,
+      "grad_norm": 1.4317964102175593,
+      "learning_rate": 1.123269003892096e-05,
+      "loss": 0.1973,
+      "step": 6520
+    },
+    {
+      "epoch": 362.77777777777777,
+      "grad_norm": 1.583126866113261,
+      "learning_rate": 1.1151123417138555e-05,
+      "loss": 0.1972,
+      "step": 6530
+    },
+    {
+      "epoch": 363.3333333333333,
+      "grad_norm": 1.527567503166504,
+      "learning_rate": 1.1070046333915014e-05,
+      "loss": 0.1862,
+      "step": 6540
+    },
+    {
+      "epoch": 363.8888888888889,
+      "grad_norm": 1.40081278504311,
+      "learning_rate": 1.0989460055335845e-05,
+      "loss": 0.1899,
+      "step": 6550
+    },
+    {
+      "epoch": 364.44444444444446,
+      "grad_norm": 1.343806332139319,
+      "learning_rate": 1.0909365839822227e-05,
+      "loss": 0.1884,
+      "step": 6560
+    },
+    {
+      "epoch": 365.0,
+      "grad_norm": 1.2712546051602143,
+      "learning_rate": 1.0829764938111404e-05,
+      "loss": 0.19,
+      "step": 6570
+    },
+    {
+      "epoch": 365.55555555555554,
+      "grad_norm": 1.2246343436730125,
+      "learning_rate": 1.0750658593237084e-05,
+      "loss": 0.1871,
+      "step": 6580
+    },
+    {
+      "epoch": 366.1111111111111,
+      "grad_norm": 1.3891773189295686,
+      "learning_rate": 1.0672048040510083e-05,
+      "loss": 0.1834,
+      "step": 6590
+    },
+    {
+      "epoch": 366.6666666666667,
+      "grad_norm": 1.2760318937381523,
+      "learning_rate": 1.0593934507498998e-05,
+      "loss": 0.1941,
+      "step": 6600
+    },
+    {
+      "epoch": 367.22222222222223,
+      "grad_norm": 1.4709586723170276,
+      "learning_rate": 1.051631921401107e-05,
+      "loss": 0.1985,
+      "step": 6610
+    },
+    {
+      "epoch": 367.77777777777777,
+      "grad_norm": 1.4455331170040162,
+      "learning_rate": 1.0439203372073096e-05,
+      "loss": 0.2012,
+      "step": 6620
+    },
+    {
+      "epoch": 368.3333333333333,
+      "grad_norm": 1.3604688080419298,
+      "learning_rate": 1.0362588185912561e-05,
+      "loss": 0.2038,
+      "step": 6630
+    },
+    {
+      "epoch": 368.8888888888889,
+      "grad_norm": 1.3344901649395677,
+      "learning_rate": 1.0286474851938762e-05,
+      "loss": 0.1796,
+      "step": 6640
+    },
+    {
+      "epoch": 369.44444444444446,
+      "grad_norm": 1.3985403864855765,
+      "learning_rate": 1.0210864558724168e-05,
+      "loss": 0.1925,
+      "step": 6650
+    },
+    {
+      "epoch": 370.0,
+      "grad_norm": 1.3461247241883973,
+      "learning_rate": 1.013575848698587e-05,
+      "loss": 0.1734,
+      "step": 6660
+    },
+    {
+      "epoch": 370.55555555555554,
+      "grad_norm": 1.411205503710029,
+      "learning_rate": 1.0061157809567101e-05,
+      "loss": 0.1838,
+      "step": 6670
+    },
+    {
+      "epoch": 371.1111111111111,
+      "grad_norm": 1.355532246182537,
+      "learning_rate": 9.987063691418975e-06,
+      "loss": 0.1876,
+      "step": 6680
+    },
+    {
+      "epoch": 371.6666666666667,
+      "grad_norm": 1.3387069105378433,
+      "learning_rate": 9.91347728958222e-06,
+      "loss": 0.1845,
+      "step": 6690
+    },
+    {
+      "epoch": 372.22222222222223,
+      "grad_norm": 1.282925045824001,
+      "learning_rate": 9.840399753169205e-06,
+      "loss": 0.189,
+      "step": 6700
+    },
+    {
+      "epoch": 372.77777777777777,
