diff --git "a/checkpoint-52810/trainer_state.json" "b/checkpoint-52810/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-52810/trainer_state.json"
@@ -0,0 +1,37055 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.0,
+  "eval_steps": 500,
+  "global_step": 52810,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0009467903806097331,
+      "grad_norm": 1.3941882848739624,
+      "learning_rate": 1.999621283847756e-05,
+      "loss": 0.6629,
+      "step": 10
+    },
+    {
+      "epoch": 0.0018935807612194661,
+      "grad_norm": 1.248525619506836,
+      "learning_rate": 1.9992425676955122e-05,
+      "loss": 0.554,
+      "step": 20
+    },
+    {
+      "epoch": 0.002840371141829199,
+      "grad_norm": 1.0281944274902344,
+      "learning_rate": 1.9988638515432685e-05,
+      "loss": 0.462,
+      "step": 30
+    },
+    {
+      "epoch": 0.0037871615224389322,
+      "grad_norm": 0.8631435036659241,
+      "learning_rate": 1.9984851353910245e-05,
+      "loss": 0.3908,
+      "step": 40
+    },
+    {
+      "epoch": 0.004733951903048665,
+      "grad_norm": 0.697015106678009,
+      "learning_rate": 1.9981064192387805e-05,
+      "loss": 0.3349,
+      "step": 50
+    },
+    {
+      "epoch": 0.005680742283658398,
+      "grad_norm": 0.6066704988479614,
+      "learning_rate": 1.997727703086537e-05,
+      "loss": 0.2918,
+      "step": 60
+    },
+    {
+      "epoch": 0.006627532664268131,
+      "grad_norm": 0.5035427808761597,
+      "learning_rate": 1.997348986934293e-05,
+      "loss": 0.2571,
+      "step": 70
+    },
+    {
+      "epoch": 0.0075743230448778644,
+      "grad_norm": 0.4318634271621704,
+      "learning_rate": 1.9969702707820492e-05,
+      "loss": 0.2356,
+      "step": 80
+    },
+    {
+      "epoch": 0.008521113425487597,
+      "grad_norm": 0.38201332092285156,
+      "learning_rate": 1.9965915546298052e-05,
+      "loss": 0.2173,
+      "step": 90
+    },
+    {
+      "epoch": 0.00946790380609733,
+      "grad_norm": 0.35830944776535034,
+      "learning_rate": 1.9962128384775612e-05,
+      "loss": 0.2039,
+      "step": 100
+    },
+    {
+      "epoch": 0.010414694186707063,
+      "grad_norm": 0.31125694513320923,
+      "learning_rate": 1.9958341223253172e-05,
+      "loss": 0.1964,
+      "step": 110
+    },
+    {
+      "epoch": 0.011361484567316796,
+      "grad_norm": 0.3396235406398773,
+      "learning_rate": 1.9954554061730736e-05,
+      "loss": 0.1988,
+      "step": 120
+    },
+    {
+      "epoch": 0.01230827494792653,
+      "grad_norm": 0.272406667470932,
+      "learning_rate": 1.9950766900208296e-05,
+      "loss": 0.1848,
+      "step": 130
+    },
+    {
+      "epoch": 0.013255065328536262,
+      "grad_norm": 0.27646365761756897,
+      "learning_rate": 1.9946979738685856e-05,
+      "loss": 0.1815,
+      "step": 140
+    },
+    {
+      "epoch": 0.014201855709145995,
+      "grad_norm": 0.3154810070991516,
+      "learning_rate": 1.9943192577163416e-05,
+      "loss": 0.1741,
+      "step": 150
+    },
+    {
+      "epoch": 0.015148646089755729,
+      "grad_norm": 0.2073918879032135,
+      "learning_rate": 1.993940541564098e-05,
+      "loss": 0.1785,
+      "step": 160
+    },
+    {
+      "epoch": 0.01609543647036546,
+      "grad_norm": 0.23369823396205902,
+      "learning_rate": 1.993561825411854e-05,
+      "loss": 0.1729,
+      "step": 170
+    },
+    {
+      "epoch": 0.017042226850975194,
+      "grad_norm": 0.21863120794296265,
+      "learning_rate": 1.99318310925961e-05,
+      "loss": 0.1715,
+      "step": 180
+    },
+    {
+      "epoch": 0.017989017231584928,
+      "grad_norm": 0.24868248403072357,
+      "learning_rate": 1.9928043931073663e-05,
+      "loss": 0.1679,
+      "step": 190
+    },
+    {
+      "epoch": 0.01893580761219466,
+      "grad_norm": 0.25883275270462036,
+      "learning_rate": 1.9924256769551223e-05,
+      "loss": 0.1668,
+      "step": 200
+    },
+    {
+      "epoch": 0.019882597992804393,
+      "grad_norm": 0.23148854076862335,
+      "learning_rate": 1.9920469608028786e-05,
+      "loss": 0.1624,
+      "step": 210
+    },
+    {
+      "epoch": 0.020829388373414127,
+      "grad_norm": 0.2580638825893402,
+      "learning_rate": 1.9916682446506346e-05,
+      "loss": 0.1667,
+      "step": 220
+    },
+    {
+      "epoch": 0.021776178754023857,
+      "grad_norm": 0.28563928604125977,
+      "learning_rate": 1.9912895284983906e-05,
+      "loss": 0.1673,
+      "step": 230
+    },
+    {
+      "epoch": 0.02272296913463359,
+      "grad_norm": 0.23611009120941162,
+      "learning_rate": 1.9909108123461467e-05,
+      "loss": 0.1687,
+      "step": 240
+    },
+    {
+      "epoch": 0.023669759515243326,
+      "grad_norm": 0.27900001406669617,
+      "learning_rate": 1.9905320961939027e-05,
+      "loss": 0.1626,
+      "step": 250
+    },
+    {
+      "epoch": 0.02461654989585306,
+      "grad_norm": 0.32736676931381226,
+      "learning_rate": 1.990153380041659e-05,
+      "loss": 0.1647,
+      "step": 260
+    },
+    {
+      "epoch": 0.02556334027646279,
+      "grad_norm": 0.25693920254707336,
+      "learning_rate": 1.989774663889415e-05,
+      "loss": 0.154,
+      "step": 270
+    },
+    {
+      "epoch": 0.026510130657072525,
+      "grad_norm": 0.2522478401660919,
+      "learning_rate": 1.989395947737171e-05,
+      "loss": 0.1672,
+      "step": 280
+    },
+    {
+      "epoch": 0.02745692103768226,
+      "grad_norm": 0.20121590793132782,
+      "learning_rate": 1.9890172315849274e-05,
+      "loss": 0.1585,
+      "step": 290
+    },
+    {
+      "epoch": 0.02840371141829199,
+      "grad_norm": 0.2583962380886078,
+      "learning_rate": 1.9886385154326834e-05,
+      "loss": 0.1558,
+      "step": 300
+    },
+    {
+      "epoch": 0.029350501798901724,
+      "grad_norm": 0.2211063653230667,
+      "learning_rate": 1.9882597992804397e-05,
+      "loss": 0.1645,
+      "step": 310
+    },
+    {
+      "epoch": 0.030297292179511458,
+      "grad_norm": 0.2027604579925537,
+      "learning_rate": 1.9878810831281957e-05,
+      "loss": 0.1608,
+      "step": 320
+    },
+    {
+      "epoch": 0.03124408256012119,
+      "grad_norm": 0.2854134738445282,
+      "learning_rate": 1.9875023669759517e-05,
+      "loss": 0.1599,
+      "step": 330
+    },
+    {
+      "epoch": 0.03219087294073092,
+      "grad_norm": 0.18335621058940887,
+      "learning_rate": 1.9871236508237077e-05,
+      "loss": 0.1688,
+      "step": 340
+    },
+    {
+      "epoch": 0.03313766332134065,
+      "grad_norm": 0.24929380416870117,
+      "learning_rate": 1.986744934671464e-05,
+      "loss": 0.1578,
+      "step": 350
+    },
+    {
+      "epoch": 0.03408445370195039,
+      "grad_norm": 0.20108000934123993,
+      "learning_rate": 1.98636621851922e-05,
+      "loss": 0.156,
+      "step": 360
+    },
+    {
+      "epoch": 0.03503124408256012,
+      "grad_norm": 0.24530109763145447,
+      "learning_rate": 1.985987502366976e-05,
+      "loss": 0.1497,
+      "step": 370
+    },
+    {
+      "epoch": 0.035978034463169856,
+      "grad_norm": 0.2246169000864029,
+      "learning_rate": 1.985608786214732e-05,
+      "loss": 0.1634,
+      "step": 380
+    },
+    {
+      "epoch": 0.03692482484377959,
+      "grad_norm": 0.2763858735561371,
+      "learning_rate": 1.985230070062488e-05,
+      "loss": 0.1592,
+      "step": 390
+    },
+    {
+      "epoch": 0.03787161522438932,
+      "grad_norm": 0.19560927152633667,
+      "learning_rate": 1.9848513539102444e-05,
+      "loss": 0.1488,
+      "step": 400
+    },
+    {
+      "epoch": 0.03881840560499905,
+      "grad_norm": 0.22450771927833557,
+      "learning_rate": 1.9844726377580004e-05,
+      "loss": 0.1575,
+      "step": 410
+    },
+    {
+      "epoch": 0.039765195985608785,
+      "grad_norm": 0.29325568675994873,
+      "learning_rate": 1.9840939216057568e-05,
+      "loss": 0.1591,
+      "step": 420
+    },
+    {
+      "epoch": 0.04071198636621852,
+      "grad_norm": 0.2534486651420593,
+      "learning_rate": 1.9837152054535128e-05,
+      "loss": 0.1588,
+      "step": 430
+    },
+    {
+      "epoch": 0.041658776746828254,
+      "grad_norm": 0.21553802490234375,
+      "learning_rate": 1.9833364893012688e-05,
+      "loss": 0.1572,
+      "step": 440
+    },
+    {
+      "epoch": 0.04260556712743799,
+      "grad_norm": 0.20270265638828278,
+      "learning_rate": 1.982957773149025e-05,
+      "loss": 0.1664,
+      "step": 450
+    },
+    {
+      "epoch": 0.043552357508047715,
+      "grad_norm": 0.26474466919898987,
+      "learning_rate": 1.982579056996781e-05,
+      "loss": 0.1549,
+      "step": 460
+    },
+    {
+      "epoch": 0.04449914788865745,
+      "grad_norm": 0.23204047977924347,
+      "learning_rate": 1.982200340844537e-05,
+      "loss": 0.157,
+      "step": 470
+    },
+    {
+      "epoch": 0.04544593826926718,
+      "grad_norm": 0.22077463567256927,
+      "learning_rate": 1.981821624692293e-05,
+      "loss": 0.1601,
+      "step": 480
+    },
+    {
+      "epoch": 0.04639272864987692,
+      "grad_norm": 0.24180015921592712,
+      "learning_rate": 1.9814429085400495e-05,
+      "loss": 0.1679,
+      "step": 490
+    },
+    {
+      "epoch": 0.04733951903048665,
+      "grad_norm": 0.20309610664844513,
+      "learning_rate": 1.9810641923878055e-05,
+      "loss": 0.15,
+      "step": 500
+    },
+    {
+      "epoch": 0.048286309411096386,
+      "grad_norm": 0.2009934037923813,
+      "learning_rate": 1.9806854762355615e-05,
+      "loss": 0.1614,
+      "step": 510
+    },
+    {
+      "epoch": 0.04923309979170612,
+      "grad_norm": 0.29323482513427734,
+      "learning_rate": 1.980306760083318e-05,
+      "loss": 0.1635,
+      "step": 520
+    },
+    {
+      "epoch": 0.05017989017231585,
+      "grad_norm": 0.20960304141044617,
+      "learning_rate": 1.979928043931074e-05,
+      "loss": 0.1522,
+      "step": 530
+    },
+    {
+      "epoch": 0.05112668055292558,
+      "grad_norm": 0.22783887386322021,
+      "learning_rate": 1.97954932777883e-05,
+      "loss": 0.1573,
+      "step": 540
+    },
+    {
+      "epoch": 0.052073470933535315,
+      "grad_norm": 0.22936321794986725,
+      "learning_rate": 1.9791706116265862e-05,
+      "loss": 0.1436,
+      "step": 550
+    },
+    {
+      "epoch": 0.05302026131414505,
+      "grad_norm": 0.22300918400287628,
+      "learning_rate": 1.9787918954743422e-05,
+      "loss": 0.1546,
+      "step": 560
+    },
+    {
+      "epoch": 0.053967051694754783,
+      "grad_norm": 0.22048968076705933,
+      "learning_rate": 1.9784131793220982e-05,
+      "loss": 0.1416,
+      "step": 570
+    },
+    {
+      "epoch": 0.05491384207536452,
+      "grad_norm": 0.254375696182251,
+      "learning_rate": 1.9780344631698542e-05,
+      "loss": 0.1532,
+      "step": 580
+    },
+    {
+      "epoch": 0.055860632455974245,
+      "grad_norm": 0.3245167136192322,
+      "learning_rate": 1.9776557470176106e-05,
+      "loss": 0.15,
+      "step": 590
+    },
+    {
+      "epoch": 0.05680742283658398,
+      "grad_norm": 0.21500037610530853,
+      "learning_rate": 1.9772770308653666e-05,
+      "loss": 0.1524,
+      "step": 600
+    },
+    {
+      "epoch": 0.05775421321719371,
+      "grad_norm": 0.19425217807292938,
+      "learning_rate": 1.9768983147131226e-05,
+      "loss": 0.1462,
+      "step": 610
+    },
+    {
+      "epoch": 0.05870100359780345,
+      "grad_norm": 0.23880882561206818,
+      "learning_rate": 1.9765195985608786e-05,
+      "loss": 0.1477,
+      "step": 620
+    },
+    {
+      "epoch": 0.05964779397841318,
+      "grad_norm": 0.2194243222475052,
+      "learning_rate": 1.976140882408635e-05,
+      "loss": 0.1501,
+      "step": 630
+    },
+    {
+      "epoch": 0.060594584359022916,
+      "grad_norm": 0.3316631317138672,
+      "learning_rate": 1.975762166256391e-05,
+      "loss": 0.158,
+      "step": 640
+    },
+    {
+      "epoch": 0.06154137473963264,
+      "grad_norm": 0.23250705003738403,
+      "learning_rate": 1.9753834501041473e-05,
+      "loss": 0.1486,
+      "step": 650
+    },
+    {
+      "epoch": 0.06248816512024238,
+      "grad_norm": 0.211704283952713,
+      "learning_rate": 1.9750047339519033e-05,
+      "loss": 0.1487,
+      "step": 660
+    },
+    {
+      "epoch": 0.06343495550085211,
+      "grad_norm": 0.30874431133270264,
+      "learning_rate": 1.9746260177996593e-05,
+      "loss": 0.1529,
+      "step": 670
+    },
+    {
+      "epoch": 0.06438174588146184,
+      "grad_norm": 0.26447850465774536,
+      "learning_rate": 1.9742473016474156e-05,
+      "loss": 0.1521,
+      "step": 680
+    },
+    {
+      "epoch": 0.06532853626207158,
+      "grad_norm": 0.23446421325206757,
+      "learning_rate": 1.9738685854951716e-05,
+      "loss": 0.1525,
+      "step": 690
+    },
+    {
+      "epoch": 0.0662753266426813,
+      "grad_norm": 0.2225397229194641,
+      "learning_rate": 1.9734898693429276e-05,
+      "loss": 0.1476,
+      "step": 700
+    },
+    {
+      "epoch": 0.06722211702329105,
+      "grad_norm": 0.20426969230175018,
+      "learning_rate": 1.9731111531906836e-05,
+      "loss": 0.1564,
+      "step": 710
+    },
+    {
+      "epoch": 0.06816890740390077,
+      "grad_norm": 0.27150651812553406,
+      "learning_rate": 1.9727324370384396e-05,
+      "loss": 0.1459,
+      "step": 720
+    },
+    {
+      "epoch": 0.06911569778451052,
+      "grad_norm": 0.23939593136310577,
+      "learning_rate": 1.972353720886196e-05,
+      "loss": 0.1399,
+      "step": 730
+    },
+    {
+      "epoch": 0.07006248816512024,
+      "grad_norm": 0.26259246468544006,
+      "learning_rate": 1.971975004733952e-05,
+      "loss": 0.1483,
+      "step": 740
+    },
+    {
+      "epoch": 0.07100927854572997,
+      "grad_norm": 0.22480525076389313,
+      "learning_rate": 1.971596288581708e-05,
+      "loss": 0.14,
+      "step": 750
+    },
+    {
+      "epoch": 0.07195606892633971,
+      "grad_norm": 0.2269698679447174,
+      "learning_rate": 1.9712175724294643e-05,
+      "loss": 0.1497,
+      "step": 760
+    },
+    {
+      "epoch": 0.07290285930694944,
+      "grad_norm": 0.216821551322937,
+      "learning_rate": 1.9708388562772203e-05,
+      "loss": 0.1422,
+      "step": 770
+    },
+    {
+      "epoch": 0.07384964968755918,
+      "grad_norm": 0.24823130667209625,
+      "learning_rate": 1.9704601401249767e-05,
+      "loss": 0.1436,
+      "step": 780
+    },
+    {
+      "epoch": 0.0747964400681689,
+      "grad_norm": 0.3634086847305298,
+      "learning_rate": 1.9700814239727327e-05,
+      "loss": 0.1579,
+      "step": 790
+    },
+    {
+      "epoch": 0.07574323044877863,
+      "grad_norm": 0.24505460262298584,
+      "learning_rate": 1.9697027078204887e-05,
+      "loss": 0.1423,
+      "step": 800
+    },
+    {
+      "epoch": 0.07669002082938838,
+      "grad_norm": 0.30174288153648376,
+      "learning_rate": 1.9693239916682447e-05,
+      "loss": 0.1504,
+      "step": 810
+    },
+    {
+      "epoch": 0.0776368112099981,
+      "grad_norm": 0.2679823637008667,
+      "learning_rate": 1.968945275516001e-05,
+      "loss": 0.1428,
+      "step": 820
+    },
+    {
+      "epoch": 0.07858360159060784,
+      "grad_norm": 0.30597180128097534,
+      "learning_rate": 1.968566559363757e-05,
+      "loss": 0.1487,
+      "step": 830
+    },
+    {
+      "epoch": 0.07953039197121757,
+      "grad_norm": 0.25198864936828613,
+      "learning_rate": 1.968187843211513e-05,
+      "loss": 0.1442,
+      "step": 840
+    },
+    {
+      "epoch": 0.08047718235182731,
+      "grad_norm": 0.23763233423233032,
+      "learning_rate": 1.967809127059269e-05,
+      "loss": 0.1437,
+      "step": 850
+    },
+    {
+      "epoch": 0.08142397273243704,
+      "grad_norm": 0.20261543989181519,
+      "learning_rate": 1.9674304109070254e-05,
+      "loss": 0.146,
+      "step": 860
+    },
+    {
+      "epoch": 0.08237076311304677,
+      "grad_norm": 0.24674880504608154,
+      "learning_rate": 1.9670516947547814e-05,
+      "loss": 0.1485,
+      "step": 870
+    },
+    {
+      "epoch": 0.08331755349365651,
+      "grad_norm": 0.23125316202640533,
+      "learning_rate": 1.9666729786025377e-05,
+      "loss": 0.1434,
+      "step": 880
+    },
+    {
+      "epoch": 0.08426434387426623,
+      "grad_norm": 0.2659028470516205,
+      "learning_rate": 1.9662942624502938e-05,
+      "loss": 0.1428,
+      "step": 890
+    },
+    {
+      "epoch": 0.08521113425487598,
+      "grad_norm": 0.22943943738937378,
+      "learning_rate": 1.9659155462980498e-05,
+      "loss": 0.1474,
+      "step": 900
+    },
+    {
+      "epoch": 0.0861579246354857,
+      "grad_norm": 0.25804975628852844,
+      "learning_rate": 1.965536830145806e-05,
+      "loss": 0.1404,
+      "step": 910
+    },
+    {
+      "epoch": 0.08710471501609543,
+      "grad_norm": 0.26774296164512634,
+      "learning_rate": 1.965158113993562e-05,
+      "loss": 0.1433,
+      "step": 920
+    },
+    {
+      "epoch": 0.08805150539670517,
+      "grad_norm": 0.2502722144126892,
+      "learning_rate": 1.964779397841318e-05,
+      "loss": 0.1391,
+      "step": 930
+    },
+    {
+      "epoch": 0.0889982957773149,
+      "grad_norm": 0.22401435673236847,
+      "learning_rate": 1.964400681689074e-05,
+      "loss": 0.145,
+      "step": 940
+    },
+    {
+      "epoch": 0.08994508615792464,
+      "grad_norm": 0.25525417923927307,
+      "learning_rate": 1.96402196553683e-05,
+      "loss": 0.1481,
+      "step": 950
+    },
+    {
+      "epoch": 0.09089187653853437,
+      "grad_norm": 0.2963325083255768,
+      "learning_rate": 1.9636432493845865e-05,
+      "loss": 0.1452,
+      "step": 960
+    },
+    {
+      "epoch": 0.09183866691914411,
+      "grad_norm": 0.257304847240448,
+      "learning_rate": 1.9632645332323425e-05,
+      "loss": 0.1409,
+      "step": 970
+    },
+    {
+      "epoch": 0.09278545729975383,
+      "grad_norm": 0.3979777991771698,
+      "learning_rate": 1.9628858170800985e-05,
+      "loss": 0.1363,
+      "step": 980
+    },
+    {
+      "epoch": 0.09373224768036356,
+      "grad_norm": 0.2767225205898285,
+      "learning_rate": 1.9625071009278548e-05,
+      "loss": 0.1402,
+      "step": 990
+    },
+    {
+      "epoch": 0.0946790380609733,
+      "grad_norm": 0.22073742747306824,
+      "learning_rate": 1.9621283847756108e-05,
+      "loss": 0.1471,
+      "step": 1000
+    },
+    {
+      "epoch": 0.09562582844158303,
+      "grad_norm": 0.2693091928958893,
+      "learning_rate": 1.961749668623367e-05,
+      "loss": 0.1505,
+      "step": 1010
+    },
+    {
+      "epoch": 0.09657261882219277,
+      "grad_norm": 0.24559836089611053,
+      "learning_rate": 1.9613709524711232e-05,
+      "loss": 0.1369,
+      "step": 1020
+    },
+    {
+      "epoch": 0.0975194092028025,
+      "grad_norm": 0.22167791426181793,
+      "learning_rate": 1.9609922363188792e-05,
+      "loss": 0.1348,
+      "step": 1030
+    },
+    {
+      "epoch": 0.09846619958341224,
+      "grad_norm": 0.23056115210056305,
+      "learning_rate": 1.9606135201666352e-05,
+      "loss": 0.1394,
+      "step": 1040
+    },
+    {
+      "epoch": 0.09941298996402197,
+      "grad_norm": 0.35368019342422485,
+      "learning_rate": 1.9602348040143915e-05,
+      "loss": 0.1485,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1003597803446317,
+      "grad_norm": 0.35556063055992126,
+      "learning_rate": 1.9598560878621475e-05,
+      "loss": 0.1489,
+      "step": 1060
+    },
+    {
+      "epoch": 0.10130657072524143,
+      "grad_norm": 0.2518173158168793,
+      "learning_rate": 1.9594773717099035e-05,
+      "loss": 0.1489,
+      "step": 1070
+    },
+    {
+      "epoch": 0.10225336110585116,
+      "grad_norm": 0.2986575663089752,
+      "learning_rate": 1.9590986555576595e-05,
+      "loss": 0.1397,
+      "step": 1080
+    },
+    {
+      "epoch": 0.1032001514864609,
+      "grad_norm": 0.22278901934623718,
+      "learning_rate": 1.958719939405416e-05,
+      "loss": 0.1322,
+      "step": 1090
+    },
+    {
+      "epoch": 0.10414694186707063,
+      "grad_norm": 0.2492545247077942,
+      "learning_rate": 1.958341223253172e-05,
+      "loss": 0.1344,
+      "step": 1100
+    },
+    {
+      "epoch": 0.10509373224768036,
+      "grad_norm": 0.348187118768692,
+      "learning_rate": 1.957962507100928e-05,
+      "loss": 0.15,
+      "step": 1110
+    },
+    {
+      "epoch": 0.1060405226282901,
+      "grad_norm": 0.3564904034137726,
+      "learning_rate": 1.9575837909486842e-05,
+      "loss": 0.1401,
+      "step": 1120
+    },
+    {
+      "epoch": 0.10698731300889983,
+      "grad_norm": 0.25616133213043213,
+      "learning_rate": 1.9572050747964402e-05,
+      "loss": 0.1451,
+      "step": 1130
+    },
+    {
+      "epoch": 0.10793410338950957,
+      "grad_norm": 0.2531684637069702,
+      "learning_rate": 1.9568263586441962e-05,
+      "loss": 0.1426,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1088808937701193,
+      "grad_norm": 0.2361394166946411,
+      "learning_rate": 1.9564476424919526e-05,
+      "loss": 0.1332,
+      "step": 1150
+    },
+    {
+      "epoch": 0.10982768415072904,
+      "grad_norm": 0.2564547657966614,
+      "learning_rate": 1.9560689263397086e-05,
+      "loss": 0.1352,
+      "step": 1160
+    },
+    {
+      "epoch": 0.11077447453133876,
+      "grad_norm": 0.2713509500026703,
+      "learning_rate": 1.9556902101874646e-05,
+      "loss": 0.1346,
+      "step": 1170
+    },
+    {
+      "epoch": 0.11172126491194849,
+      "grad_norm": 0.3002881705760956,
+      "learning_rate": 1.9553114940352206e-05,
+      "loss": 0.1434,
+      "step": 1180
+    },
+    {
+      "epoch": 0.11266805529255823,
+      "grad_norm": 0.3270877003669739,
+      "learning_rate": 1.954932777882977e-05,
+      "loss": 0.1441,
+      "step": 1190
+    },
+    {
+      "epoch": 0.11361484567316796,
+      "grad_norm": 0.31670162081718445,
+      "learning_rate": 1.954554061730733e-05,
+      "loss": 0.142,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1145616360537777,
+      "grad_norm": 0.290374755859375,
+      "learning_rate": 1.954175345578489e-05,
+      "loss": 0.1443,
+      "step": 1210
+    },
+    {
+      "epoch": 0.11550842643438743,
+      "grad_norm": 0.25081178545951843,
+      "learning_rate": 1.9537966294262453e-05,
+      "loss": 0.1308,
+      "step": 1220
+    },
+    {
+      "epoch": 0.11645521681499715,
+      "grad_norm": 0.29356658458709717,
+      "learning_rate": 1.9534179132740013e-05,
+      "loss": 0.1281,
+      "step": 1230
+    },
+    {
+      "epoch": 0.1174020071956069,
+      "grad_norm": 0.38654467463493347,
+      "learning_rate": 1.9530391971217577e-05,
+      "loss": 0.1515,
+      "step": 1240
+    },
+    {
+      "epoch": 0.11834879757621662,
+      "grad_norm": 0.2073059380054474,
+      "learning_rate": 1.9526604809695137e-05,
+      "loss": 0.13,
+      "step": 1250
+    },
+    {
+      "epoch": 0.11929558795682636,
+      "grad_norm": 0.21164889633655548,
+      "learning_rate": 1.9522817648172697e-05,
+      "loss": 0.1325,
+      "step": 1260
+    },
+    {
+      "epoch": 0.12024237833743609,
+      "grad_norm": 0.28698667883872986,
+      "learning_rate": 1.9519030486650257e-05,
+      "loss": 0.1349,
+      "step": 1270
+    },
+    {
+      "epoch": 0.12118916871804583,
+      "grad_norm": 0.24369283020496368,
+      "learning_rate": 1.9515243325127817e-05,
+      "loss": 0.1364,
+      "step": 1280
+    },
+    {
+      "epoch": 0.12213595909865556,
+      "grad_norm": 0.24442093074321747,
+      "learning_rate": 1.951145616360538e-05,
+      "loss": 0.134,
+      "step": 1290
+    },
+    {
+      "epoch": 0.12308274947926529,
+      "grad_norm": 0.21796512603759766,
+      "learning_rate": 1.950766900208294e-05,
+      "loss": 0.1447,
+      "step": 1300
+    },
+    {
+      "epoch": 0.12402953985987503,
+      "grad_norm": 0.2371196150779724,
+      "learning_rate": 1.95038818405605e-05,
+      "loss": 0.1339,
+      "step": 1310
+    },
+    {
+      "epoch": 0.12497633024048475,
+      "grad_norm": 0.31082502007484436,
+      "learning_rate": 1.950009467903806e-05,
+      "loss": 0.1288,
+      "step": 1320
+    },
+    {
+      "epoch": 0.12592312062109448,
+      "grad_norm": 0.24794170260429382,
+      "learning_rate": 1.9496307517515624e-05,
+      "loss": 0.1293,
+      "step": 1330
+    },
+    {
+      "epoch": 0.12686991100170422,
+      "grad_norm": 0.21178680658340454,
+      "learning_rate": 1.9492520355993184e-05,
+      "loss": 0.1353,
+      "step": 1340
+    },
+    {
+      "epoch": 0.12781670138231396,
+      "grad_norm": 0.3490561246871948,
+      "learning_rate": 1.9488733194470747e-05,
+      "loss": 0.1343,
+      "step": 1350
+    },
+    {
+      "epoch": 0.12876349176292368,
+      "grad_norm": 0.25812265276908875,
+      "learning_rate": 1.9484946032948307e-05,
+      "loss": 0.1304,
+      "step": 1360
+    },
+    {
+      "epoch": 0.12971028214353342,
+      "grad_norm": 0.2792973220348358,
+      "learning_rate": 1.9481158871425867e-05,
+      "loss": 0.135,
+      "step": 1370
+    },
+    {
+      "epoch": 0.13065707252414316,
+      "grad_norm": 0.4771386384963989,
+      "learning_rate": 1.947737170990343e-05,
+      "loss": 0.1367,
+      "step": 1380
+    },
+    {
+      "epoch": 0.1316038629047529,
+      "grad_norm": 0.2986657917499542,
+      "learning_rate": 1.947358454838099e-05,
+      "loss": 0.127,
+      "step": 1390
+    },
+    {
+      "epoch": 0.1325506532853626,
+      "grad_norm": 0.2369564324617386,
+      "learning_rate": 1.946979738685855e-05,
+      "loss": 0.132,
+      "step": 1400
+    },
+    {
+      "epoch": 0.13349744366597235,
+      "grad_norm": 0.27226126194000244,
+      "learning_rate": 1.946601022533611e-05,
+      "loss": 0.1347,
+      "step": 1410
+    },
+    {
+      "epoch": 0.1344442340465821,
+      "grad_norm": 0.26052626967430115,
+      "learning_rate": 1.946222306381367e-05,
+      "loss": 0.1331,
+      "step": 1420
+    },
+    {
+      "epoch": 0.1353910244271918,
+      "grad_norm": 0.34890538454055786,
+      "learning_rate": 1.9458435902291234e-05,
+      "loss": 0.1347,
+      "step": 1430
+    },
+    {
+      "epoch": 0.13633781480780155,
+      "grad_norm": 0.2817912995815277,
+      "learning_rate": 1.9454648740768794e-05,
+      "loss": 0.1307,
+      "step": 1440
+    },
+    {
+      "epoch": 0.1372846051884113,
+      "grad_norm": 0.2826266884803772,
+      "learning_rate": 1.9450861579246358e-05,
+      "loss": 0.1426,
+      "step": 1450
+    },
+    {
+      "epoch": 0.13823139556902103,
+      "grad_norm": 0.29997652769088745,
+      "learning_rate": 1.9447074417723918e-05,
+      "loss": 0.1327,
+      "step": 1460
+    },
+    {
+      "epoch": 0.13917818594963074,
+      "grad_norm": 0.30327969789505005,
+      "learning_rate": 1.9443287256201478e-05,
+      "loss": 0.1326,
+      "step": 1470
+    },
+    {
+      "epoch": 0.14012497633024049,
+      "grad_norm": 0.2640013098716736,
+      "learning_rate": 1.943950009467904e-05,
+      "loss": 0.1376,
+      "step": 1480
+    },
+    {
+      "epoch": 0.14107176671085023,
+      "grad_norm": 0.24050156772136688,
+      "learning_rate": 1.94357129331566e-05,
+      "loss": 0.132,
+      "step": 1490
+    },
+    {
+      "epoch": 0.14201855709145994,
+      "grad_norm": 0.34889718890190125,
+      "learning_rate": 1.943192577163416e-05,
+      "loss": 0.1286,
+      "step": 1500
+    },
+    {
+      "epoch": 0.14296534747206968,
+      "grad_norm": 0.24513930082321167,
+      "learning_rate": 1.942813861011172e-05,
+      "loss": 0.1351,
+      "step": 1510
+    },
+    {
+      "epoch": 0.14391213785267942,
+      "grad_norm": 0.3242141902446747,
+      "learning_rate": 1.9424351448589285e-05,
+      "loss": 0.1357,
+      "step": 1520
+    },
+    {
+      "epoch": 0.14485892823328914,
+      "grad_norm": 0.25881341099739075,
+      "learning_rate": 1.9420564287066845e-05,
+      "loss": 0.1307,
+      "step": 1530
+    },
+    {
+      "epoch": 0.14580571861389888,
+      "grad_norm": 0.24430468678474426,
+      "learning_rate": 1.9416777125544405e-05,
+      "loss": 0.1208,
+      "step": 1540
+    },
+    {
+      "epoch": 0.14675250899450862,
+      "grad_norm": 0.32587867975234985,
+      "learning_rate": 1.9412989964021965e-05,
+      "loss": 0.133,
+      "step": 1550
+    },
+    {
+      "epoch": 0.14769929937511836,
+      "grad_norm": 0.30574116110801697,
+      "learning_rate": 1.940920280249953e-05,
+      "loss": 0.1321,
+      "step": 1560
+    },
+    {
+      "epoch": 0.14864608975572807,
+      "grad_norm": 0.316580206155777,
+      "learning_rate": 1.940541564097709e-05,
+      "loss": 0.1387,
+      "step": 1570
+    },
+    {
+      "epoch": 0.1495928801363378,
+      "grad_norm": 0.4084049165248871,
+      "learning_rate": 1.9401628479454652e-05,
+      "loss": 0.134,
+      "step": 1580
+    },
+    {
+      "epoch": 0.15053967051694755,
+      "grad_norm": 0.3074231445789337,
+      "learning_rate": 1.9397841317932212e-05,
+      "loss": 0.1277,
+      "step": 1590
+    },
+    {
+      "epoch": 0.15148646089755727,
+      "grad_norm": 0.34363508224487305,
+      "learning_rate": 1.9394054156409772e-05,
+      "loss": 0.1305,
+      "step": 1600
+    },
+    {
+      "epoch": 0.152433251278167,
+      "grad_norm": 0.27845168113708496,
+      "learning_rate": 1.9390266994887336e-05,
+      "loss": 0.1259,
+      "step": 1610
+    },
+    {
+      "epoch": 0.15338004165877675,
+      "grad_norm": 0.2976890504360199,
+      "learning_rate": 1.9386479833364896e-05,
+      "loss": 0.1404,
+      "step": 1620
+    },
+    {
+      "epoch": 0.1543268320393865,
+      "grad_norm": 0.23898760974407196,
+      "learning_rate": 1.9382692671842456e-05,
+      "loss": 0.1243,
+      "step": 1630
+    },
+    {
+      "epoch": 0.1552736224199962,
+      "grad_norm": 0.28521913290023804,
+      "learning_rate": 1.9378905510320016e-05,
+      "loss": 0.1265,
+      "step": 1640
+    },
+    {
+      "epoch": 0.15622041280060595,
+      "grad_norm": 0.34148770570755005,
+      "learning_rate": 1.9375118348797576e-05,
+      "loss": 0.1347,
+      "step": 1650
+    },
+    {
+      "epoch": 0.1571672031812157,
+      "grad_norm": 0.43205875158309937,
+      "learning_rate": 1.937133118727514e-05,
+      "loss": 0.1246,
+      "step": 1660
+    },
+    {
+      "epoch": 0.1581139935618254,
+      "grad_norm": 0.33226126432418823,
+      "learning_rate": 1.93675440257527e-05,
+      "loss": 0.1273,
+      "step": 1670
+    },
+    {
+      "epoch": 0.15906078394243514,
+      "grad_norm": 0.3262731730937958,
+      "learning_rate": 1.936375686423026e-05,
+      "loss": 0.1241,
+      "step": 1680
+    },
+    {
+      "epoch": 0.16000757432304488,
+      "grad_norm": 0.27835527062416077,
+      "learning_rate": 1.9359969702707823e-05,