+      "grad_norm": 1.234313335568503,
+      "learning_rate": 9.767832223345916e-06,
+      "loss": 0.1681,
+      "step": 6710
+    },
+    {
+      "epoch": 373.3333333333333,
+      "grad_norm": 1.298997543872981,
+      "learning_rate": 9.695775833314159e-06,
+      "loss": 0.1812,
+      "step": 6720
+    },
+    {
+      "epoch": 373.8888888888889,
+      "grad_norm": 1.3175163229667686,
+      "learning_rate": 9.624231708293902e-06,
+      "loss": 0.181,
+      "step": 6730
+    },
+    {
+      "epoch": 374.44444444444446,
+      "grad_norm": 1.4615845904619011,
+      "learning_rate": 9.553200965505643e-06,
+      "loss": 0.1837,
+      "step": 6740
+    },
+    {
+      "epoch": 375.0,
+      "grad_norm": 1.2509714166682806,
+      "learning_rate": 9.482684714153013e-06,
+      "loss": 0.1817,
+      "step": 6750
+    },
+    {
+      "epoch": 375.55555555555554,
+      "grad_norm": 1.3166347045077218,
+      "learning_rate": 9.41268405540541e-06,
+      "loss": 0.1667,
+      "step": 6760
+    },
+    {
+      "epoch": 376.1111111111111,
+      "grad_norm": 1.4637728833137198,
+      "learning_rate": 9.343200082380868e-06,
+      "loss": 0.1711,
+      "step": 6770
+    },
+    {
+      "epoch": 376.6666666666667,
+      "grad_norm": 1.3398153229747896,
+      "learning_rate": 9.274233880128898e-06,
+      "loss": 0.1684,
+      "step": 6780
+    },
+    {
+      "epoch": 377.22222222222223,
+      "grad_norm": 1.2851195634515988,
+      "learning_rate": 9.205786525613647e-06,
+      "loss": 0.1722,
+      "step": 6790
+    },
+    {
+      "epoch": 377.77777777777777,
+      "grad_norm": 1.3555541877314088,
+      "learning_rate": 9.13785908769698e-06,
+      "loss": 0.1726,
+      "step": 6800
+    },
+    {
+      "epoch": 378.3333333333333,
+      "grad_norm": 1.3121554058571638,
+      "learning_rate": 9.070452627121857e-06,
+      "loss": 0.1721,
+      "step": 6810
+    },
+    {
+      "epoch": 378.8888888888889,
+      "grad_norm": 1.2695651475342327,
+      "learning_rate": 9.00356819649575e-06,
+      "loss": 0.1657,
+      "step": 6820
+    },
+    {
+      "epoch": 379.44444444444446,
+      "grad_norm": 1.4500831021139786,
+      "learning_rate": 8.937206840274185e-06,
+      "loss": 0.1723,
+      "step": 6830
+    },
+    {
+      "epoch": 380.0,
+      "grad_norm": 1.4466370982270138,
+      "learning_rate": 8.871369594744485e-06,
+      "loss": 0.1736,
+      "step": 6840
+    },
+    {
+      "epoch": 380.55555555555554,
+      "grad_norm": 1.3471304047267185,
+      "learning_rate": 8.806057488009507e-06,
+      "loss": 0.1813,
+      "step": 6850
+    },
+    {
+      "epoch": 381.1111111111111,
+      "grad_norm": 1.2472586134878676,
+      "learning_rate": 8.741271539971674e-06,
+      "loss": 0.1612,
+      "step": 6860
+    },
+    {
+      "epoch": 381.6666666666667,
+      "grad_norm": 1.3401508035067664,
+      "learning_rate": 8.677012762316979e-06,
+      "loss": 0.1592,
+      "step": 6870
+    },
+    {
+      "epoch": 382.22222222222223,
+      "grad_norm": 1.3294691576699622,
+      "learning_rate": 8.613282158499221e-06,
+      "loss": 0.1765,
+      "step": 6880
+    },
+    {
+      "epoch": 382.77777777777777,
+      "grad_norm": 1.4002583333499454,
+      "learning_rate": 8.550080723724344e-06,
+      "loss": 0.1614,
+      "step": 6890
+    },
+    {