+      "loss": 0.1314,
+      "step": 1690
+    },
+    {
+      "epoch": 0.16095436470365462,
+      "grad_norm": 0.28034350275993347,
+      "learning_rate": 1.9356182541185383e-05,
+      "loss": 0.1297,
+      "step": 1700
+    },
+    {
+      "epoch": 0.16190115508426434,
+      "grad_norm": 0.2608806788921356,
+      "learning_rate": 1.9352395379662946e-05,
+      "loss": 0.1244,
+      "step": 1710
+    },
+    {
+      "epoch": 0.16284794546487408,
+      "grad_norm": 0.32022422552108765,
+      "learning_rate": 1.9348608218140506e-05,
+      "loss": 0.1297,
+      "step": 1720
+    },
+    {
+      "epoch": 0.16379473584548382,
+      "grad_norm": 0.2928096354007721,
+      "learning_rate": 1.9344821056618066e-05,
+      "loss": 0.1246,
+      "step": 1730
+    },
+    {
+      "epoch": 0.16474152622609353,
+      "grad_norm": 0.3574795126914978,
+      "learning_rate": 1.9341033895095626e-05,
+      "loss": 0.1383,
+      "step": 1740
+    },
+    {
+      "epoch": 0.16568831660670327,
+      "grad_norm": 0.2713389992713928,
+      "learning_rate": 1.933724673357319e-05,
+      "loss": 0.1344,
+      "step": 1750
+    },
+    {
+      "epoch": 0.16663510698731301,
+      "grad_norm": 0.2822892665863037,
+      "learning_rate": 1.933345957205075e-05,
+      "loss": 0.1281,
+      "step": 1760
+    },
+    {
+      "epoch": 0.16758189736792276,
+      "grad_norm": 0.3460506200790405,
+      "learning_rate": 1.932967241052831e-05,
+      "loss": 0.1307,
+      "step": 1770
+    },
+    {
+      "epoch": 0.16852868774853247,
+      "grad_norm": 0.323542058467865,
+      "learning_rate": 1.932588524900587e-05,
+      "loss": 0.1362,
+      "step": 1780
+    },
+    {
+      "epoch": 0.1694754781291422,
+      "grad_norm": 0.30738893151283264,
+      "learning_rate": 1.9322098087483433e-05,
+      "loss": 0.1282,
+      "step": 1790
+    },
+    {
+      "epoch": 0.17042226850975195,
+      "grad_norm": 0.34989845752716064,
+      "learning_rate": 1.9318310925960994e-05,
+      "loss": 0.1234,
+      "step": 1800
+    },
+    {
+      "epoch": 0.17136905889036166,
+      "grad_norm": 0.33833956718444824,
+      "learning_rate": 1.9314523764438557e-05,
+      "loss": 0.1313,
+      "step": 1810
+    },
+    {
+      "epoch": 0.1723158492709714,
+      "grad_norm": 0.3208501636981964,
+      "learning_rate": 1.9310736602916117e-05,
+      "loss": 0.136,
+      "step": 1820
+    },
+    {
+      "epoch": 0.17326263965158115,
+      "grad_norm": 0.38496994972229004,
+      "learning_rate": 1.9306949441393677e-05,
+      "loss": 0.1278,
+      "step": 1830
+    },
+    {
+      "epoch": 0.17420943003219086,
+      "grad_norm": 0.2736138105392456,
+      "learning_rate": 1.9303162279871237e-05,
+      "loss": 0.1262,
+      "step": 1840
+    },
+    {
+      "epoch": 0.1751562204128006,
+      "grad_norm": 0.25814059376716614,
+      "learning_rate": 1.92993751183488e-05,
+      "loss": 0.1246,
+      "step": 1850
+    },
+    {
+      "epoch": 0.17610301079341034,
+      "grad_norm": 0.2308269441127777,
+      "learning_rate": 1.929558795682636e-05,
+      "loss": 0.136,
+      "step": 1860
+    },
+    {
+      "epoch": 0.17704980117402008,
+      "grad_norm": 0.3181948661804199,
+      "learning_rate": 1.929180079530392e-05,
+      "loss": 0.1206,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1779965915546298,
+      "grad_norm": 0.2826565206050873,
+      "learning_rate": 1.928801363378148e-05,
+      "loss": 0.1329,
+      "step": 1880
+    },
+    {
+      "epoch": 0.17894338193523954,
+      "grad_norm": 0.4134462773799896,
+      "learning_rate": 1.9284226472259044e-05,
+      "loss": 0.126,
+      "step": 1890
+    },
+    {
+      "epoch": 0.17989017231584928,
+      "grad_norm": 0.3727562725543976,
+      "learning_rate": 1.9280439310736604e-05,
+      "loss": 0.1319,
+      "step": 1900
+    },
+    {
+      "epoch": 0.180836962696459,
+      "grad_norm": 0.25525522232055664,
+      "learning_rate": 1.9276652149214164e-05,
+      "loss": 0.1312,
+      "step": 1910
+    },
+    {
+      "epoch": 0.18178375307706873,
+      "grad_norm": 0.2545289397239685,
+      "learning_rate": 1.9272864987691728e-05,
+      "loss": 0.1367,
+      "step": 1920
+    },
+    {
+      "epoch": 0.18273054345767847,
+      "grad_norm": 0.3217427730560303,
+      "learning_rate": 1.9269077826169288e-05,
+      "loss": 0.1367,
+      "step": 1930
+    },
+    {
+      "epoch": 0.18367733383828821,
+      "grad_norm": 0.25008267164230347,
+      "learning_rate": 1.926529066464685e-05,
+      "loss": 0.1259,
+      "step": 1940
+    },
+    {
+      "epoch": 0.18462412421889793,
+      "grad_norm": 0.28599464893341064,
+      "learning_rate": 1.926150350312441e-05,
+      "loss": 0.1283,
+      "step": 1950
+    },
+    {
+      "epoch": 0.18557091459950767,
+      "grad_norm": 0.30091628432273865,
+      "learning_rate": 1.925771634160197e-05,
+      "loss": 0.1424,
+      "step": 1960
+    },
+    {
+      "epoch": 0.1865177049801174,
+      "grad_norm": 0.2841487228870392,
+      "learning_rate": 1.925392918007953e-05,
+      "loss": 0.1225,
+      "step": 1970
+    },
+    {
+      "epoch": 0.18746449536072712,
+      "grad_norm": 0.2743360996246338,
+      "learning_rate": 1.925014201855709e-05,
+      "loss": 0.1307,
+      "step": 1980
+    },
+    {
+      "epoch": 0.18841128574133686,
+      "grad_norm": 0.2886848449707031,
+      "learning_rate": 1.9246354857034655e-05,
+      "loss": 0.1292,
+      "step": 1990
+    },
+    {
+      "epoch": 0.1893580761219466,
+      "grad_norm": 0.35317230224609375,
+      "learning_rate": 1.9242567695512215e-05,
+      "loss": 0.1357,
+      "step": 2000
+    },
+    {
+      "epoch": 0.19030486650255635,
+      "grad_norm": 0.2847851514816284,
+      "learning_rate": 1.9238780533989775e-05,
+      "loss": 0.1222,
+      "step": 2010
+    },
+    {
+      "epoch": 0.19125165688316606,
+      "grad_norm": 0.3323739469051361,
+      "learning_rate": 1.9234993372467335e-05,
+      "loss": 0.1251,
+      "step": 2020
+    },
+    {
+      "epoch": 0.1921984472637758,
+      "grad_norm": 0.35399678349494934,
+      "learning_rate": 1.92312062109449e-05,
+      "loss": 0.1334,
+      "step": 2030
+    },
+    {
+      "epoch": 0.19314523764438554,
+      "grad_norm": 0.2679152488708496,
+      "learning_rate": 1.922741904942246e-05,
+      "loss": 0.1271,
+      "step": 2040
+    },
+    {
+      "epoch": 0.19409202802499526,
+      "grad_norm": 0.23062950372695923,
+      "learning_rate": 1.9223631887900022e-05,
+      "loss": 0.1259,
+      "step": 2050
+    },
+    {
+      "epoch": 0.195038818405605,
+      "grad_norm": 0.25115498900413513,
+      "learning_rate": 1.9219844726377582e-05,
+      "loss": 0.1283,
+      "step": 2060
+    },
+    {
+      "epoch": 0.19598560878621474,
+      "grad_norm": 0.2893475294113159,
+      "learning_rate": 1.9216057564855142e-05,
+      "loss": 0.1277,
+      "step": 2070
+    },
+    {
+      "epoch": 0.19693239916682448,
+      "grad_norm": 0.3177919089794159,
+      "learning_rate": 1.9212270403332705e-05,
+      "loss": 0.1205,
+      "step": 2080
+    },
+    {
+      "epoch": 0.1978791895474342,
+      "grad_norm": 0.35728374123573303,
+      "learning_rate": 1.9208483241810265e-05,
+      "loss": 0.1345,
+      "step": 2090
+    },
+    {
+      "epoch": 0.19882597992804393,
+      "grad_norm": 0.3154493570327759,
+      "learning_rate": 1.9204696080287825e-05,
+      "loss": 0.1249,
+      "step": 2100
+    },
+    {
+      "epoch": 0.19977277030865367,
+      "grad_norm": 0.28626155853271484,
+      "learning_rate": 1.9200908918765386e-05,
+      "loss": 0.1241,
+      "step": 2110
+    },
+    {
+      "epoch": 0.2007195606892634,
+      "grad_norm": 0.35578519105911255,
+      "learning_rate": 1.9197121757242946e-05,
+      "loss": 0.1339,
+      "step": 2120
+    },
+    {
+      "epoch": 0.20166635106987313,
+      "grad_norm": 0.2937614917755127,
+      "learning_rate": 1.919333459572051e-05,
+      "loss": 0.1289,
+      "step": 2130
+    },
+    {
+      "epoch": 0.20261314145048287,
+      "grad_norm": 0.30598610639572144,
+      "learning_rate": 1.918954743419807e-05,
+      "loss": 0.1265,
+      "step": 2140
+    },
+    {
+      "epoch": 0.20355993183109258,
+      "grad_norm": 0.2843713164329529,
+      "learning_rate": 1.9185760272675633e-05,
+      "loss": 0.1256,
+      "step": 2150
+    },
+    {
+      "epoch": 0.20450672221170232,
+      "grad_norm": 0.2970130145549774,
+      "learning_rate": 1.9181973111153193e-05,
+      "loss": 0.1184,
+      "step": 2160
+    },
+    {
+      "epoch": 0.20545351259231207,
+      "grad_norm": 0.30623194575309753,
+      "learning_rate": 1.9178185949630756e-05,
+      "loss": 0.1272,
+      "step": 2170
+    },
+    {
+      "epoch": 0.2064003029729218,
+      "grad_norm": 0.3569609224796295,
+      "learning_rate": 1.9174398788108316e-05,
+      "loss": 0.1224,
+      "step": 2180
+    },
+    {
+      "epoch": 0.20734709335353152,
+      "grad_norm": 0.29475313425064087,
+      "learning_rate": 1.9170611626585876e-05,
+      "loss": 0.128,
+      "step": 2190
+    },
+    {
+      "epoch": 0.20829388373414126,
+      "grad_norm": 0.2869347035884857,
+      "learning_rate": 1.9166824465063436e-05,
+      "loss": 0.1289,
+      "step": 2200
+    },
+    {
+      "epoch": 0.209240674114751,
+      "grad_norm": 0.2653000056743622,
+      "learning_rate": 1.9163037303540996e-05,
+      "loss": 0.1174,
+      "step": 2210
+    },
+    {
+      "epoch": 0.21018746449536072,
+      "grad_norm": 0.21829605102539062,
+      "learning_rate": 1.915925014201856e-05,
+      "loss": 0.1203,
+      "step": 2220
+    },
+    {
+      "epoch": 0.21113425487597046,
+      "grad_norm": 0.32572606205940247,
+      "learning_rate": 1.915546298049612e-05,
+      "loss": 0.1249,
+      "step": 2230
+    },
+    {
+      "epoch": 0.2120810452565802,
+      "grad_norm": 0.5544735789299011,
+      "learning_rate": 1.915167581897368e-05,
+      "loss": 0.1258,
+      "step": 2240
+    },
+    {
+      "epoch": 0.21302783563718994,
+      "grad_norm": 0.3420076370239258,
+      "learning_rate": 1.914788865745124e-05,
+      "loss": 0.1225,
+      "step": 2250
+    },
+    {
+      "epoch": 0.21397462601779965,
+      "grad_norm": 0.32641270756721497,
+      "learning_rate": 1.9144101495928803e-05,
+      "loss": 0.1334,
+      "step": 2260
+    },
+    {
+      "epoch": 0.2149214163984094,
+      "grad_norm": 0.3167610168457031,
+      "learning_rate": 1.9140314334406363e-05,
+      "loss": 0.1267,
+      "step": 2270
+    },
+    {
+      "epoch": 0.21586820677901913,
+      "grad_norm": 0.37354180216789246,
+      "learning_rate": 1.9136527172883927e-05,
+      "loss": 0.1408,
+      "step": 2280
+    },
+    {
+      "epoch": 0.21681499715962885,
+      "grad_norm": 0.2738534212112427,
+      "learning_rate": 1.9132740011361487e-05,
+      "loss": 0.1231,
+      "step": 2290
+    },
+    {
+      "epoch": 0.2177617875402386,
+      "grad_norm": 0.3062466084957123,
+      "learning_rate": 1.9128952849839047e-05,
+      "loss": 0.1302,
+      "step": 2300
+    },
+    {
+      "epoch": 0.21870857792084833,
+      "grad_norm": 0.3265305161476135,
+      "learning_rate": 1.912516568831661e-05,
+      "loss": 0.121,
+      "step": 2310
+    },
+    {
+      "epoch": 0.21965536830145807,
+      "grad_norm": 0.3429173231124878,
+      "learning_rate": 1.912137852679417e-05,
+      "loss": 0.1344,
+      "step": 2320
+    },
+    {
+      "epoch": 0.22060215868206778,
+      "grad_norm": 0.2771424949169159,
+      "learning_rate": 1.911759136527173e-05,
+      "loss": 0.1188,
+      "step": 2330
+    },
+    {
+      "epoch": 0.22154894906267752,
+      "grad_norm": 0.29070237278938293,
+      "learning_rate": 1.911380420374929e-05,
+      "loss": 0.1218,
+      "step": 2340
+    },
+    {
+      "epoch": 0.22249573944328727,
+      "grad_norm": 0.2981018126010895,
+      "learning_rate": 1.911001704222685e-05,
+      "loss": 0.1361,
+      "step": 2350
+    },
+    {
+      "epoch": 0.22344252982389698,
+      "grad_norm": 0.3472117781639099,
+      "learning_rate": 1.9106229880704414e-05,
+      "loss": 0.1211,
+      "step": 2360
+    },
+    {
+      "epoch": 0.22438932020450672,
+      "grad_norm": 0.271899551153183,
+      "learning_rate": 1.9102442719181974e-05,
+      "loss": 0.133,
+      "step": 2370
+    },
+    {
+      "epoch": 0.22533611058511646,
+      "grad_norm": 0.34610268473625183,
+      "learning_rate": 1.9098655557659534e-05,
+      "loss": 0.1273,
+      "step": 2380
+    },
+    {
+      "epoch": 0.22628290096572617,
+      "grad_norm": 0.2797755002975464,
+      "learning_rate": 1.9094868396137097e-05,
+      "loss": 0.1243,
+      "step": 2390
+    },
+    {
+      "epoch": 0.22722969134633592,
+      "grad_norm": 0.32911232113838196,
+      "learning_rate": 1.9091081234614657e-05,
+      "loss": 0.1184,
+      "step": 2400
+    },
+    {
+      "epoch": 0.22817648172694566,
+      "grad_norm": 0.27326205372810364,
+      "learning_rate": 1.908729407309222e-05,
+      "loss": 0.1177,
+      "step": 2410
+    },
+    {
+      "epoch": 0.2291232721075554,
+      "grad_norm": 0.2909172475337982,
+      "learning_rate": 1.908350691156978e-05,
+      "loss": 0.1177,
+      "step": 2420
+    },
+    {
+      "epoch": 0.2300700624881651,
+      "grad_norm": 0.2706824839115143,
+      "learning_rate": 1.907971975004734e-05,
+      "loss": 0.1202,
+      "step": 2430
+    },
+    {
+      "epoch": 0.23101685286877485,
+      "grad_norm": 0.31379246711730957,
+      "learning_rate": 1.90759325885249e-05,
+      "loss": 0.1164,
+      "step": 2440
+    },
+    {
+      "epoch": 0.2319636432493846,
+      "grad_norm": 0.33181753754615784,
+      "learning_rate": 1.9072145427002464e-05,
+      "loss": 0.1225,
+      "step": 2450
+    },
+    {
+      "epoch": 0.2329104336299943,
+      "grad_norm": 0.23471438884735107,
+      "learning_rate": 1.9068358265480025e-05,
+      "loss": 0.1273,
+      "step": 2460
+    },
+    {
+      "epoch": 0.23385722401060405,
+      "grad_norm": 0.22987720370292664,
+      "learning_rate": 1.9064571103957585e-05,
+      "loss": 0.1291,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2348040143912138,
+      "grad_norm": 0.26784586906433105,
+      "learning_rate": 1.9060783942435145e-05,
+      "loss": 0.1334,
+      "step": 2480
+    },
+    {
+      "epoch": 0.23575080477182353,
+      "grad_norm": 0.3212006390094757,
+      "learning_rate": 1.9056996780912708e-05,
+      "loss": 0.1257,
+      "step": 2490
+    },
+    {
+      "epoch": 0.23669759515243324,
+      "grad_norm": 0.2957512140274048,
+      "learning_rate": 1.9053209619390268e-05,
+      "loss": 0.137,
+      "step": 2500
+    },
+    {
+      "epoch": 0.23764438553304298,
+      "grad_norm": 0.2556230425834656,
+      "learning_rate": 1.904942245786783e-05,
+      "loss": 0.1283,
+      "step": 2510
+    },
+    {
+      "epoch": 0.23859117591365273,
+      "grad_norm": 0.3201919198036194,
+      "learning_rate": 1.904563529634539e-05,
+      "loss": 0.1261,
+      "step": 2520
+    },
+    {
+      "epoch": 0.23953796629426244,
+      "grad_norm": 0.2649897336959839,
+      "learning_rate": 1.904184813482295e-05,
+      "loss": 0.1234,
+      "step": 2530
+    },
+    {
+      "epoch": 0.24048475667487218,
+      "grad_norm": 0.2947181165218353,
+      "learning_rate": 1.9038060973300512e-05,
+      "loss": 0.1229,