+      "epoch": 383.3333333333333,
+      "grad_norm": 1.2234101332603566,
+      "learning_rate": 8.487409444934853e-06,
+      "loss": 0.1685,
+      "step": 6900
+    },
+    {
+      "epoch": 383.8888888888889,
+      "grad_norm": 1.3103819514509327,
+      "learning_rate": 8.425269300794467e-06,
+      "loss": 0.1685,
+      "step": 6910
+    },
+    {
+      "epoch": 384.44444444444446,
+      "grad_norm": 1.4041337938545284,
+      "learning_rate": 8.363661261672756e-06,
+      "loss": 0.1571,
+      "step": 6920
+    },
+    {
+      "epoch": 385.0,
+      "grad_norm": 1.2768979475785343,
+      "learning_rate": 8.302586289630078e-06,
+      "loss": 0.1514,
+      "step": 6930
+    },
+    {
+      "epoch": 385.55555555555554,
+      "grad_norm": 1.4770726528597609,
+      "learning_rate": 8.242045338402464e-06,
+      "loss": 0.1696,
+      "step": 6940
+    },
+    {
+      "epoch": 386.1111111111111,
+      "grad_norm": 1.3486631750491231,
+      "learning_rate": 8.182039353386808e-06,
+      "loss": 0.1626,
+      "step": 6950
+    },
+    {
+      "epoch": 386.6666666666667,
+      "grad_norm": 1.3250615861363344,
+      "learning_rate": 8.122569271626042e-06,
+      "loss": 0.1573,
+      "step": 6960
+    },
+    {
+      "epoch": 387.22222222222223,
+      "grad_norm": 1.3243012529176694,
+      "learning_rate": 8.063636021794522e-06,
+      "loss": 0.1483,
+      "step": 6970
+    },
+    {
+      "epoch": 387.77777777777777,
+      "grad_norm": 1.5110626768471902,
+      "learning_rate": 8.005240524183561e-06,
+      "loss": 0.1641,
+      "step": 6980
+    },
+    {
+      "epoch": 388.3333333333333,
+      "grad_norm": 1.3952905633959831,
+      "learning_rate": 7.94738369068699e-06,
+      "loss": 0.1577,
+      "step": 6990
+    },
+    {
+      "epoch": 388.8888888888889,
+      "grad_norm": 1.3705054431126644,
+      "learning_rate": 7.890066424786995e-06,
+      "loss": 0.1619,
+      "step": 7000
+    },
+    {
+      "epoch": 389.44444444444446,
+      "grad_norm": 1.3773523928695124,
+      "learning_rate": 7.833289621539925e-06,
+      "loss": 0.1573,
+      "step": 7010
+    },
+    {
+      "epoch": 390.0,
+      "grad_norm": 1.3147191859607699,
+      "learning_rate": 7.777054167562405e-06,
+      "loss": 0.1546,
+      "step": 7020
+    },
+    {
+      "epoch": 390.55555555555554,
+      "grad_norm": 1.2858794707951735,
+      "learning_rate": 7.721360941017406e-06,
+      "loss": 0.1472,
+      "step": 7030
+    },
+    {
+      "epoch": 391.1111111111111,
+      "grad_norm": 1.1899738142995278,
+      "learning_rate": 7.66621081160059e-06,
+      "loss": 0.1613,
+      "step": 7040
+    },
+    {
+      "epoch": 391.6666666666667,
+      "grad_norm": 1.2314832108282845,
+      "learning_rate": 7.611604640526709e-06,
+      "loss": 0.1621,
+      "step": 7050
+    },
+    {
+      "epoch": 392.22222222222223,
+      "grad_norm": 1.3272860850740373,
+      "learning_rate": 7.557543280516148e-06,
+      "loss": 0.1507,
+      "step": 7060
+    },
+    {
+      "epoch": 392.77777777777777,
+      "grad_norm": 1.1510539449782005,
+      "learning_rate": 7.50402757578163e-06,
+      "loss": 0.1574,
+      "step": 7070
+    },
+    {
+      "epoch": 393.3333333333333,
+      "grad_norm": 1.214961889477732,
+      "learning_rate": 7.451058362015012e-06,
+      "loss": 0.1458,
+      "step": 7080
+    },
+    {