+      "step": 2540
+    },
+    {
+      "epoch": 0.24143154705548192,
+      "grad_norm": 0.2702978253364563,
+      "learning_rate": 1.9034273811778075e-05,
+      "loss": 0.13,
+      "step": 2550
+    },
+    {
+      "epoch": 0.24237833743609166,
+      "grad_norm": 0.2793761193752289,
+      "learning_rate": 1.9030486650255635e-05,
+      "loss": 0.1213,
+      "step": 2560
+    },
+    {
+      "epoch": 0.24332512781670138,
+      "grad_norm": 0.36399927735328674,
+      "learning_rate": 1.9026699488733195e-05,
+      "loss": 0.1341,
+      "step": 2570
+    },
+    {
+      "epoch": 0.24427191819731112,
+      "grad_norm": 0.2572638690471649,
+      "learning_rate": 1.9022912327210755e-05,
+      "loss": 0.1205,
+      "step": 2580
+    },
+    {
+      "epoch": 0.24521870857792086,
+      "grad_norm": 0.5421980023384094,
+      "learning_rate": 1.901912516568832e-05,
+      "loss": 0.1418,
+      "step": 2590
+    },
+    {
+      "epoch": 0.24616549895853057,
+      "grad_norm": 0.3563353419303894,
+      "learning_rate": 1.901533800416588e-05,
+      "loss": 0.1109,
+      "step": 2600
+    },
+    {
+      "epoch": 0.2471122893391403,
+      "grad_norm": 0.2500777244567871,
+      "learning_rate": 1.901155084264344e-05,
+      "loss": 0.1221,
+      "step": 2610
+    },
+    {
+      "epoch": 0.24805907971975005,
+      "grad_norm": 0.2641560137271881,
+      "learning_rate": 1.9007763681121002e-05,
+      "loss": 0.1204,
+      "step": 2620
+    },
+    {
+      "epoch": 0.2490058701003598,
+      "grad_norm": 0.2920982837677002,
+      "learning_rate": 1.9003976519598562e-05,
+      "loss": 0.1281,
+      "step": 2630
+    },
+    {
+      "epoch": 0.2499526604809695,
+      "grad_norm": 0.31924280524253845,
+      "learning_rate": 1.9000189358076126e-05,
+      "loss": 0.1245,
+      "step": 2640
+    },
+    {
+      "epoch": 0.2508994508615792,
+      "grad_norm": 0.30418962240219116,
+      "learning_rate": 1.8996402196553686e-05,
+      "loss": 0.1322,
+      "step": 2650
+    },
+    {
+      "epoch": 0.25184624124218896,
+      "grad_norm": 0.37784865498542786,
+      "learning_rate": 1.8992615035031246e-05,
+      "loss": 0.1239,
+      "step": 2660
+    },
+    {
+      "epoch": 0.2527930316227987,
+      "grad_norm": 0.28060677647590637,
+      "learning_rate": 1.8988827873508806e-05,
+      "loss": 0.1282,
+      "step": 2670
+    },
+    {
+      "epoch": 0.25373982200340844,
+      "grad_norm": 0.3838202953338623,
+      "learning_rate": 1.8985040711986366e-05,
+      "loss": 0.1369,
+      "step": 2680
+    },
+    {
+      "epoch": 0.2546866123840182,
+      "grad_norm": 0.3191165626049042,
+      "learning_rate": 1.898125355046393e-05,
+      "loss": 0.1177,
+      "step": 2690
+    },
+    {
+      "epoch": 0.2556334027646279,
+      "grad_norm": 0.28924256563186646,
+      "learning_rate": 1.897746638894149e-05,
+      "loss": 0.1258,
+      "step": 2700
+    },
+    {
+      "epoch": 0.25658019314523767,
+      "grad_norm": 0.33843380212783813,
+      "learning_rate": 1.897367922741905e-05,
+      "loss": 0.1202,
+      "step": 2710
+    },
+    {
+      "epoch": 0.25752698352584735,
+      "grad_norm": 0.313869446516037,
+      "learning_rate": 1.8969892065896613e-05,
+      "loss": 0.1167,
+      "step": 2720
+    },
+    {
+      "epoch": 0.2584737739064571,
+      "grad_norm": 0.3136950433254242,
+      "learning_rate": 1.8966104904374173e-05,
+      "loss": 0.1229,
+      "step": 2730
+    },
+    {
+      "epoch": 0.25942056428706683,
+      "grad_norm": 0.3564135730266571,
+      "learning_rate": 1.8962317742851733e-05,
+      "loss": 0.1269,
+      "step": 2740
+    },
+    {
+      "epoch": 0.2603673546676766,
+      "grad_norm": 0.28279656171798706,
+      "learning_rate": 1.8958530581329296e-05,
+      "loss": 0.1223,
+      "step": 2750
+    },
+    {
+      "epoch": 0.2613141450482863,
+      "grad_norm": 0.38073378801345825,
+      "learning_rate": 1.8954743419806857e-05,
+      "loss": 0.1264,
+      "step": 2760
+    },
+    {
+      "epoch": 0.26226093542889606,
+      "grad_norm": 0.3409043252468109,
+      "learning_rate": 1.8950956258284417e-05,
+      "loss": 0.1241,
+      "step": 2770
+    },
+    {
+      "epoch": 0.2632077258095058,
+      "grad_norm": 0.3403298854827881,
+      "learning_rate": 1.894716909676198e-05,
+      "loss": 0.1297,
+      "step": 2780
+    },
+    {
+      "epoch": 0.2641545161901155,
+      "grad_norm": 0.29332926869392395,
+      "learning_rate": 1.894338193523954e-05,
+      "loss": 0.127,
+      "step": 2790
+    },
+    {
+      "epoch": 0.2651013065707252,
+      "grad_norm": 0.26315629482269287,
+      "learning_rate": 1.89395947737171e-05,
+      "loss": 0.1265,
+      "step": 2800
+    },
+    {
+      "epoch": 0.26604809695133497,
+      "grad_norm": 0.32936838269233704,
+      "learning_rate": 1.893580761219466e-05,
+      "loss": 0.123,
+      "step": 2810
+    },
+    {
+      "epoch": 0.2669948873319447,
+      "grad_norm": 0.2917657196521759,
+      "learning_rate": 1.8932020450672224e-05,
+      "loss": 0.1061,
+      "step": 2820
+    },
+    {
+      "epoch": 0.26794167771255445,
+      "grad_norm": 0.2700554132461548,
+      "learning_rate": 1.8928233289149784e-05,
+      "loss": 0.1099,
+      "step": 2830
+    },
+    {
+      "epoch": 0.2688884680931642,
+      "grad_norm": 0.29582569003105164,
+      "learning_rate": 1.8924446127627344e-05,
+      "loss": 0.1299,
+      "step": 2840
+    },
+    {
+      "epoch": 0.26983525847377393,
+      "grad_norm": 0.28645405173301697,
+      "learning_rate": 1.8920658966104907e-05,
+      "loss": 0.127,
+      "step": 2850
+    },
+    {
+      "epoch": 0.2707820488543836,
+      "grad_norm": 0.5851916670799255,
+      "learning_rate": 1.8916871804582467e-05,
+      "loss": 0.1162,
+      "step": 2860
+    },
+    {
+      "epoch": 0.27172883923499336,
+      "grad_norm": 0.35853055119514465,
+      "learning_rate": 1.891308464306003e-05,
+      "loss": 0.1322,
+      "step": 2870
+    },
+    {
+      "epoch": 0.2726756296156031,
+      "grad_norm": 0.3862152695655823,
+      "learning_rate": 1.890929748153759e-05,
+      "loss": 0.126,
+      "step": 2880
+    },
+    {
+      "epoch": 0.27362241999621284,
+      "grad_norm": 0.27488529682159424,
+      "learning_rate": 1.890551032001515e-05,
+      "loss": 0.1203,
+      "step": 2890
+    },
+    {
+      "epoch": 0.2745692103768226,
+      "grad_norm": 0.3433137834072113,
+      "learning_rate": 1.890172315849271e-05,
+      "loss": 0.1208,
+      "step": 2900
+    },
+    {
+      "epoch": 0.2755160007574323,
+      "grad_norm": 0.38652151823043823,
+      "learning_rate": 1.889793599697027e-05,
+      "loss": 0.1227,
+      "step": 2910
+    },
+    {
+      "epoch": 0.27646279113804206,
+      "grad_norm": 0.31607943773269653,
+      "learning_rate": 1.8894148835447834e-05,
+      "loss": 0.1151,
+      "step": 2920
+    },
+    {
+      "epoch": 0.27740958151865175,
+      "grad_norm": 0.319286584854126,
+      "learning_rate": 1.8890361673925394e-05,
+      "loss": 0.1263,
+      "step": 2930
+    },
+    {
+      "epoch": 0.2783563718992615,
+      "grad_norm": 0.27554312348365784,
+      "learning_rate": 1.8886574512402954e-05,
+      "loss": 0.113,
+      "step": 2940
+    },
+    {
+      "epoch": 0.27930316227987123,
+      "grad_norm": 0.3367822766304016,
+      "learning_rate": 1.8882787350880514e-05,
+      "loss": 0.1137,
+      "step": 2950
+    },
+    {
+      "epoch": 0.28024995266048097,
+      "grad_norm": 0.3277094066143036,
+      "learning_rate": 1.8879000189358078e-05,
+      "loss": 0.1163,
+      "step": 2960
+    },
+    {
+      "epoch": 0.2811967430410907,
+      "grad_norm": 0.39818716049194336,
+      "learning_rate": 1.8875213027835638e-05,
+      "loss": 0.1318,
+      "step": 2970
+    },
+    {
+      "epoch": 0.28214353342170045,
+      "grad_norm": 0.2829638421535492,
+      "learning_rate": 1.88714258663132e-05,
+      "loss": 0.1228,
+      "step": 2980
+    },
+    {
+      "epoch": 0.2830903238023102,
+      "grad_norm": 0.2840532958507538,
+      "learning_rate": 1.886763870479076e-05,
+      "loss": 0.1181,
+      "step": 2990
+    },
+    {
+      "epoch": 0.2840371141829199,
+      "grad_norm": 0.4064909815788269,
+      "learning_rate": 1.886385154326832e-05,
+      "loss": 0.1163,
+      "step": 3000
+    },
+    {
+      "epoch": 0.2849839045635296,
+      "grad_norm": 0.373701274394989,
+      "learning_rate": 1.8860064381745885e-05,
+      "loss": 0.1209,
+      "step": 3010
+    },
+    {
+      "epoch": 0.28593069494413936,
+      "grad_norm": 0.26867496967315674,
+      "learning_rate": 1.8856277220223445e-05,
+      "loss": 0.1264,
+      "step": 3020
+    },
+    {
+      "epoch": 0.2868774853247491,
+      "grad_norm": 0.3012067675590515,
+      "learning_rate": 1.8852490058701005e-05,
+      "loss": 0.1161,
+      "step": 3030
+    },
+    {
+      "epoch": 0.28782427570535885,
+      "grad_norm": 0.26876571774482727,
+      "learning_rate": 1.8848702897178565e-05,
+      "loss": 0.1156,
+      "step": 3040
+    },
+    {
+      "epoch": 0.2887710660859686,
+      "grad_norm": 0.2857358455657959,
+      "learning_rate": 1.8844915735656125e-05,
+      "loss": 0.1286,
+      "step": 3050
+    },
+    {
+      "epoch": 0.28971785646657827,
+      "grad_norm": 0.2931133210659027,
+      "learning_rate": 1.884112857413369e-05,
+      "loss": 0.1203,
+      "step": 3060
+    },
+    {
+      "epoch": 0.290664646847188,
+      "grad_norm": 0.2736462950706482,
+      "learning_rate": 1.883734141261125e-05,
+      "loss": 0.1189,
+      "step": 3070
+    },
+    {
+      "epoch": 0.29161143722779775,
+      "grad_norm": 0.32691168785095215,
+      "learning_rate": 1.8833554251088812e-05,
+      "loss": 0.1231,
+      "step": 3080
+    },
+    {
+      "epoch": 0.2925582276084075,
+      "grad_norm": 0.3366908133029938,
+      "learning_rate": 1.8829767089566372e-05,
+      "loss": 0.1254,
+      "step": 3090
+    },
+    {
+      "epoch": 0.29350501798901724,
+      "grad_norm": 0.26143720746040344,
+      "learning_rate": 1.8825979928043932e-05,
+      "loss": 0.1268,
+      "step": 3100
+    },
+    {
+      "epoch": 0.294451808369627,
+      "grad_norm": 0.3320617079734802,
+      "learning_rate": 1.8822192766521496e-05,
+      "loss": 0.1234,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2953985987502367,
+      "grad_norm": 0.28845691680908203,
+      "learning_rate": 1.8818405604999056e-05,
+      "loss": 0.1201,
+      "step": 3120
+    },
+    {
+      "epoch": 0.2963453891308464,
+      "grad_norm": 0.31063926219940186,
+      "learning_rate": 1.8814618443476616e-05,
+      "loss": 0.1173,
+      "step": 3130
+    },
+    {
+      "epoch": 0.29729217951145614,
+      "grad_norm": 0.24950078129768372,
+      "learning_rate": 1.8810831281954176e-05,
+      "loss": 0.1283,
+      "step": 3140
+    },
+    {
+      "epoch": 0.2982389698920659,
+      "grad_norm": 0.2675219178199768,
+      "learning_rate": 1.880704412043174e-05,
+      "loss": 0.1175,
+      "step": 3150
+    },
+    {
+      "epoch": 0.2991857602726756,
+      "grad_norm": 0.24839593470096588,
+      "learning_rate": 1.88032569589093e-05,
+      "loss": 0.1228,
+      "step": 3160
+    },
+    {
+      "epoch": 0.30013255065328537,
+      "grad_norm": 0.5048424005508423,
+      "learning_rate": 1.879946979738686e-05,
+      "loss": 0.1129,
+      "step": 3170
+    },
+    {
+      "epoch": 0.3010793410338951,
+      "grad_norm": 0.34761127829551697,
+      "learning_rate": 1.879568263586442e-05,
+      "loss": 0.1219,
+      "step": 3180
+    },
+    {
+      "epoch": 0.30202613141450485,
+      "grad_norm": 0.3296762704849243,
+      "learning_rate": 1.8791895474341983e-05,
+      "loss": 0.131,
+      "step": 3190
+    },
+    {
+      "epoch": 0.30297292179511454,
+      "grad_norm": 0.3156794011592865,
+      "learning_rate": 1.8788108312819543e-05,
+      "loss": 0.1151,
+      "step": 3200
+    },
+    {
+      "epoch": 0.3039197121757243,
+      "grad_norm": 0.29802405834198,
+      "learning_rate": 1.8784321151297106e-05,
+      "loss": 0.1267,
+      "step": 3210
+    },
+    {
+      "epoch": 0.304866502556334,
+      "grad_norm": 0.285553902387619,
+      "learning_rate": 1.8780533989774666e-05,
+      "loss": 0.1202,
+      "step": 3220
+    },
+    {
+      "epoch": 0.30581329293694376,
+      "grad_norm": 0.29214930534362793,
+      "learning_rate": 1.8776746828252226e-05,
+      "loss": 0.1183,
+      "step": 3230
+    },
+    {
+      "epoch": 0.3067600833175535,
+      "grad_norm": 0.3107370138168335,
+      "learning_rate": 1.877295966672979e-05,
+      "loss": 0.1187,
+      "step": 3240
+    },
+    {
+      "epoch": 0.30770687369816324,
+      "grad_norm": 0.3791562020778656,
+      "learning_rate": 1.876917250520735e-05,
+      "loss": 0.119,
+      "step": 3250
+    },
+    {
+      "epoch": 0.308653664078773,
+      "grad_norm": 0.3186882734298706,
+      "learning_rate": 1.876538534368491e-05,
+      "loss": 0.1164,
+      "step": 3260
+    },
+    {
+      "epoch": 0.30960045445938267,
+      "grad_norm": 0.3271386921405792,
+      "learning_rate": 1.876159818216247e-05,
+      "loss": 0.1066,
+      "step": 3270
+    },
+    {
+      "epoch": 0.3105472448399924,
+      "grad_norm": 0.27886903285980225,
+      "learning_rate": 1.875781102064003e-05,
+      "loss": 0.1237,
+      "step": 3280
+    },
+    {
+      "epoch": 0.31149403522060215,
+      "grad_norm": 0.2511722147464752,
+      "learning_rate": 1.8754023859117593e-05,
+      "loss": 0.128,
+      "step": 3290
+    },
+    {
+      "epoch": 0.3124408256012119,
+      "grad_norm": 0.3614468276500702,
+      "learning_rate": 1.8750236697595153e-05,
+      "loss": 0.1192,
+      "step": 3300
+    },
+    {
+      "epoch": 0.31338761598182163,
+      "grad_norm": 0.3021714687347412,
+      "learning_rate": 1.8746449536072713e-05,
+      "loss": 0.1235,
+      "step": 3310
+    },
+    {
+      "epoch": 0.3143344063624314,
+      "grad_norm": 0.31286343932151794,
+      "learning_rate": 1.8742662374550277e-05,
+      "loss": 0.1263,
+      "step": 3320
+    },
+    {
+      "epoch": 0.3152811967430411,
+      "grad_norm": 0.31820154190063477,
+      "learning_rate": 1.8738875213027837e-05,
+      "loss": 0.1271,
+      "step": 3330
+    },
+    {
+      "epoch": 0.3162279871236508,
+      "grad_norm": 0.3829352557659149,
+      "learning_rate": 1.87350880515054e-05,
+      "loss": 0.1096,
+      "step": 3340
+    },
+    {
+      "epoch": 0.31717477750426054,
+      "grad_norm": 0.24418611824512482,
+      "learning_rate": 1.873130088998296e-05,
+      "loss": 0.1299,
+      "step": 3350
+    },
+    {
+      "epoch": 0.3181215678848703,
+      "grad_norm": 0.25075456500053406,
+      "learning_rate": 1.872751372846052e-05,
+      "loss": 0.111,
+      "step": 3360
+    },
+    {
+      "epoch": 0.31906835826548,
+      "grad_norm": 0.41566210985183716,
+      "learning_rate": 1.872372656693808e-05,
+      "loss": 0.1201,
+      "step": 3370
+    },
+    {
+      "epoch": 0.32001514864608976,
+      "grad_norm": 0.32048362493515015,
+      "learning_rate": 1.8719939405415644e-05,
+      "loss": 0.1253,
+      "step": 3380
+    },
+    {
+      "epoch": 0.3209619390266995,
+      "grad_norm": 0.3619791567325592,
+      "learning_rate": 1.8716152243893204e-05,
+      "loss": 0.126,
+      "step": 3390
+    },
+    {
+      "epoch": 0.32190872940730925,
+      "grad_norm": 0.3619197905063629,
+      "learning_rate": 1.8712365082370764e-05,
+      "loss": 0.1155,
+      "step": 3400
+    },
+    {
+      "epoch": 0.32285551978791893,
+      "grad_norm": 0.29323554039001465,
+      "learning_rate": 1.8708577920848324e-05,
+      "loss": 0.1118,
+      "step": 3410
+    },
+    {
+      "epoch": 0.3238023101685287,
+      "grad_norm": 0.3184221386909485,
+      "learning_rate": 1.8704790759325888e-05,
+      "loss": 0.1265,
+      "step": 3420
+    },
+    {
+      "epoch": 0.3247491005491384,
+      "grad_norm": 0.33619818091392517,
+      "learning_rate": 1.8701003597803448e-05,
+      "loss": 0.1219,
+      "step": 3430
+    },
+    {
+      "epoch": 0.32569589092974816,
+      "grad_norm": 0.3194987177848816,
+      "learning_rate": 1.869721643628101e-05,
+      "loss": 0.1231,
+      "step": 3440
+    },
+    {
+      "epoch": 0.3266426813103579,
+      "grad_norm": 0.239166721701622,
+      "learning_rate": 1.869342927475857e-05,
+      "loss": 0.1021,
+      "step": 3450
+    },
+    {
+      "epoch": 0.32758947169096764,
+      "grad_norm": 0.315458208322525,
+      "learning_rate": 1.868964211323613e-05,
+      "loss": 0.1284,
+      "step": 3460
+    },
+    {
+      "epoch": 0.3285362620715774,
+      "grad_norm": 0.3272011876106262,
+      "learning_rate": 1.868585495171369e-05,
+      "loss": 0.1158,
+      "step": 3470
+    },
+    {
+      "epoch": 0.32948305245218706,
+      "grad_norm": 0.38707324862480164,
+      "learning_rate": 1.8682067790191255e-05,
+      "loss": 0.1166,
+      "step": 3480
+    },
+    {
+      "epoch": 0.3304298428327968,
+      "grad_norm": 0.3211500644683838,
+      "learning_rate": 1.8678280628668815e-05,
+      "loss": 0.129,
+      "step": 3490
+    },
+    {
+      "epoch": 0.33137663321340655,
+      "grad_norm": 0.35265690088272095,
+      "learning_rate": 1.8674493467146375e-05,
+      "loss": 0.1185,
+      "step": 3500
+    },
+    {
+      "epoch": 0.3323234235940163,
+      "grad_norm": 0.27759236097335815,
+      "learning_rate": 1.8670706305623935e-05,
+      "loss": 0.1313,
+      "step": 3510
+    },
+    {
+      "epoch": 0.33327021397462603,
+      "grad_norm": 0.3936869502067566,
+      "learning_rate": 1.8666919144101498e-05,
+      "loss": 0.1238,
+      "step": 3520
+    },
+    {
+      "epoch": 0.33421700435523577,
+      "grad_norm": 0.2950145900249481,
+      "learning_rate": 1.8663131982579058e-05,
+      "loss": 0.1189,
+      "step": 3530
+    },
+    {
+      "epoch": 0.3351637947358455,
+      "grad_norm": 0.46038469672203064,
+      "learning_rate": 1.8659344821056618e-05,
+      "loss": 0.1254,
+      "step": 3540
+    },
+    {
+      "epoch": 0.3361105851164552,
+      "grad_norm": 0.3949941098690033,
+      "learning_rate": 1.8655557659534182e-05,
+      "loss": 0.1166,
+      "step": 3550
+    },
+    {
+      "epoch": 0.33705737549706494,
+      "grad_norm": 0.218438521027565,
+      "learning_rate": 1.8651770498011742e-05,
+      "loss": 0.1096,
+      "step": 3560
+    },
+    {
+      "epoch": 0.3380041658776747,
+      "grad_norm": 0.34671449661254883,
+      "learning_rate": 1.8647983336489305e-05,
+      "loss": 0.1163,
+      "step": 3570
+    },
+    {
+      "epoch": 0.3389509562582844,
+      "grad_norm": 0.29920464754104614,
+      "learning_rate": 1.8644196174966865e-05,
+      "loss": 0.1261,
+      "step": 3580
+    },
+    {
+      "epoch": 0.33989774663889416,
+      "grad_norm": 0.2960605025291443,
+      "learning_rate": 1.8640409013444425e-05,
+      "loss": 0.1181,
+      "step": 3590
+    },
+    {
+      "epoch": 0.3408445370195039,
+      "grad_norm": 0.38953059911727905,
+      "learning_rate": 1.8636621851921985e-05,
+      "loss": 0.1218,
+      "step": 3600
+    },
+    {
+      "epoch": 0.34179132740011364,
+      "grad_norm": 0.2678033709526062,
+      "learning_rate": 1.8632834690399545e-05,
+      "loss": 0.114,
+      "step": 3610
+    },
+    {
+      "epoch": 0.34273811778072333,
+      "grad_norm": 0.3713693916797638,
+      "learning_rate": 1.862904752887711e-05,
+      "loss": 0.1207,
+      "step": 3620
+    },
+    {
+      "epoch": 0.34368490816133307,
+      "grad_norm": 0.43956953287124634,
+      "learning_rate": 1.862526036735467e-05,
+      "loss": 0.1235,
+      "step": 3630
+    },
+    {
+      "epoch": 0.3446316985419428,
+      "grad_norm": 0.30412259697914124,
+      "learning_rate": 1.862147320583223e-05,
+      "loss": 0.1205,
+      "step": 3640
+    },
+    {
+      "epoch": 0.34557848892255255,
+      "grad_norm": 0.3188311755657196,
+      "learning_rate": 1.861768604430979e-05,
+      "loss": 0.1262,
+      "step": 3650
+    },
+    {
+      "epoch": 0.3465252793031623,
+      "grad_norm": 0.3703117072582245,
+      "learning_rate": 1.8613898882787352e-05,
+      "loss": 0.1157,
+      "step": 3660
+    },
+    {
+      "epoch": 0.34747206968377203,
+      "grad_norm": 0.314065158367157,
+      "learning_rate": 1.8610111721264912e-05,
+      "loss": 0.1174,
+      "step": 3670
+    },
+    {
+      "epoch": 0.3484188600643817,
+      "grad_norm": 0.3832274377346039,
+      "learning_rate": 1.8606324559742476e-05,
+      "loss": 0.1245,
+      "step": 3680
+    },
+    {
+      "epoch": 0.34936565044499146,
+      "grad_norm": 0.3195415139198303,
+      "learning_rate": 1.8602537398220036e-05,
+      "loss": 0.1175,
+      "step": 3690
+    },
+    {
+      "epoch": 0.3503124408256012,
+      "grad_norm": 0.2775132656097412,
+      "learning_rate": 1.8598750236697596e-05,
+      "loss": 0.1166,
+      "step": 3700
+    },
+    {
+      "epoch": 0.35125923120621094,
+      "grad_norm": 0.43748944997787476,
+      "learning_rate": 1.859496307517516e-05,
+      "loss": 0.1247,
+      "step": 3710
+    },
+    {
+      "epoch": 0.3522060215868207,
+      "grad_norm": 0.313702791929245,
+      "learning_rate": 1.859117591365272e-05,
+      "loss": 0.1267,
+      "step": 3720
+    },
+    {
+      "epoch": 0.3531528119674304,
+      "grad_norm": 0.37444302439689636,
+      "learning_rate": 1.858738875213028e-05,
+      "loss": 0.118,
+      "step": 3730
+    },
+    {
+      "epoch": 0.35409960234804017,
+      "grad_norm": 0.32261762022972107,
+      "learning_rate": 1.858360159060784e-05,
+      "loss": 0.1237,
+      "step": 3740
+    },
+    {
+      "epoch": 0.35504639272864985,
+      "grad_norm": 0.2619532346725464,
+      "learning_rate": 1.85798144290854e-05,
+      "loss": 0.1119,
+      "step": 3750
+    },
+    {
+      "epoch": 0.3559931831092596,
+      "grad_norm": 0.31362977623939514,
+      "learning_rate": 1.8576027267562963e-05,
+      "loss": 0.115,
+      "step": 3760
+    },
+    {
+      "epoch": 0.35693997348986933,
+      "grad_norm": 0.34366849064826965,
+      "learning_rate": 1.8572240106040523e-05,
+      "loss": 0.124,
+      "step": 3770
+    },
+    {
+      "epoch": 0.3578867638704791,
+      "grad_norm": 0.31962850689888,
+      "learning_rate": 1.8568452944518087e-05,
+      "loss": 0.1269,
+      "step": 3780
+    },
+    {
+      "epoch": 0.3588335542510888,
+      "grad_norm": 0.35521697998046875,
+      "learning_rate": 1.8564665782995647e-05,
+      "loss": 0.129,
+      "step": 3790
+    },
+    {
+      "epoch": 0.35978034463169856,
+      "grad_norm": 0.37336549162864685,
+      "learning_rate": 1.856087862147321e-05,
+      "loss": 0.1244,
+      "step": 3800
+    },
+    {
+      "epoch": 0.3607271350123083,
+      "grad_norm": 0.3989594578742981,
+      "learning_rate": 1.855709145995077e-05,
+      "loss": 0.1272,
+      "step": 3810
+    },
+    {
+      "epoch": 0.361673925392918,
+      "grad_norm": 0.2903200387954712,
+      "learning_rate": 1.855330429842833e-05,
+      "loss": 0.1198,
+      "step": 3820
+    },
+    {
+      "epoch": 0.3626207157735277,
+      "grad_norm": 0.2967371940612793,
+      "learning_rate": 1.854951713690589e-05,
+      "loss": 0.1171,
+      "step": 3830
+    },
+    {
+      "epoch": 0.36356750615413747,
+      "grad_norm": 0.33660125732421875,
+      "learning_rate": 1.854572997538345e-05,
+      "loss": 0.1186,
+      "step": 3840
+    },
+    {
+      "epoch": 0.3645142965347472,
+      "grad_norm": 0.27733585238456726,
+      "learning_rate": 1.8541942813861014e-05,
+      "loss": 0.1234,
+      "step": 3850
+    },
+    {
+      "epoch": 0.36546108691535695,
+      "grad_norm": 0.3161308169364929,
+      "learning_rate": 1.8538155652338574e-05,
+      "loss": 0.119,
+      "step": 3860
+    },
+    {
+      "epoch": 0.3664078772959667,
+      "grad_norm": 0.2673567533493042,
+      "learning_rate": 1.8534368490816134e-05,
+      "loss": 0.1249,
+      "step": 3870
+    },
+    {
+      "epoch": 0.36735466767657643,
+      "grad_norm": 0.2674126625061035,
+      "learning_rate": 1.8530581329293694e-05,
+      "loss": 0.1212,
+      "step": 3880
+    },
+    {
+      "epoch": 0.3683014580571861,
+      "grad_norm": 0.4858635663986206,
+      "learning_rate": 1.8526794167771257e-05,
+      "loss": 0.123,
+      "step": 3890
+    },
+    {
+      "epoch": 0.36924824843779586,
+      "grad_norm": 0.35940200090408325,
+      "learning_rate": 1.8523007006248817e-05,
+      "loss": 0.1192,
+      "step": 3900
+    },
+    {
+      "epoch": 0.3701950388184056,
+      "grad_norm": 0.3293863534927368,
+      "learning_rate": 1.851921984472638e-05,
+      "loss": 0.1125,
+      "step": 3910
+    },
+    {
+      "epoch": 0.37114182919901534,
+      "grad_norm": 0.44395145773887634,
+      "learning_rate": 1.851543268320394e-05,
+      "loss": 0.1143,
+      "step": 3920
+    },
+    {
+      "epoch": 0.3720886195796251,
+      "grad_norm": 0.317997545003891,
+      "learning_rate": 1.85116455216815e-05,
+      "loss": 0.1155,
+      "step": 3930
+    },
+    {
+      "epoch": 0.3730354099602348,
+      "grad_norm": 0.3860480785369873,
+      "learning_rate": 1.8507858360159064e-05,
+      "loss": 0.1309,
+      "step": 3940
+    },
+    {
+      "epoch": 0.37398220034084456,
+      "grad_norm": 0.2992136478424072,
+      "learning_rate": 1.8504071198636624e-05,
+      "loss": 0.1177,
+      "step": 3950
+    },
+    {
+      "epoch": 0.37492899072145425,
+      "grad_norm": 0.2919655442237854,
+      "learning_rate": 1.8500284037114184e-05,
+      "loss": 0.1184,
+      "step": 3960
+    },
+    {
+      "epoch": 0.375875781102064,
+      "grad_norm": 0.3052353262901306,
+      "learning_rate": 1.8496496875591744e-05,
+      "loss": 0.1317,
+      "step": 3970
+    },
+    {
+      "epoch": 0.37682257148267373,
+      "grad_norm": 0.3412197530269623,
+      "learning_rate": 1.8492709714069305e-05,
+      "loss": 0.1256,
+      "step": 3980
+    },
+    {
+      "epoch": 0.37776936186328347,
+      "grad_norm": 0.31510818004608154,
+      "learning_rate": 1.8488922552546868e-05,
+      "loss": 0.1226,
+      "step": 3990
+    },
+    {
+      "epoch": 0.3787161522438932,
+      "grad_norm": 0.3692219853401184,
+      "learning_rate": 1.8485135391024428e-05,
+      "loss": 0.13,
+      "step": 4000
+    },
+    {
+      "epoch": 0.37966294262450295,
+      "grad_norm": 0.3020903766155243,
+      "learning_rate": 1.8481348229501988e-05,
+      "loss": 0.1188,
+      "step": 4010
+    },
+    {
+      "epoch": 0.3806097330051127,
+      "grad_norm": 0.36188632249832153,
+      "learning_rate": 1.847756106797955e-05,
+      "loss": 0.1203,
+      "step": 4020
+    },
+    {
+      "epoch": 0.3815565233857224,
+      "grad_norm": 0.3624081313610077,
+      "learning_rate": 1.847377390645711e-05,
+      "loss": 0.1289,
+      "step": 4030
+    },
+    {
+      "epoch": 0.3825033137663321,
+      "grad_norm": 0.2747180461883545,
+      "learning_rate": 1.8469986744934675e-05,
+      "loss": 0.1174,
+      "step": 4040
+    },
+    {
+      "epoch": 0.38345010414694186,
+      "grad_norm": 0.3396485149860382,
+      "learning_rate": 1.8466199583412235e-05,
+      "loss": 0.119,
+      "step": 4050
+    },
+    {
+      "epoch": 0.3843968945275516,
+      "grad_norm": 0.3224307894706726,
+      "learning_rate": 1.8462412421889795e-05,
+      "loss": 0.1141,
+      "step": 4060
+    },
+    {
+      "epoch": 0.38534368490816134,
+      "grad_norm": 0.30942174792289734,
+      "learning_rate": 1.8458625260367355e-05,
+      "loss": 0.1165,
+      "step": 4070
+    },
+    {
+      "epoch": 0.3862904752887711,
+      "grad_norm": 0.2940969467163086,
+      "learning_rate": 1.845483809884492e-05,
+      "loss": 0.1162,
+      "step": 4080
+    },
+    {
+      "epoch": 0.3872372656693808,
+      "grad_norm": 0.35114723443984985,
+      "learning_rate": 1.845105093732248e-05,
+      "loss": 0.1306,
+      "step": 4090
+    },
+    {
+      "epoch": 0.3881840560499905,
+      "grad_norm": 0.42864593863487244,
+      "learning_rate": 1.844726377580004e-05,
+      "loss": 0.1226,
+      "step": 4100
+    },
+    {
+      "epoch": 0.38913084643060025,
+      "grad_norm": 0.3231365978717804,
+      "learning_rate": 1.84434766142776e-05,
+      "loss": 0.1194,
+      "step": 4110
+    },
+    {
+      "epoch": 0.39007763681121,
+      "grad_norm": 0.30363401770591736,
+      "learning_rate": 1.8439689452755162e-05,
+      "loss": 0.1075,
+      "step": 4120
+    },
+    {
+      "epoch": 0.39102442719181973,
+      "grad_norm": 0.39980143308639526,
+      "learning_rate": 1.8435902291232722e-05,
+      "loss": 0.1149,
+      "step": 4130
+    },
+    {
+      "epoch": 0.3919712175724295,
+      "grad_norm": 0.37222790718078613,
+      "learning_rate": 1.8432115129710286e-05,
+      "loss": 0.117,
+      "step": 4140
+    },
+    {
+      "epoch": 0.3929180079530392,
+      "grad_norm": 0.29298582673072815,
+      "learning_rate": 1.8428327968187846e-05,
+      "loss": 0.125,
+      "step": 4150
+    },
+    {
+      "epoch": 0.39386479833364896,
+      "grad_norm": 0.339240700006485,
+      "learning_rate": 1.8424540806665406e-05,
+      "loss": 0.1182,
+      "step": 4160
+    },
+    {
+      "epoch": 0.39481158871425864,
+      "grad_norm": 0.33925309777259827,
+      "learning_rate": 1.8420753645142966e-05,
+      "loss": 0.1227,
+      "step": 4170
+    },
+    {
+      "epoch": 0.3957583790948684,
+      "grad_norm": 0.2940181791782379,
+      "learning_rate": 1.841696648362053e-05,
+      "loss": 0.1207,
+      "step": 4180
+    },
+    {
+      "epoch": 0.3967051694754781,
+      "grad_norm": 0.27949029207229614,
+      "learning_rate": 1.841317932209809e-05,
+      "loss": 0.121,
+      "step": 4190
+    },
+    {
+      "epoch": 0.39765195985608787,
+      "grad_norm": 0.2940124571323395,
+      "learning_rate": 1.840939216057565e-05,
+      "loss": 0.1181,
+      "step": 4200
+    },
+    {
+      "epoch": 0.3985987502366976,
+      "grad_norm": 0.31642553210258484,
+      "learning_rate": 1.840560499905321e-05,
+      "loss": 0.125,
+      "step": 4210
+    },
+    {
+      "epoch": 0.39954554061730735,
+      "grad_norm": 0.23605939745903015,
+      "learning_rate": 1.8401817837530773e-05,
+      "loss": 0.1207,
+      "step": 4220
+    },
+    {
+      "epoch": 0.40049233099791703,
+      "grad_norm": 0.36963286995887756,
+      "learning_rate": 1.8398030676008333e-05,
+      "loss": 0.1217,
+      "step": 4230
+    },
+    {
+      "epoch": 0.4014391213785268,
+      "grad_norm": 0.254743754863739,
+      "learning_rate": 1.8394243514485893e-05,
+      "loss": 0.1243,
+      "step": 4240
+    },
+    {
+      "epoch": 0.4023859117591365,
+      "grad_norm": 0.326809287071228,