+      "epoch": 393.8888888888889,
+      "grad_norm": 1.224636679897074,
+      "learning_rate": 7.398636466374252e-06,
+      "loss": 0.1427,
+      "step": 7090
+    },
+    {
+      "epoch": 394.44444444444446,
+      "grad_norm": 1.4017386233305398,
+      "learning_rate": 7.3467627074704676e-06,
+      "loss": 0.1672,
+      "step": 7100
+    },
+    {
+      "epoch": 395.0,
+      "grad_norm": 1.211167295248333,
+      "learning_rate": 7.295437895355191e-06,
+      "loss": 0.1421,
+      "step": 7110
+    },
+    {
+      "epoch": 395.55555555555554,
+      "grad_norm": 1.1544847674349978,
+      "learning_rate": 7.244662831507682e-06,
+      "loss": 0.1446,
+      "step": 7120
+    },
+    {
+      "epoch": 396.1111111111111,
+      "grad_norm": 1.3057864009860112,
+      "learning_rate": 7.194438308822426e-06,
+      "loss": 0.1557,
+      "step": 7130
+    },
+    {
+      "epoch": 396.6666666666667,
+      "grad_norm": 1.208871863927433,
+      "learning_rate": 7.144765111596779e-06,
+      "loss": 0.1348,
+      "step": 7140
+    },
+    {
+      "epoch": 397.22222222222223,
+      "grad_norm": 1.3242932414030395,
+      "learning_rate": 7.095644015518659e-06,
+      "loss": 0.1554,
+      "step": 7150
+    },
+    {
+      "epoch": 397.77777777777777,
+      "grad_norm": 1.075012579555985,
+      "learning_rate": 7.0470757876545026e-06,
+      "loss": 0.1544,
+      "step": 7160
+    },
+    {
+      "epoch": 398.3333333333333,
+      "grad_norm": 1.2954913191922273,
+      "learning_rate": 6.999061186437224e-06,
+      "loss": 0.1502,
+      "step": 7170
+    },
+    {
+      "epoch": 398.8888888888889,
+      "grad_norm": 1.3649822215085752,
+      "learning_rate": 6.951600961654432e-06,
+      "loss": 0.1396,
+      "step": 7180
+    },
+    {
+      "epoch": 399.44444444444446,
+      "grad_norm": 1.3632436927497,
+      "learning_rate": 6.9046958544366604e-06,
+      "loss": 0.1437,
+      "step": 7190
+    },
+    {
+      "epoch": 400.0,
+      "grad_norm": 1.2172988174718529,
+      "learning_rate": 6.8583465972458365e-06,
+      "loss": 0.1384,
+      "step": 7200
+    },
+    {
+      "epoch": 400.55555555555554,
+      "grad_norm": 1.151434588218825,
+      "learning_rate": 6.81255391386384e-06,
+      "loss": 0.1461,
+      "step": 7210
+    },
+    {
+      "epoch": 401.1111111111111,
+      "grad_norm": 1.155787401191997,
+      "learning_rate": 6.767318519381177e-06,
+      "loss": 0.147,
+      "step": 7220
+    },
+    {
+      "epoch": 401.6666666666667,
+      "grad_norm": 1.2304794553260745,
+      "learning_rate": 6.722641120185843e-06,
+      "loss": 0.1404,
+      "step": 7230
+    },
+    {
+      "epoch": 402.22222222222223,
+      "grad_norm": 1.1783220080884016,
+      "learning_rate": 6.678522413952258e-06,
+      "loss": 0.1315,
+      "step": 7240
+    },
+    {
+      "epoch": 402.77777777777777,
+      "grad_norm": 1.3233647926652508,
+      "learning_rate": 6.63496308963041e-06,
+      "loss": 0.1482,
+      "step": 7250
+    },
+    {
+      "epoch": 403.3333333333333,
+      "grad_norm": 1.2916709171758467,
+      "learning_rate": 6.5919638274350614e-06,
+      "loss": 0.1304,
+      "step": 7260
+    },
+    {
+      "epoch": 403.8888888888889,
+      "grad_norm": 1.2753555737012041,
+      "learning_rate": 6.549525298835157e-06,
+      "loss": 0.1373,
+      "step": 7270
+    },
+    {
+      "epoch": 404.44444444444446,
+      "grad_norm": 1.2444687534883363,
+      "learning_rate": 6.50764816654331e-06,
+      "loss": 0.1346,
+      "step": 7280
+    },
+    {
+      "epoch": 405.0,
+      "grad_norm": 1.3409327760611869,
+      "learning_rate": 6.4663330845054775e-06,
+      "loss": 0.1395,
+      "step": 7290
+    },
+    {
+      "epoch": 405.55555555555554,
+      "grad_norm": 1.325656481723797,
+      "learning_rate": 6.425580697890733e-06,
+      "loss": 0.1404,
+      "step": 7300
+    },
+    {
+      "epoch": 406.1111111111111,
+      "grad_norm": 1.1924410469149982,
+      "learning_rate": 6.385391643081197e-06,
+      "loss": 0.1305,
+      "step": 7310
+    },
+    {
+      "epoch": 406.6666666666667,
+      "grad_norm": 1.2496581563821765,
+      "learning_rate": 6.34576654766211e-06,
+      "loss": 0.1351,
+      "step": 7320
+    },
+    {
+      "epoch": 407.22222222222223,
+      "grad_norm": 1.3387201786370682,
+      "learning_rate": 6.306706030412002e-06,
+      "loss": 0.1312,
+      "step": 7330
+    },
+    {
+      "epoch": 407.77777777777777,
+      "grad_norm": 1.293492191077314,
+      "learning_rate": 6.268210701293069e-06,
+      "loss": 0.1483,
+      "step": 7340
+    },
+    {
+      "epoch": 408.3333333333333,
+      "grad_norm": 1.1534449601222827,
+      "learning_rate": 6.23028116144161e-06,
+      "loss": 0.1384,
+      "step": 7350
+    },
+    {
+      "epoch": 408.8888888888889,
+      "grad_norm": 1.2072350289875367,
+      "learning_rate": 6.192918003158677e-06,
+      "loss": 0.1312,
+      "step": 7360
+    },
+    {
+      "epoch": 409.44444444444446,
+      "grad_norm": 1.3364439779428876,
+      "learning_rate": 6.156121809900792e-06,
+      "loss": 0.1363,
+      "step": 7370
+    },
+    {
+      "epoch": 410.0,
+      "grad_norm": 1.284655487916259,
+      "learning_rate": 6.1198931562708504e-06,
+      "loss": 0.1367,
+      "step": 7380
+    },
+    {
+      "epoch": 410.55555555555554,
+      "grad_norm": 1.132283843761233,
+      "learning_rate": 6.084232608009166e-06,
+      "loss": 0.1332,
+      "step": 7390
+    },
+    {
+      "epoch": 411.1111111111111,
+      "grad_norm": 1.130339560676534,
+      "learning_rate": 6.049140721984598e-06,
+      "loss": 0.1303,
+      "step": 7400
+    },
+    {
+      "epoch": 411.6666666666667,
+      "grad_norm": 1.3367118153038267,
+      "learning_rate": 6.014618046185887e-06,
+      "loss": 0.1388,
+      "step": 7410
+    },
+    {
+      "epoch": 412.22222222222223,
+      "grad_norm": 1.3591909612842323,
+      "learning_rate": 5.980665119713077e-06,
+      "loss": 0.1337,
+      "step": 7420
+    },
+    {
+      "epoch": 412.77777777777777,
+      "grad_norm": 1.1582848454122034,
+      "learning_rate": 5.947282472769127e-06,
+      "loss": 0.1348,
+      "step": 7430
+    },
+    {
+      "epoch": 413.3333333333333,
+      "grad_norm": 1.290043128423759,
+      "learning_rate": 5.914470626651588e-06,
+      "loss": 0.1307,
+      "step": 7440
+    },
+    {
+      "epoch": 413.8888888888889,
+      "grad_norm": 1.2934967530294215,
+      "learning_rate": 5.882230093744495e-06,
+      "loss": 0.1341,
+      "step": 7450
+    },
+    {
+      "epoch": 414.44444444444446,
+      "grad_norm": 1.0518393028942963,
+      "learning_rate": 5.8505613775103555e-06,
+      "loss": 0.1245,
+      "step": 7460
+    },
+    {