+      "learning_rate": 1.8390456352963456e-05,
+      "loss": 0.1242,
+      "step": 4250
+    },
+    {
+      "epoch": 0.40333270213974626,
+      "grad_norm": 0.3072037100791931,
+      "learning_rate": 1.8386669191441016e-05,
+      "loss": 0.1195,
+      "step": 4260
+    },
+    {
+      "epoch": 0.404279492520356,
+      "grad_norm": 0.37460359930992126,
+      "learning_rate": 1.838288202991858e-05,
+      "loss": 0.1114,
+      "step": 4270
+    },
+    {
+      "epoch": 0.40522628290096574,
+      "grad_norm": 0.24516154825687408,
+      "learning_rate": 1.837909486839614e-05,
+      "loss": 0.1088,
+      "step": 4280
+    },
+    {
+      "epoch": 0.4061730732815755,
+      "grad_norm": 0.3073451519012451,
+      "learning_rate": 1.83753077068737e-05,
+      "loss": 0.1094,
+      "step": 4290
+    },
+    {
+      "epoch": 0.40711986366218517,
+      "grad_norm": 0.3263719379901886,
+      "learning_rate": 1.837152054535126e-05,
+      "loss": 0.1217,
+      "step": 4300
+    },
+    {
+      "epoch": 0.4080666540427949,
+      "grad_norm": 0.3760577142238617,
+      "learning_rate": 1.836773338382882e-05,
+      "loss": 0.1236,
+      "step": 4310
+    },
+    {
+      "epoch": 0.40901344442340465,
+      "grad_norm": 0.30230146646499634,
+      "learning_rate": 1.8363946222306383e-05,
+      "loss": 0.1188,
+      "step": 4320
+    },
+    {
+      "epoch": 0.4099602348040144,
+      "grad_norm": 0.4337088465690613,
+      "learning_rate": 1.8360159060783944e-05,
+      "loss": 0.1251,
+      "step": 4330
+    },
+    {
+      "epoch": 0.41090702518462413,
+      "grad_norm": 0.35297223925590515,
+      "learning_rate": 1.8356371899261504e-05,
+      "loss": 0.1171,
+      "step": 4340
+    },
+    {
+      "epoch": 0.41185381556523387,
+      "grad_norm": 0.3118896186351776,
+      "learning_rate": 1.8352584737739067e-05,
+      "loss": 0.1216,
+      "step": 4350
+    },
+    {
+      "epoch": 0.4128006059458436,
+      "grad_norm": 0.47326090931892395,
+      "learning_rate": 1.8348797576216627e-05,
+      "loss": 0.1247,
+      "step": 4360
+    },
+    {
+      "epoch": 0.4137473963264533,
+      "grad_norm": 0.4229494035243988,
+      "learning_rate": 1.8345010414694187e-05,
+      "loss": 0.116,
+      "step": 4370
+    },
+    {
+      "epoch": 0.41469418670706304,
+      "grad_norm": 0.3024285137653351,
+      "learning_rate": 1.834122325317175e-05,
+      "loss": 0.1235,
+      "step": 4380
+    },
+    {
+      "epoch": 0.4156409770876728,
+      "grad_norm": 0.3133857548236847,
+      "learning_rate": 1.833743609164931e-05,
+      "loss": 0.1137,
+      "step": 4390
+    },
+    {
+      "epoch": 0.4165877674682825,
+      "grad_norm": 0.2643986940383911,
+      "learning_rate": 1.833364893012687e-05,
+      "loss": 0.1218,
+      "step": 4400
+    },
+    {
+      "epoch": 0.41753455784889226,
+      "grad_norm": 0.3025350272655487,
+      "learning_rate": 1.8329861768604434e-05,
+      "loss": 0.1174,
+      "step": 4410
+    },
+    {
+      "epoch": 0.418481348229502,
+      "grad_norm": 0.31136399507522583,
+      "learning_rate": 1.8326074607081994e-05,
+      "loss": 0.1197,
+      "step": 4420
+    },
+    {
+      "epoch": 0.41942813861011174,
+      "grad_norm": 0.3025195896625519,
+      "learning_rate": 1.8322287445559554e-05,
+      "loss": 0.1119,
+      "step": 4430
+    },
+    {
+      "epoch": 0.42037492899072143,
+      "grad_norm": 0.23095518350601196,
+      "learning_rate": 1.8318500284037114e-05,
+      "loss": 0.1224,
+      "step": 4440
+    },
+    {
+      "epoch": 0.42132171937133117,
+      "grad_norm": 0.45459040999412537,
+      "learning_rate": 1.8314713122514674e-05,
+      "loss": 0.1166,
+      "step": 4450
+    },
+    {
+      "epoch": 0.4222685097519409,
+      "grad_norm": 0.39385083317756653,
+      "learning_rate": 1.8310925960992238e-05,
+      "loss": 0.1238,
+      "step": 4460
+    },
+    {
+      "epoch": 0.42321530013255065,
+      "grad_norm": 0.2855985462665558,
+      "learning_rate": 1.8307138799469798e-05,
+      "loss": 0.1288,
+      "step": 4470
+    },
+    {
+      "epoch": 0.4241620905131604,
+      "grad_norm": 0.29849758744239807,
+      "learning_rate": 1.830335163794736e-05,
+      "loss": 0.1121,
+      "step": 4480
+    },
+    {
+      "epoch": 0.42510888089377014,
+      "grad_norm": 0.40433022379875183,
+      "learning_rate": 1.829956447642492e-05,
+      "loss": 0.1129,
+      "step": 4490
+    },
+    {
+      "epoch": 0.4260556712743799,
+      "grad_norm": 0.29155251383781433,
+      "learning_rate": 1.8295777314902485e-05,
+      "loss": 0.1194,
+      "step": 4500
+    },
+    {
+      "epoch": 0.42700246165498956,
+      "grad_norm": 0.2591720223426819,
+      "learning_rate": 1.8291990153380045e-05,
+      "loss": 0.115,
+      "step": 4510
+    },
+    {
+      "epoch": 0.4279492520355993,
+      "grad_norm": 0.3257406949996948,
+      "learning_rate": 1.8288202991857605e-05,
+      "loss": 0.1137,
+      "step": 4520
+    },
+    {
+      "epoch": 0.42889604241620904,
+      "grad_norm": 0.3112312853336334,
+      "learning_rate": 1.8284415830335165e-05,
+      "loss": 0.1228,
+      "step": 4530
+    },
+    {
+      "epoch": 0.4298428327968188,
+      "grad_norm": 0.2938838601112366,
+      "learning_rate": 1.8280628668812725e-05,
+      "loss": 0.1111,
+      "step": 4540
+    },
+    {
+      "epoch": 0.4307896231774285,
+      "grad_norm": 0.3180755376815796,
+      "learning_rate": 1.827684150729029e-05,
+      "loss": 0.1355,
+      "step": 4550
+    },
+    {
+      "epoch": 0.43173641355803827,
+      "grad_norm": 0.39242005348205566,
+      "learning_rate": 1.827305434576785e-05,
+      "loss": 0.12,
+      "step": 4560
+    },
+    {
+      "epoch": 0.432683203938648,
+      "grad_norm": 0.3311753273010254,
+      "learning_rate": 1.826926718424541e-05,
+      "loss": 0.1233,
+      "step": 4570
+    },
+    {
+      "epoch": 0.4336299943192577,
+      "grad_norm": 0.26584166288375854,
+      "learning_rate": 1.826548002272297e-05,
+      "loss": 0.1149,
+      "step": 4580
+    },
+    {
+      "epoch": 0.43457678469986744,
+      "grad_norm": 0.282243549823761,
+      "learning_rate": 1.8261692861200532e-05,
+      "loss": 0.1295,
+      "step": 4590
+    },
+    {
+      "epoch": 0.4355235750804772,
+      "grad_norm": 0.4552994668483734,
+      "learning_rate": 1.8257905699678092e-05,
+      "loss": 0.1123,
+      "step": 4600
+    },
+    {
+      "epoch": 0.4364703654610869,
+      "grad_norm": 0.2721503674983978,
+      "learning_rate": 1.8254118538155655e-05,
+      "loss": 0.117,
+      "step": 4610
+    },
+    {
+      "epoch": 0.43741715584169666,
+      "grad_norm": 0.30223289132118225,
+      "learning_rate": 1.8250331376633215e-05,
+      "loss": 0.1135,
+      "step": 4620
+    },
+    {
+      "epoch": 0.4383639462223064,
+      "grad_norm": 0.3740018904209137,
+      "learning_rate": 1.8246544215110776e-05,
+      "loss": 0.1199,
+      "step": 4630
+    },
+    {
+      "epoch": 0.43931073660291614,
+      "grad_norm": 0.25978952646255493,
+      "learning_rate": 1.824275705358834e-05,
+      "loss": 0.1128,
+      "step": 4640
+    },
+    {
+      "epoch": 0.4402575269835258,
+      "grad_norm": 0.2803758680820465,
+      "learning_rate": 1.82389698920659e-05,
+      "loss": 0.1174,
+      "step": 4650
+    },
+    {
+      "epoch": 0.44120431736413557,
+      "grad_norm": 0.2670730948448181,
+      "learning_rate": 1.823518273054346e-05,
+      "loss": 0.1126,
+      "step": 4660
+    },
+    {
+      "epoch": 0.4421511077447453,
+      "grad_norm": 0.22530563175678253,
+      "learning_rate": 1.823139556902102e-05,
+      "loss": 0.1259,
+      "step": 4670
+    },
+    {
+      "epoch": 0.44309789812535505,
+      "grad_norm": 0.3665080964565277,
+      "learning_rate": 1.822760840749858e-05,
+      "loss": 0.1191,
+      "step": 4680
+    },
+    {
+      "epoch": 0.4440446885059648,
+      "grad_norm": 0.2968369722366333,
+      "learning_rate": 1.8223821245976143e-05,
+      "loss": 0.1222,
+      "step": 4690
+    },
+    {
+      "epoch": 0.44499147888657453,
+      "grad_norm": 0.38494449853897095,
+      "learning_rate": 1.8220034084453703e-05,
+      "loss": 0.1204,
+      "step": 4700
+    },
+    {
+      "epoch": 0.4459382692671843,
+      "grad_norm": 0.3491016924381256,
+      "learning_rate": 1.8216246922931266e-05,
+      "loss": 0.1261,
+      "step": 4710
+    },
+    {
+      "epoch": 0.44688505964779396,
+      "grad_norm": 0.5060954689979553,
+      "learning_rate": 1.8212459761408826e-05,
+      "loss": 0.125,
+      "step": 4720
+    },
+    {
+      "epoch": 0.4478318500284037,
+      "grad_norm": 0.36539798974990845,
+      "learning_rate": 1.8208672599886386e-05,
+      "loss": 0.1177,
+      "step": 4730
+    },
+    {
+      "epoch": 0.44877864040901344,
+      "grad_norm": 0.30213627219200134,
+      "learning_rate": 1.820488543836395e-05,
+      "loss": 0.1146,
+      "step": 4740
+    },
+    {
+      "epoch": 0.4497254307896232,
+      "grad_norm": 0.3161025047302246,
+      "learning_rate": 1.820109827684151e-05,
+      "loss": 0.1229,
+      "step": 4750
+    },
+    {
+      "epoch": 0.4506722211702329,
+      "grad_norm": 0.3496663272380829,
+      "learning_rate": 1.819731111531907e-05,
+      "loss": 0.1175,
+      "step": 4760
+    },
+    {
+      "epoch": 0.45161901155084266,
+      "grad_norm": 0.4112556576728821,
+      "learning_rate": 1.819352395379663e-05,
+      "loss": 0.1297,
+      "step": 4770
+    },
+    {
+      "epoch": 0.45256580193145235,
+      "grad_norm": 0.31184515357017517,
+      "learning_rate": 1.8189736792274193e-05,
+      "loss": 0.113,
+      "step": 4780
+    },
+    {
+      "epoch": 0.4535125923120621,
+      "grad_norm": 0.2756892144680023,
+      "learning_rate": 1.8185949630751753e-05,
+      "loss": 0.1189,
+      "step": 4790
+    },
+    {
+      "epoch": 0.45445938269267183,
+      "grad_norm": 0.38942059874534607,
+      "learning_rate": 1.8182162469229313e-05,
+      "loss": 0.1213,
+      "step": 4800
+    },
+    {
+      "epoch": 0.4554061730732816,
+      "grad_norm": 0.38971012830734253,
+      "learning_rate": 1.8178375307706873e-05,
+      "loss": 0.1186,
+      "step": 4810
+    },
+    {
+      "epoch": 0.4563529634538913,
+      "grad_norm": 0.36139774322509766,
+      "learning_rate": 1.8174588146184437e-05,
+      "loss": 0.1099,
+      "step": 4820
+    },
+    {
+      "epoch": 0.45729975383450105,
+      "grad_norm": 0.30894285440444946,
+      "learning_rate": 1.8170800984661997e-05,
+      "loss": 0.1125,
+      "step": 4830
+    },
+    {
+      "epoch": 0.4582465442151108,
+      "grad_norm": 0.39547091722488403,
+      "learning_rate": 1.816701382313956e-05,
+      "loss": 0.1159,
+      "step": 4840
+    },
+    {
+      "epoch": 0.4591933345957205,
+      "grad_norm": 0.35442787408828735,
+      "learning_rate": 1.816322666161712e-05,
+      "loss": 0.1199,
+      "step": 4850
+    },
+    {
+      "epoch": 0.4601401249763302,
+      "grad_norm": 0.31647929549217224,
+      "learning_rate": 1.815943950009468e-05,
+      "loss": 0.123,
+      "step": 4860
+    },
+    {
+      "epoch": 0.46108691535693996,
+      "grad_norm": 0.23882287740707397,
+      "learning_rate": 1.815565233857224e-05,
+      "loss": 0.1281,
+      "step": 4870
+    },
+    {
+      "epoch": 0.4620337057375497,
+      "grad_norm": 0.28302937746047974,
+      "learning_rate": 1.8151865177049804e-05,
+      "loss": 0.1178,
+      "step": 4880
+    },
+    {
+      "epoch": 0.46298049611815945,
+      "grad_norm": 0.30785325169563293,
+      "learning_rate": 1.8148078015527364e-05,
+      "loss": 0.115,
+      "step": 4890
+    },
+    {
+      "epoch": 0.4639272864987692,
+      "grad_norm": 0.291075199842453,
+      "learning_rate": 1.8144290854004924e-05,
+      "loss": 0.1119,
+      "step": 4900
+    },
+    {
+      "epoch": 0.46487407687937893,
+      "grad_norm": 0.3640170395374298,
+      "learning_rate": 1.8140503692482484e-05,
+      "loss": 0.1153,
+      "step": 4910
+    },
+    {
+      "epoch": 0.4658208672599886,
+      "grad_norm": 0.25291305780410767,
+      "learning_rate": 1.8136716530960047e-05,
+      "loss": 0.115,
+      "step": 4920
+    },
+    {
+      "epoch": 0.46676765764059835,
+      "grad_norm": 0.34131917357444763,
+      "learning_rate": 1.8132929369437607e-05,
+      "loss": 0.1164,
+      "step": 4930
+    },
+    {
+      "epoch": 0.4677144480212081,
+      "grad_norm": 0.40944379568099976,
+      "learning_rate": 1.8129142207915168e-05,
+      "loss": 0.1251,
+      "step": 4940
+    },
+    {
+      "epoch": 0.46866123840181784,
+      "grad_norm": 0.36214330792427063,
+      "learning_rate": 1.812535504639273e-05,
+      "loss": 0.1176,
+      "step": 4950
+    },
+    {
+      "epoch": 0.4696080287824276,
+      "grad_norm": 0.2422647923231125,
+      "learning_rate": 1.812156788487029e-05,
+      "loss": 0.1047,
+      "step": 4960
+    },
+    {
+      "epoch": 0.4705548191630373,
+      "grad_norm": 0.33104845881462097,
+      "learning_rate": 1.8117780723347854e-05,
+      "loss": 0.1178,
+      "step": 4970
+    },
+    {
+      "epoch": 0.47150160954364706,
+      "grad_norm": 0.30480530858039856,
+      "learning_rate": 1.8113993561825415e-05,
+      "loss": 0.1267,
+      "step": 4980
+    },
+    {
+      "epoch": 0.47244839992425675,
+      "grad_norm": 0.29709336161613464,
+      "learning_rate": 1.8110206400302975e-05,
+      "loss": 0.1172,
+      "step": 4990
+    },
+    {
+      "epoch": 0.4733951903048665,
+      "grad_norm": 0.3976782560348511,
+      "learning_rate": 1.8106419238780535e-05,
+      "loss": 0.1195,
+      "step": 5000
+    },
+    {
+      "epoch": 0.4743419806854762,
+      "grad_norm": 0.29639732837677,
+      "learning_rate": 1.8102632077258095e-05,
+      "loss": 0.12,
+      "step": 5010
+    },
+    {
+      "epoch": 0.47528877106608597,
+      "grad_norm": 0.31762635707855225,
+      "learning_rate": 1.8098844915735658e-05,
+      "loss": 0.1197,
+      "step": 5020
+    },
+    {
+      "epoch": 0.4762355614466957,
+      "grad_norm": 0.29064318537712097,
+      "learning_rate": 1.8095057754213218e-05,
+      "loss": 0.11,
+      "step": 5030
+    },
+    {
+      "epoch": 0.47718235182730545,
+      "grad_norm": 0.37592342495918274,
+      "learning_rate": 1.8091270592690778e-05,
+      "loss": 0.1122,
+      "step": 5040
+    },
+    {
+      "epoch": 0.4781291422079152,
+      "grad_norm": 0.3141964077949524,
+      "learning_rate": 1.808748343116834e-05,
+      "loss": 0.1202,
+      "step": 5050
+    },
+    {
+      "epoch": 0.4790759325885249,
+      "grad_norm": 0.4769488275051117,
+      "learning_rate": 1.80836962696459e-05,
+      "loss": 0.1191,
+      "step": 5060
+    },
+    {
+      "epoch": 0.4800227229691346,
+      "grad_norm": 0.30844876170158386,
+      "learning_rate": 1.8079909108123465e-05,
+      "loss": 0.113,
+      "step": 5070
+    },
+    {
+      "epoch": 0.48096951334974436,
+      "grad_norm": 0.3101666569709778,
+      "learning_rate": 1.8076121946601025e-05,
+      "loss": 0.1099,
+      "step": 5080
+    },
+    {
+      "epoch": 0.4819163037303541,
+      "grad_norm": 0.2982792854309082,
+      "learning_rate": 1.8072334785078585e-05,
+      "loss": 0.1315,
+      "step": 5090
+    },
+    {
+      "epoch": 0.48286309411096384,
+      "grad_norm": 0.33043378591537476,
+      "learning_rate": 1.8068547623556145e-05,
+      "loss": 0.1187,