+      "epoch": 415.0,
+      "grad_norm": 1.2841087176652428,
+      "learning_rate": 5.819464972482287e-06,
+      "loss": 0.1274,
+      "step": 7470
+    },
+    {
+      "epoch": 415.55555555555554,
+      "grad_norm": 1.1869416178909626,
+      "learning_rate": 5.788941364256304e-06,
+      "loss": 0.1208,
+      "step": 7480
+    },
+    {
+      "epoch": 416.1111111111111,
+      "grad_norm": 1.3108921647900047,
+      "learning_rate": 5.758991029483714e-06,
+      "loss": 0.1294,
+      "step": 7490
+    },
+    {
+      "epoch": 416.6666666666667,
+      "grad_norm": 1.2439785407107473,
+      "learning_rate": 5.7296144358636925e-06,
+      "loss": 0.1297,
+      "step": 7500
+    },
+    {
+      "epoch": 417.22222222222223,
+      "grad_norm": 1.3611405406201007,
+      "learning_rate": 5.700812042135969e-06,
+      "loss": 0.1257,
+      "step": 7510
+    },
+    {
+      "epoch": 417.77777777777777,
+      "grad_norm": 1.209536326291414,
+      "learning_rate": 5.672584298073685e-06,
+      "loss": 0.1182,
+      "step": 7520
+    },
+    {
+      "epoch": 418.3333333333333,
+      "grad_norm": 1.2886591887276007,
+      "learning_rate": 5.644931644476331e-06,
+      "loss": 0.1332,
+      "step": 7530
+    },
+    {
+      "epoch": 418.8888888888889,
+      "grad_norm": 1.1468635724821037,
+      "learning_rate": 5.617854513162896e-06,
+      "loss": 0.1198,
+      "step": 7540
+    },
+    {
+      "epoch": 419.44444444444446,
+      "grad_norm": 1.2608106433535349,
+      "learning_rate": 5.5913533269651185e-06,
+      "loss": 0.1194,
+      "step": 7550
+    },
+    {
+      "epoch": 420.0,
+      "grad_norm": 1.1414235598735303,
+      "learning_rate": 5.565428499720873e-06,
+      "loss": 0.1233,
+      "step": 7560
+    },
+    {
+      "epoch": 420.55555555555554,
+      "grad_norm": 1.1777337676848558,
+      "learning_rate": 5.540080436267716e-06,
+      "loss": 0.1369,
+      "step": 7570
+    },
+    {
+      "epoch": 421.1111111111111,
+      "grad_norm": 1.1363421468436776,
+      "learning_rate": 5.515309532436559e-06,
+      "loss": 0.135,
+      "step": 7580
+    },
+    {
+      "epoch": 421.6666666666667,
+      "grad_norm": 1.4319375265015812,
+      "learning_rate": 5.491116175045491e-06,
+      "loss": 0.1242,
+      "step": 7590
+    },
+    {
+      "epoch": 422.22222222222223,
+      "grad_norm": 1.0959333154203064,
+      "learning_rate": 5.467500741893746e-06,
+      "loss": 0.1144,
+      "step": 7600
+    },
+    {
+      "epoch": 422.77777777777777,
+      "grad_norm": 1.3149867342145618,
+      "learning_rate": 5.444463601755772e-06,
+      "loss": 0.1244,
+      "step": 7610
+    },
+    {
+      "epoch": 423.3333333333333,
+      "grad_norm": 1.375747347516361,
+      "learning_rate": 5.422005114375519e-06,
+      "loss": 0.1442,
+      "step": 7620
+    },
+    {
+      "epoch": 423.8888888888889,
+      "grad_norm": 1.2101739752833829,
+      "learning_rate": 5.400125630460786e-06,
+      "loss": 0.1342,
+      "step": 7630
+    },
+    {
+      "epoch": 424.44444444444446,
+      "grad_norm": 1.272360223164354,
+      "learning_rate": 5.37882549167776e-06,
+      "loss": 0.1262,
+      "step": 7640
+    },
+    {
+      "epoch": 425.0,
+      "grad_norm": 1.2479030185944493,
+      "learning_rate": 5.358105030645666e-06,
+      "loss": 0.1165,
+      "step": 7650
+    },
+    {
+      "epoch": 425.55555555555554,
+      "grad_norm": 1.2277330777837763,
+      "learning_rate": 5.337964570931595e-06,
+      "loss": 0.1148,
+      "step": 7660
+    },
+    {
+      "epoch": 426.1111111111111,
+      "grad_norm": 1.316428934235771,
+      "learning_rate": 5.31840442704543e-06,
+      "loss": 0.1217,
+      "step": 7670
+    },
+    {
+      "epoch": 426.6666666666667,
+      "grad_norm": 1.1844826058092455,
+      "learning_rate": 5.299424904434947e-06,
+      "loss": 0.1147,
+      "step": 7680
+    },
+    {
+      "epoch": 427.22222222222223,
+      "grad_norm": 1.0935277440546316,
+      "learning_rate": 5.281026299481048e-06,
+      "loss": 0.1222,
+      "step": 7690
+    },
+    {
+      "epoch": 427.77777777777777,
+      "grad_norm": 1.2621320868254722,
+      "learning_rate": 5.263208899493116e-06,
+      "loss": 0.1141,
+      "step": 7700
+    },
+    {
+      "epoch": 428.3333333333333,
+      "grad_norm": 1.3621615208150106,
+      "learning_rate": 5.245972982704546e-06,
+      "loss": 0.1135,
+      "step": 7710
+    },
+    {
+      "epoch": 428.8888888888889,
+      "grad_norm": 1.1423472114085822,
+      "learning_rate": 5.229318818268392e-06,
+      "loss": 0.1236,
+      "step": 7720
+    },
+    {
+      "epoch": 429.44444444444446,
+      "grad_norm": 1.2176600252766276,
+      "learning_rate": 5.2132466662531654e-06,
+      "loss": 0.122,
+      "step": 7730
+    },
+    {
+      "epoch": 430.0,
+      "grad_norm": 1.3601825989828291,
+      "learning_rate": 5.197756777638772e-06,
+      "loss": 0.129,
+      "step": 7740
+    },
+    {
+      "epoch": 430.55555555555554,
+      "grad_norm": 1.0886680797323232,
+      "learning_rate": 5.182849394312596e-06,
+      "loss": 0.1109,
+      "step": 7750
+    },
+    {
+      "epoch": 431.1111111111111,
+      "grad_norm": 1.3540655783066489,
+      "learning_rate": 5.1685247490657234e-06,
+      "loss": 0.129,
+      "step": 7760
+    },
+    {
+      "epoch": 431.6666666666667,
+      "grad_norm": 1.3522264911274609,
+      "learning_rate": 5.154783065589303e-06,
+      "loss": 0.1148,
+      "step": 7770
+    },
+    {
+      "epoch": 432.22222222222223,
+      "grad_norm": 1.330606274488716,
+      "learning_rate": 5.141624558471049e-06,
+      "loss": 0.1146,
+      "step": 7780
+    },
+    {
+      "epoch": 432.77777777777777,
+      "grad_norm": 1.1848808818614405,
+      "learning_rate": 5.1290494331919025e-06,
+      "loss": 0.1181,
+      "step": 7790
+    },
+    {
+      "epoch": 433.3333333333333,
+      "grad_norm": 1.1701685879362898,
+      "learning_rate": 5.117057886122817e-06,
+      "loss": 0.1112,
+      "step": 7800
+    },
+    {
+      "epoch": 433.8888888888889,
+      "grad_norm": 1.1245603762065808,
+      "learning_rate": 5.105650104521688e-06,
+      "loss": 0.102,
+      "step": 7810
+    },
+    {
+      "epoch": 434.44444444444446,
+      "grad_norm": 1.2962067042654888,
+      "learning_rate": 5.094826266530429e-06,
+      "loss": 0.115,
+      "step": 7820
+    },
+    {
+      "epoch": 435.0,
+      "grad_norm": 1.2978117443719066,
+      "learning_rate": 5.0845865411722e-06,
+      "loss": 0.1145,
+      "step": 7830
+    },
+    {
+      "epoch": 435.55555555555554,
+      "grad_norm": 1.204839055167584,
+      "learning_rate": 5.074931088348753e-06,
+      "loss": 0.1172,
+      "step": 7840
+    },
+    {