+      "step": 5100
+    },
+    {
+      "epoch": 0.4838098844915736,
+      "grad_norm": 0.30517569184303284,
+      "learning_rate": 1.806476046203371e-05,
+      "loss": 0.108,
+      "step": 5110
+    },
+    {
+      "epoch": 0.4847566748721833,
+      "grad_norm": 0.393460750579834,
+      "learning_rate": 1.806097330051127e-05,
+      "loss": 0.1212,
+      "step": 5120
+    },
+    {
+      "epoch": 0.485703465252793,
+      "grad_norm": 0.2593437731266022,
+      "learning_rate": 1.805718613898883e-05,
+      "loss": 0.1095,
+      "step": 5130
+    },
+    {
+      "epoch": 0.48665025563340275,
+      "grad_norm": 0.33833277225494385,
+      "learning_rate": 1.805339897746639e-05,
+      "loss": 0.1239,
+      "step": 5140
+    },
+    {
+      "epoch": 0.4875970460140125,
+      "grad_norm": 0.2983340620994568,
+      "learning_rate": 1.804961181594395e-05,
+      "loss": 0.1141,
+      "step": 5150
+    },
+    {
+      "epoch": 0.48854383639462223,
+      "grad_norm": 0.286362886428833,
+      "learning_rate": 1.8045824654421512e-05,
+      "loss": 0.1089,
+      "step": 5160
+    },
+    {
+      "epoch": 0.489490626775232,
+      "grad_norm": 0.27127572894096375,
+      "learning_rate": 1.8042037492899072e-05,
+      "loss": 0.1031,
+      "step": 5170
+    },
+    {
+      "epoch": 0.4904374171558417,
+      "grad_norm": 0.28876861929893494,
+      "learning_rate": 1.8038250331376636e-05,
+      "loss": 0.1184,
+      "step": 5180
+    },
+    {
+      "epoch": 0.49138420753645146,
+      "grad_norm": 0.33738434314727783,
+      "learning_rate": 1.8034463169854196e-05,
+      "loss": 0.1157,
+      "step": 5190
+    },
+    {
+      "epoch": 0.49233099791706114,
+      "grad_norm": 0.40151458978652954,
+      "learning_rate": 1.803067600833176e-05,
+      "loss": 0.1124,
+      "step": 5200
+    },
+    {
+      "epoch": 0.4932777882976709,
+      "grad_norm": 0.30535009503364563,
+      "learning_rate": 1.802688884680932e-05,
+      "loss": 0.1254,
+      "step": 5210
+    },
+    {
+      "epoch": 0.4942245786782806,
+      "grad_norm": 0.29737281799316406,
+      "learning_rate": 1.802310168528688e-05,
+      "loss": 0.1118,
+      "step": 5220
+    },
+    {
+      "epoch": 0.49517136905889036,
+      "grad_norm": 0.27557647228240967,
+      "learning_rate": 1.801931452376444e-05,
+      "loss": 0.1042,
+      "step": 5230
+    },
+    {
+      "epoch": 0.4961181594395001,
+      "grad_norm": 0.4273216724395752,
+      "learning_rate": 1.8015527362242e-05,
+      "loss": 0.1263,
+      "step": 5240
+    },
+    {
+      "epoch": 0.49706494982010985,
+      "grad_norm": 0.3010781407356262,
+      "learning_rate": 1.8011740200719563e-05,
+      "loss": 0.1181,
+      "step": 5250
+    },
+    {
+      "epoch": 0.4980117402007196,
+      "grad_norm": 0.3851463496685028,
+      "learning_rate": 1.8007953039197123e-05,
+      "loss": 0.1186,
+      "step": 5260
+    },
+    {
+      "epoch": 0.4989585305813293,
+      "grad_norm": 0.2870292365550995,
+      "learning_rate": 1.8004165877674683e-05,
+      "loss": 0.1272,
+      "step": 5270
+    },
+    {
+      "epoch": 0.499905320961939,
+      "grad_norm": 0.3452846109867096,
+      "learning_rate": 1.8000378716152243e-05,
+      "loss": 0.1264,
+      "step": 5280
+    },
+    {
+      "epoch": 0.5008521113425488,
+      "grad_norm": 0.32496556639671326,
+      "learning_rate": 1.7996591554629807e-05,
+      "loss": 0.1183,
+      "step": 5290
+    },
+    {
+      "epoch": 0.5017989017231584,
+      "grad_norm": 0.29744213819503784,
+      "learning_rate": 1.7992804393107367e-05,
+      "loss": 0.1225,
+      "step": 5300
+    },
+    {
+      "epoch": 0.5027456921037682,
+      "grad_norm": 0.3163803815841675,
+      "learning_rate": 1.798901723158493e-05,
+      "loss": 0.1197,
+      "step": 5310
+    },
+    {
+      "epoch": 0.5036924824843779,
+      "grad_norm": 0.2961333096027374,
+      "learning_rate": 1.798523007006249e-05,
+      "loss": 0.1266,
+      "step": 5320
+    },
+    {
+      "epoch": 0.5046392728649877,
+      "grad_norm": 0.41541510820388794,
+      "learning_rate": 1.798144290854005e-05,
+      "loss": 0.1249,
+      "step": 5330
+    },
+    {
+      "epoch": 0.5055860632455974,
+      "grad_norm": 0.3027608096599579,
+      "learning_rate": 1.7977655747017614e-05,
+      "loss": 0.1138,
+      "step": 5340
+    },
+    {
+      "epoch": 0.5065328536262071,
+      "grad_norm": 0.34629517793655396,
+      "learning_rate": 1.7973868585495174e-05,
+      "loss": 0.1143,
+      "step": 5350
+    },
+    {
+      "epoch": 0.5074796440068169,
+      "grad_norm": 0.3278519809246063,
+      "learning_rate": 1.7970081423972734e-05,
+      "loss": 0.1258,
+      "step": 5360
+    },
+    {
+      "epoch": 0.5084264343874266,
+      "grad_norm": 0.39441338181495667,
+      "learning_rate": 1.7966294262450294e-05,
+      "loss": 0.1111,
+      "step": 5370
+    },
+    {
+      "epoch": 0.5093732247680364,
+      "grad_norm": 0.19750452041625977,
+      "learning_rate": 1.7962507100927854e-05,
+      "loss": 0.1132,
+      "step": 5380
+    },
+    {
+      "epoch": 0.5103200151486461,
+      "grad_norm": 0.32753902673721313,
+      "learning_rate": 1.7958719939405417e-05,
+      "loss": 0.1089,
+      "step": 5390
+    },
+    {
+      "epoch": 0.5112668055292559,
+      "grad_norm": 0.283546507358551,
+      "learning_rate": 1.7954932777882977e-05,
+      "loss": 0.1187,
+      "step": 5400
+    },
+    {
+      "epoch": 0.5122135959098656,
+      "grad_norm": 0.2674579322338104,
+      "learning_rate": 1.795114561636054e-05,
+      "loss": 0.1144,
+      "step": 5410
+    },
+    {
+      "epoch": 0.5131603862904753,
+      "grad_norm": 0.3309236168861389,
+      "learning_rate": 1.79473584548381e-05,
+      "loss": 0.1087,
+      "step": 5420
+    },
+    {
+      "epoch": 0.5141071766710851,
+      "grad_norm": 0.3469752371311188,
+      "learning_rate": 1.794357129331566e-05,
+      "loss": 0.1207,
+      "step": 5430
+    },
+    {
+      "epoch": 0.5150539670516947,
+      "grad_norm": 0.28428488969802856,
+      "learning_rate": 1.7939784131793224e-05,
+      "loss": 0.1221,
+      "step": 5440
+    },
+    {
+      "epoch": 0.5160007574323044,
+      "grad_norm": 0.32135307788848877,
+      "learning_rate": 1.7935996970270784e-05,
+      "loss": 0.1193,
+      "step": 5450
+    },
+    {
+      "epoch": 0.5169475478129142,
+      "grad_norm": 0.26351454854011536,
+      "learning_rate": 1.7932209808748344e-05,
+      "loss": 0.1218,
+      "step": 5460
+    },
+    {
+      "epoch": 0.5178943381935239,
+      "grad_norm": 0.2696927785873413,
+      "learning_rate": 1.7928422647225904e-05,
+      "loss": 0.1108,
+      "step": 5470
+    },
+    {
+      "epoch": 0.5188411285741337,
+      "grad_norm": 0.4349839985370636,
+      "learning_rate": 1.7924635485703468e-05,
+      "loss": 0.125,
+      "step": 5480
+    },
+    {
+      "epoch": 0.5197879189547434,
+      "grad_norm": 0.3225243091583252,
+      "learning_rate": 1.7920848324181028e-05,
+      "loss": 0.1191,
+      "step": 5490
+    },
+    {
+      "epoch": 0.5207347093353532,
+      "grad_norm": 0.3667551577091217,
+      "learning_rate": 1.7917061162658588e-05,
+      "loss": 0.112,
+      "step": 5500
+    },
+    {
+      "epoch": 0.5216814997159629,
+      "grad_norm": 0.3296608328819275,
+      "learning_rate": 1.7913274001136148e-05,
+      "loss": 0.1181,
+      "step": 5510
+    },
+    {
+      "epoch": 0.5226282900965726,
+      "grad_norm": 0.3578593134880066,
+      "learning_rate": 1.790948683961371e-05,
+      "loss": 0.1193,
+      "step": 5520
+    },
+    {
+      "epoch": 0.5235750804771824,
+      "grad_norm": 0.3956611752510071,
+      "learning_rate": 1.790569967809127e-05,
+      "loss": 0.1203,
+      "step": 5530
+    },
+    {
+      "epoch": 0.5245218708577921,
+      "grad_norm": 0.34705013036727905,
+      "learning_rate": 1.7901912516568835e-05,
+      "loss": 0.118,
+      "step": 5540
+    },
+    {
+      "epoch": 0.5254686612384019,
+      "grad_norm": 0.30799970030784607,
+      "learning_rate": 1.7898125355046395e-05,
+      "loss": 0.117,
+      "step": 5550
+    },
+    {
+      "epoch": 0.5264154516190116,
+      "grad_norm": 0.2643110156059265,
+      "learning_rate": 1.7894338193523955e-05,
+      "loss": 0.1139,
+      "step": 5560
+    },
+    {
+      "epoch": 0.5273622419996212,
+      "grad_norm": 0.2462998777627945,
+      "learning_rate": 1.789055103200152e-05,
+      "loss": 0.1148,
+      "step": 5570
+    },
+    {
+      "epoch": 0.528309032380231,
+      "grad_norm": 0.3135557472705841,
+      "learning_rate": 1.788676387047908e-05,
+      "loss": 0.1123,
+      "step": 5580
+    },
+    {
+      "epoch": 0.5292558227608407,
+      "grad_norm": 0.3650512397289276,
+      "learning_rate": 1.788297670895664e-05,
+      "loss": 0.1127,
+      "step": 5590
+    },
+    {
+      "epoch": 0.5302026131414505,
+      "grad_norm": 0.24632461369037628,
+      "learning_rate": 1.78791895474342e-05,
+      "loss": 0.1248,
+      "step": 5600
+    },
+    {
+      "epoch": 0.5311494035220602,
+      "grad_norm": 0.3508126139640808,
+      "learning_rate": 1.787540238591176e-05,
+      "loss": 0.1208,
+      "step": 5610
+    },
+    {
+      "epoch": 0.5320961939026699,
+      "grad_norm": 0.33299440145492554,
+      "learning_rate": 1.7871615224389322e-05,
+      "loss": 0.1119,
+      "step": 5620
+    },
+    {
+      "epoch": 0.5330429842832797,
+      "grad_norm": 0.3384663164615631,
+      "learning_rate": 1.7867828062866882e-05,
+      "loss": 0.1196,
+      "step": 5630
+    },
+    {
+      "epoch": 0.5339897746638894,
+      "grad_norm": 0.389332115650177,
+      "learning_rate": 1.7864040901344442e-05,
+      "loss": 0.1137,
+      "step": 5640
+    },
+    {
+      "epoch": 0.5349365650444992,
+      "grad_norm": 0.3215329349040985,
+      "learning_rate": 1.7860253739822006e-05,
+      "loss": 0.1155,
+      "step": 5650
+    },
+    {
+      "epoch": 0.5358833554251089,
+      "grad_norm": 0.32901594042778015,
+      "learning_rate": 1.7856466578299566e-05,
+      "loss": 0.1133,
+      "step": 5660
+    },
+    {
+      "epoch": 0.5368301458057186,
+      "grad_norm": 0.364347904920578,
+      "learning_rate": 1.785267941677713e-05,
+      "loss": 0.1272,
+      "step": 5670
+    },
+    {
+      "epoch": 0.5377769361863284,
+      "grad_norm": 0.29443469643592834,
+      "learning_rate": 1.784889225525469e-05,
+      "loss": 0.1136,
+      "step": 5680
+    },
+    {
+      "epoch": 0.5387237265669381,
+      "grad_norm": 0.3348941206932068,
+      "learning_rate": 1.784510509373225e-05,
+      "loss": 0.1164,
+      "step": 5690
+    },
+    {
+      "epoch": 0.5396705169475479,
+      "grad_norm": 0.26882919669151306,
+      "learning_rate": 1.784131793220981e-05,
+      "loss": 0.1211,
+      "step": 5700
+    },
+    {
+      "epoch": 0.5406173073281575,
+      "grad_norm": 0.22887469828128815,
+      "learning_rate": 1.7837530770687373e-05,
+      "loss": 0.1103,
+      "step": 5710
+    },
+    {
+      "epoch": 0.5415640977087672,
+      "grad_norm": 0.40600207448005676,
+      "learning_rate": 1.7833743609164933e-05,
+      "loss": 0.113,
+      "step": 5720
+    },
+    {
+      "epoch": 0.542510888089377,
+      "grad_norm": 0.3073524534702301,
+      "learning_rate": 1.7829956447642493e-05,
+      "loss": 0.104,
+      "step": 5730
+    },
+    {
+      "epoch": 0.5434576784699867,
+      "grad_norm": 0.3362687826156616,
+      "learning_rate": 1.7826169286120053e-05,
+      "loss": 0.1145,
+      "step": 5740
+    },
+    {
+      "epoch": 0.5444044688505965,
+      "grad_norm": 0.30210235714912415,
+      "learning_rate": 1.7822382124597616e-05,
+      "loss": 0.1132,
+      "step": 5750
+    },
+    {
+      "epoch": 0.5453512592312062,
+      "grad_norm": 0.4413982927799225,
+      "learning_rate": 1.7818594963075176e-05,
+      "loss": 0.1256,
+      "step": 5760
+    },
+    {
+      "epoch": 0.5462980496118159,
+      "grad_norm": 0.32046809792518616,
+      "learning_rate": 1.781480780155274e-05,
+      "loss": 0.1178,
+      "step": 5770
+    },
+    {
+      "epoch": 0.5472448399924257,
+      "grad_norm": 0.36214178800582886,
+      "learning_rate": 1.78110206400303e-05,
+      "loss": 0.1179,
+      "step": 5780
+    },
+    {
+      "epoch": 0.5481916303730354,
+      "grad_norm": 0.2953844666481018,
+      "learning_rate": 1.780723347850786e-05,
+      "loss": 0.1229,
+      "step": 5790
+    },
+    {
+      "epoch": 0.5491384207536452,
+      "grad_norm": 0.36594676971435547,
+      "learning_rate": 1.780344631698542e-05,
+      "loss": 0.1182,
+      "step": 5800
+    },
+    {
+      "epoch": 0.5500852111342549,
+      "grad_norm": 0.38411974906921387,
+      "learning_rate": 1.7799659155462983e-05,
+      "loss": 0.1322,
+      "step": 5810
+    },
+    {
+      "epoch": 0.5510320015148646,
+      "grad_norm": 0.3273877799510956,
+      "learning_rate": 1.7795871993940543e-05,
+      "loss": 0.1226,
+      "step": 5820
+    },
+    {
+      "epoch": 0.5519787918954744,
+      "grad_norm": 0.3491196930408478,
+      "learning_rate": 1.7792084832418103e-05,
+      "loss": 0.1172,
+      "step": 5830
+    },
+    {
+      "epoch": 0.5529255822760841,
+      "grad_norm": 0.27573728561401367,
+      "learning_rate": 1.7788297670895663e-05,
+      "loss": 0.117,
+      "step": 5840
+    },
+    {
+      "epoch": 0.5538723726566938,
+      "grad_norm": 0.31645190715789795,
+      "learning_rate": 1.7784510509373227e-05,
+      "loss": 0.1122,
+      "step": 5850
+    },
+    {
+      "epoch": 0.5548191630373035,
+      "grad_norm": 0.3352598249912262,
+      "learning_rate": 1.7780723347850787e-05,
+      "loss": 0.1193,
+      "step": 5860
+    },
+    {
+      "epoch": 0.5557659534179132,
+      "grad_norm": 0.49130764603614807,
+      "learning_rate": 1.7776936186328347e-05,
+      "loss": 0.1306,
+      "step": 5870
+    },
+    {
+      "epoch": 0.556712743798523,
+      "grad_norm": 0.27824321389198303,
+      "learning_rate": 1.777314902480591e-05,
+      "loss": 0.1095,
+      "step": 5880
+    },
+    {
+      "epoch": 0.5576595341791327,
+      "grad_norm": 0.3100660443305969,
+      "learning_rate": 1.776936186328347e-05,
+      "loss": 0.1367,
+      "step": 5890
+    },
+    {
+      "epoch": 0.5586063245597425,
+      "grad_norm": 0.404834508895874,
+      "learning_rate": 1.7765574701761034e-05,
+      "loss": 0.1166,
+      "step": 5900
+    },
+    {
+      "epoch": 0.5595531149403522,
+      "grad_norm": 0.2933177053928375,
+      "learning_rate": 1.7761787540238594e-05,
+      "loss": 0.1202,
+      "step": 5910
+    },
+    {
+      "epoch": 0.5604999053209619,
+      "grad_norm": 0.3327297270298004,
+      "learning_rate": 1.7758000378716154e-05,
+      "loss": 0.1185,
+      "step": 5920
+    },
+    {
+      "epoch": 0.5614466957015717,
+      "grad_norm": 0.2998553514480591,
+      "learning_rate": 1.7754213217193714e-05,
+      "loss": 0.1163,
+      "step": 5930
+    },
+    {
+      "epoch": 0.5623934860821814,
+      "grad_norm": 0.3321799337863922,
+      "learning_rate": 1.7750426055671274e-05,
+      "loss": 0.1192,
+      "step": 5940
+    },
+    {
+      "epoch": 0.5633402764627912,
+      "grad_norm": 0.2803283929824829,
+      "learning_rate": 1.7746638894148838e-05,
+      "loss": 0.1112,
+      "step": 5950
+    },
+    {
+      "epoch": 0.5642870668434009,