+      "epoch": 436.1111111111111,
+      "grad_norm": 1.110792624339809,
+      "learning_rate": 5.06586005883795e-06,
+      "loss": 0.1129,
+      "step": 7850
+    },
+    {
+      "epoch": 436.6666666666667,
+      "grad_norm": 1.147070245987994,
+      "learning_rate": 5.057373594291389e-06,
+      "loss": 0.1279,
+      "step": 7860
+    },
+    {
+      "epoch": 437.22222222222223,
+      "grad_norm": 1.2149107690705474,
+      "learning_rate": 5.049471827232223e-06,
+      "loss": 0.1058,
+      "step": 7870
+    },
+    {
+      "epoch": 437.77777777777777,
+      "grad_norm": 1.2743545693728435,
+      "learning_rate": 5.042154881053054e-06,
+      "loss": 0.112,
+      "step": 7880
+    },
+    {
+      "epoch": 438.3333333333333,
+      "grad_norm": 1.1795760853682948,
+      "learning_rate": 5.0354228700140305e-06,
+      "loss": 0.1102,
+      "step": 7890
+    },
+    {
+      "epoch": 438.8888888888889,
+      "grad_norm": 1.096099074793285,
+      "learning_rate": 5.0292758992410565e-06,
+      "loss": 0.1109,
+      "step": 7900
+    },
+    {
+      "epoch": 439.44444444444446,
+      "grad_norm": 1.157268333485969,
+      "learning_rate": 5.023714064724151e-06,
+      "loss": 0.111,
+      "step": 7910
+    },
+    {
+      "epoch": 440.0,
+      "grad_norm": 1.1579891734135959,
+      "learning_rate": 5.018737453315944e-06,
+      "loss": 0.1021,
+      "step": 7920
+    },
+    {
+      "epoch": 440.55555555555554,
+      "grad_norm": 1.354900977749943,
+      "learning_rate": 5.014346142730328e-06,
+      "loss": 0.1062,
+      "step": 7930
+    },
+    {
+      "epoch": 441.1111111111111,
+      "grad_norm": 1.1966533246443283,
+      "learning_rate": 5.010540201541244e-06,
+      "loss": 0.1068,
+      "step": 7940
+    },
+    {
+      "epoch": 441.6666666666667,
+      "grad_norm": 1.0928533966292313,
+      "learning_rate": 5.007319689181595e-06,
+      "loss": 0.1103,
+      "step": 7950
+    },
+    {
+      "epoch": 442.22222222222223,
+      "grad_norm": 1.1450966229544974,
+      "learning_rate": 5.004684655942338e-06,
+      "loss": 0.1052,
+      "step": 7960
+    },
+    {
+      "epoch": 442.77777777777777,
+      "grad_norm": 1.3138677645529337,
+      "learning_rate": 5.0026351429716924e-06,
+      "loss": 0.1049,
+      "step": 7970
+    },
+    {
+      "epoch": 443.3333333333333,
+      "grad_norm": 1.2042515422295628,
+      "learning_rate": 5.0011711822744915e-06,
+      "loss": 0.1027,
+      "step": 7980
+    },
+    {
+      "epoch": 443.8888888888889,
+      "grad_norm": 1.3227655319429585,
+      "learning_rate": 5.000292796711689e-06,
+      "loss": 0.1119,
+      "step": 7990
+    },
+    {
+      "epoch": 444.44444444444446,
+      "grad_norm": 1.1895495846977757,
+      "learning_rate": 5e-06,
+      "loss": 0.106,
+      "step": 8000
+    },
+    {
+      "epoch": 444.44444444444446,
+      "step": 8000,
+      "total_flos": 7851246354432000.0,
+      "train_loss": 1.1700455456227064,
+      "train_runtime": 120013.2681,
+      "train_samples_per_second": 12.799,
+      "train_steps_per_second": 0.067
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 8000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 445,
+  "save_steps": 2000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7851246354432000.0,
+  "train_batch_size": 3,
+  "trial_name": null,
+  "trial_params": null
+}