+      "grad_norm": 0.28032103180885315,
+      "learning_rate": 1.7742851732626398e-05,
+      "loss": 0.1198,
+      "step": 5960
+    },
+    {
+      "epoch": 0.5652338572240106,
+      "grad_norm": 0.2938515543937683,
+      "learning_rate": 1.7739064571103958e-05,
+      "loss": 0.1118,
+      "step": 5970
+    },
+    {
+      "epoch": 0.5661806476046204,
+      "grad_norm": 0.33386483788490295,
+      "learning_rate": 1.773527740958152e-05,
+      "loss": 0.1095,
+      "step": 5980
+    },
+    {
+      "epoch": 0.56712743798523,
+      "grad_norm": 0.35478469729423523,
+      "learning_rate": 1.773149024805908e-05,
+      "loss": 0.1235,
+      "step": 5990
+    },
+    {
+      "epoch": 0.5680742283658398,
+      "grad_norm": 0.29097840189933777,
+      "learning_rate": 1.772770308653664e-05,
+      "loss": 0.1247,
+      "step": 6000
+    },
+    {
+      "epoch": 0.5690210187464495,
+      "grad_norm": 0.36649322509765625,
+      "learning_rate": 1.7723915925014205e-05,
+      "loss": 0.1206,
+      "step": 6010
+    },
+    {
+      "epoch": 0.5699678091270592,
+      "grad_norm": 0.3363392651081085,
+      "learning_rate": 1.7720128763491765e-05,
+      "loss": 0.1144,
+      "step": 6020
+    },
+    {
+      "epoch": 0.570914599507669,
+      "grad_norm": 0.31241416931152344,
+      "learning_rate": 1.7716341601969325e-05,
+      "loss": 0.1145,
+      "step": 6030
+    },
+    {
+      "epoch": 0.5718613898882787,
+      "grad_norm": 0.34137117862701416,
+      "learning_rate": 1.7712554440446888e-05,
+      "loss": 0.1245,
+      "step": 6040
+    },
+    {
+      "epoch": 0.5728081802688885,
+      "grad_norm": 0.4431675374507904,
+      "learning_rate": 1.7708767278924448e-05,
+      "loss": 0.1218,
+      "step": 6050
+    },
+    {
+      "epoch": 0.5737549706494982,
+      "grad_norm": 0.28357037901878357,
+      "learning_rate": 1.7704980117402008e-05,
+      "loss": 0.1097,
+      "step": 6060
+    },
+    {
+      "epoch": 0.574701761030108,
+      "grad_norm": 0.35272452235221863,
+      "learning_rate": 1.770119295587957e-05,
+      "loss": 0.1246,
+      "step": 6070
+    },
+    {
+      "epoch": 0.5756485514107177,
+      "grad_norm": 0.29270246624946594,
+      "learning_rate": 1.769740579435713e-05,
+      "loss": 0.1067,
+      "step": 6080
+    },
+    {
+      "epoch": 0.5765953417913274,
+      "grad_norm": 0.3070552349090576,
+      "learning_rate": 1.7693618632834692e-05,
+      "loss": 0.1131,
+      "step": 6090
+    },
+    {
+      "epoch": 0.5775421321719372,
+      "grad_norm": 0.42108526825904846,
+      "learning_rate": 1.7689831471312252e-05,
+      "loss": 0.1244,
+      "step": 6100
+    },
+    {
+      "epoch": 0.5784889225525469,
+      "grad_norm": 0.3098716735839844,
+      "learning_rate": 1.7686044309789815e-05,
+      "loss": 0.1226,
+      "step": 6110
+    },
+    {
+      "epoch": 0.5794357129331565,
+      "grad_norm": 0.2832401990890503,
+      "learning_rate": 1.7682257148267375e-05,
+      "loss": 0.1135,
+      "step": 6120
+    },
+    {
+      "epoch": 0.5803825033137663,
+      "grad_norm": 0.3348044455051422,
+      "learning_rate": 1.767846998674494e-05,
+      "loss": 0.1198,
+      "step": 6130
+    },
+    {
+      "epoch": 0.581329293694376,
+      "grad_norm": 0.3181207776069641,
+      "learning_rate": 1.76746828252225e-05,
+      "loss": 0.1148,
+      "step": 6140
+    },
+    {
+      "epoch": 0.5822760840749858,
+      "grad_norm": 0.46395212411880493,
+      "learning_rate": 1.767089566370006e-05,
+      "loss": 0.1257,
+      "step": 6150
+    },
+    {
+      "epoch": 0.5832228744555955,
+      "grad_norm": 0.3275673985481262,
+      "learning_rate": 1.766710850217762e-05,
+      "loss": 0.1176,
+      "step": 6160
+    },
+    {
+      "epoch": 0.5841696648362052,
+      "grad_norm": 0.28788167238235474,
+      "learning_rate": 1.766332134065518e-05,
+      "loss": 0.1076,
+      "step": 6170
+    },
+    {
+      "epoch": 0.585116455216815,
+      "grad_norm": 0.29496243596076965,
+      "learning_rate": 1.7659534179132742e-05,
+      "loss": 0.1083,
+      "step": 6180
+    },
+    {
+      "epoch": 0.5860632455974247,
+      "grad_norm": 0.2746005356311798,
+      "learning_rate": 1.7655747017610302e-05,
+      "loss": 0.1139,
+      "step": 6190
+    },
+    {
+      "epoch": 0.5870100359780345,
+      "grad_norm": 0.3565037250518799,
+      "learning_rate": 1.7651959856087863e-05,
+      "loss": 0.1207,
+      "step": 6200
+    },
+    {
+      "epoch": 0.5879568263586442,
+      "grad_norm": 0.28780317306518555,
+      "learning_rate": 1.7648172694565423e-05,
+      "loss": 0.1172,
+      "step": 6210
+    },
+    {
+      "epoch": 0.588903616739254,
+      "grad_norm": 0.3613559305667877,
+      "learning_rate": 1.7644385533042986e-05,
+      "loss": 0.1152,
+      "step": 6220
+    },
+    {
+      "epoch": 0.5898504071198637,
+      "grad_norm": 0.3410801589488983,
+      "learning_rate": 1.7640598371520546e-05,
+      "loss": 0.1113,
+      "step": 6230
+    },
+    {
+      "epoch": 0.5907971975004734,
+      "grad_norm": 0.26274338364601135,
+      "learning_rate": 1.763681120999811e-05,
+      "loss": 0.1147,
+      "step": 6240
+    },
+    {
+      "epoch": 0.5917439878810832,
+      "grad_norm": 0.3763429820537567,
+      "learning_rate": 1.763302404847567e-05,
+      "loss": 0.1142,
+      "step": 6250
+    },
+    {
+      "epoch": 0.5926907782616928,
+      "grad_norm": 0.4751342833042145,
+      "learning_rate": 1.762923688695323e-05,
+      "loss": 0.1187,
+      "step": 6260
+    },
+    {
+      "epoch": 0.5936375686423025,
+      "grad_norm": 0.3844728171825409,
+      "learning_rate": 1.7625449725430793e-05,
+      "loss": 0.1208,
+      "step": 6270
+    },
+    {
+      "epoch": 0.5945843590229123,
+      "grad_norm": 0.32040935754776,
+      "learning_rate": 1.7621662563908353e-05,
+      "loss": 0.1158,
+      "step": 6280
+    },
+    {
+      "epoch": 0.595531149403522,
+      "grad_norm": 0.34830090403556824,
+      "learning_rate": 1.7617875402385913e-05,
+      "loss": 0.114,
+      "step": 6290
+    },
+    {
+      "epoch": 0.5964779397841318,
+      "grad_norm": 0.3176250159740448,
+      "learning_rate": 1.7614088240863473e-05,
+      "loss": 0.1155,
+      "step": 6300
+    },
+    {
+      "epoch": 0.5974247301647415,
+      "grad_norm": 0.3590896427631378,
+      "learning_rate": 1.7610301079341033e-05,
+      "loss": 0.1153,
+      "step": 6310
+    },
+    {
+      "epoch": 0.5983715205453513,
+      "grad_norm": 0.2998102903366089,
+      "learning_rate": 1.7606513917818597e-05,
+      "loss": 0.1179,
+      "step": 6320
+    },
+    {
+      "epoch": 0.599318310925961,
+      "grad_norm": 0.31011080741882324,
+      "learning_rate": 1.7602726756296157e-05,
+      "loss": 0.1295,
+      "step": 6330
+    },
+    {
+      "epoch": 0.6002651013065707,
+      "grad_norm": 0.30341920256614685,
+      "learning_rate": 1.759893959477372e-05,
+      "loss": 0.1214,
+      "step": 6340
+    },
+    {
+      "epoch": 0.6012118916871805,
+      "grad_norm": 0.372761607170105,
+      "learning_rate": 1.759515243325128e-05,
+      "loss": 0.1208,
+      "step": 6350
+    },
+    {
+      "epoch": 0.6021586820677902,
+      "grad_norm": 0.33373719453811646,
+      "learning_rate": 1.759136527172884e-05,
+      "loss": 0.1034,
+      "step": 6360
+    },
+    {
+      "epoch": 0.6031054724484,
+      "grad_norm": 0.34052538871765137,
+      "learning_rate": 1.7587578110206404e-05,
+      "loss": 0.1103,
+      "step": 6370
+    },
+    {
+      "epoch": 0.6040522628290097,
+      "grad_norm": 0.28972190618515015,
+      "learning_rate": 1.7583790948683964e-05,
+      "loss": 0.121,
+      "step": 6380
+    },
+    {
+      "epoch": 0.6049990532096194,
+      "grad_norm": 0.29406800866127014,
+      "learning_rate": 1.7580003787161524e-05,
+      "loss": 0.1238,
+      "step": 6390
+    },
+    {
+      "epoch": 0.6059458435902291,
+      "grad_norm": 0.4161909520626068,
+      "learning_rate": 1.7576216625639084e-05,
+      "loss": 0.1135,
+      "step": 6400
+    },
+    {
+      "epoch": 0.6068926339708388,
+      "grad_norm": 0.3617647886276245,
+      "learning_rate": 1.7572429464116647e-05,
+      "loss": 0.1175,
+      "step": 6410
+    },
+    {
+      "epoch": 0.6078394243514486,
+      "grad_norm": 0.33809611201286316,
+      "learning_rate": 1.7568642302594207e-05,
+      "loss": 0.1043,
+      "step": 6420
+    },
+    {
+      "epoch": 0.6087862147320583,
+      "grad_norm": 0.3799417316913605,
+      "learning_rate": 1.7564855141071767e-05,
+      "loss": 0.1058,
+      "step": 6430
+    },
+    {
+      "epoch": 0.609733005112668,
+      "grad_norm": 0.3288790285587311,
+      "learning_rate": 1.7561067979549327e-05,
+      "loss": 0.1073,
+      "step": 6440
+    },
+    {
+      "epoch": 0.6106797954932778,
+      "grad_norm": 0.2925897240638733,
+      "learning_rate": 1.755728081802689e-05,
+      "loss": 0.1144,
+      "step": 6450
+    },
+    {
+      "epoch": 0.6116265858738875,
+      "grad_norm": 0.29820024967193604,
+      "learning_rate": 1.755349365650445e-05,
+      "loss": 0.1199,
+      "step": 6460
+    },
+    {
+      "epoch": 0.6125733762544973,
+      "grad_norm": 0.2849036753177643,
+      "learning_rate": 1.7549706494982014e-05,
+      "loss": 0.1192,
+      "step": 6470
+    },
+    {
+      "epoch": 0.613520166635107,
+      "grad_norm": 0.2699211537837982,
+      "learning_rate": 1.7545919333459574e-05,
+      "loss": 0.1087,
+      "step": 6480
+    },
+    {
+      "epoch": 0.6144669570157167,
+      "grad_norm": 0.31521305441856384,
+      "learning_rate": 1.7542132171937134e-05,
+      "loss": 0.1141,
+      "step": 6490
+    },
+    {
+      "epoch": 0.6154137473963265,
+      "grad_norm": 0.28813159465789795,
+      "learning_rate": 1.7538345010414694e-05,
+      "loss": 0.1147,
+      "step": 6500
+    },
+    {
+      "epoch": 0.6163605377769362,
+      "grad_norm": 0.4241039752960205,
+      "learning_rate": 1.7534557848892258e-05,
+      "loss": 0.114,
+      "step": 6510
+    },
+    {
+      "epoch": 0.617307328157546,
+      "grad_norm": 0.2915937602519989,
+      "learning_rate": 1.7530770687369818e-05,
+      "loss": 0.1121,
+      "step": 6520
+    },
+    {
+      "epoch": 0.6182541185381557,
+      "grad_norm": 0.311350554227829,
+      "learning_rate": 1.7526983525847378e-05,
+      "loss": 0.1207,
+      "step": 6530
+    },
+    {
+      "epoch": 0.6192009089187653,
+      "grad_norm": 0.38931339979171753,
+      "learning_rate": 1.7523196364324938e-05,
+      "loss": 0.1166,
+      "step": 6540
+    },
+    {
+      "epoch": 0.6201476992993751,
+      "grad_norm": 0.3369103968143463,
+      "learning_rate": 1.75194092028025e-05,
+      "loss": 0.1208,
+      "step": 6550
+    },
+    {
+      "epoch": 0.6210944896799848,
+      "grad_norm": 0.2888890504837036,
+      "learning_rate": 1.751562204128006e-05,
+      "loss": 0.1224,
+      "step": 6560
+    },
+    {
+      "epoch": 0.6220412800605946,
+      "grad_norm": 0.4478912949562073,
+      "learning_rate": 1.751183487975762e-05,
+      "loss": 0.1218,
+      "step": 6570
+    },
+    {
+      "epoch": 0.6229880704412043,
+      "grad_norm": 0.34411540627479553,
+      "learning_rate": 1.7508047718235185e-05,
+      "loss": 0.1205,
+      "step": 6580
+    },
+    {
+      "epoch": 0.623934860821814,
+      "grad_norm": 0.2995246648788452,
+      "learning_rate": 1.7504260556712745e-05,
+      "loss": 0.109,
+      "step": 6590
+    },
+    {
+      "epoch": 0.6248816512024238,
+      "grad_norm": 0.2678649425506592,
+      "learning_rate": 1.750047339519031e-05,
+      "loss": 0.1122,
+      "step": 6600
+    },
+    {
+      "epoch": 0.6258284415830335,
+      "grad_norm": 0.3918863832950592,
+      "learning_rate": 1.749668623366787e-05,
+      "loss": 0.1132,
+      "step": 6610
+    },
+    {
+      "epoch": 0.6267752319636433,
+      "grad_norm": 0.3421894311904907,
+      "learning_rate": 1.749289907214543e-05,
+      "loss": 0.101,
+      "step": 6620
+    },
+    {
+      "epoch": 0.627722022344253,
+      "grad_norm": 0.32719123363494873,
+      "learning_rate": 1.748911191062299e-05,
+      "loss": 0.1193,
+      "step": 6630
+    },
+    {
+      "epoch": 0.6286688127248627,
+      "grad_norm": 0.43975284695625305,
+      "learning_rate": 1.748532474910055e-05,
+      "loss": 0.1129,
+      "step": 6640
+    },
+    {
+      "epoch": 0.6296156031054725,
+      "grad_norm": 0.27118706703186035,
+      "learning_rate": 1.7481537587578112e-05,
+      "loss": 0.1099,
+      "step": 6650
+    },
+    {
+      "epoch": 0.6305623934860822,
+      "grad_norm": 0.32624804973602295,
+      "learning_rate": 1.7477750426055672e-05,
+      "loss": 0.1249,
+      "step": 6660
+    },
+    {
+      "epoch": 0.631509183866692,
+      "grad_norm": 0.3094474971294403,
+      "learning_rate": 1.7473963264533232e-05,
+      "loss": 0.122,
+      "step": 6670
+    },
+    {
+      "epoch": 0.6324559742473016,
+      "grad_norm": 0.31549081206321716,
+      "learning_rate": 1.7470176103010796e-05,
+      "loss": 0.1085,
+      "step": 6680
+    },
+    {
+      "epoch": 0.6334027646279113,
+      "grad_norm": 0.24869810044765472,
+      "learning_rate": 1.7466388941488356e-05,
+      "loss": 0.1182,
+      "step": 6690
+    },
+    {
+      "epoch": 0.6343495550085211,
+      "grad_norm": 0.30127742886543274,
+      "learning_rate": 1.746260177996592e-05,
+      "loss": 0.1004,
+      "step": 6700
+    },
+    {
+      "epoch": 0.6352963453891308,
+      "grad_norm": 0.47941163182258606,
+      "learning_rate": 1.745881461844348e-05,
+      "loss": 0.1176,
+      "step": 6710
+    },
+    {
+      "epoch": 0.6362431357697406,
+      "grad_norm": 0.35430312156677246,
+      "learning_rate": 1.745502745692104e-05,
+      "loss": 0.1153,
+      "step": 6720
+    },
+    {
+      "epoch": 0.6371899261503503,
+      "grad_norm": 0.2501283288002014,
+      "learning_rate": 1.74512402953986e-05,
+      "loss": 0.1101,
+      "step": 6730
+    },
+    {
+      "epoch": 0.63813671653096,
+      "grad_norm": 0.3357498049736023,
+      "learning_rate": 1.7447453133876163e-05,
+      "loss": 0.1191,
+      "step": 6740
+    },
+    {
+      "epoch": 0.6390835069115698,
+      "grad_norm": 0.30266913771629333,
+      "learning_rate": 1.7443665972353723e-05,
+      "loss": 0.1137,
+      "step": 6750
+    },
+    {
+      "epoch": 0.6400302972921795,
+      "grad_norm": 0.39451736211776733,
+      "learning_rate": 1.7439878810831283e-05,
+      "loss": 0.1113,
+      "step": 6760
+    },
+    {
+      "epoch": 0.6409770876727893,
+      "grad_norm": 0.33438143134117126,
+      "learning_rate": 1.7436091649308843e-05,
+      "loss": 0.118,
+      "step": 6770
+    },
+    {
+      "epoch": 0.641923878053399,
+      "grad_norm": 0.2927396893501282,
+      "learning_rate": 1.7432304487786403e-05,
+      "loss": 0.1172,
+      "step": 6780
+    },
+    {
+      "epoch": 0.6428706684340088,
+      "grad_norm": 0.3146035969257355,
+      "learning_rate": 1.7428517326263966e-05,
+      "loss": 0.1073,
+      "step": 6790
+    },
+    {
+      "epoch": 0.6438174588146185,
+      "grad_norm": 0.27291423082351685,
+      "learning_rate": 1.7424730164741526e-05,
+      "loss": 0.1177,
+      "step": 6800
+    },
+    {
+      "epoch": 0.6447642491952281,
+      "grad_norm": 0.3319913148880005,
+      "learning_rate": 1.742094300321909e-05,
+      "loss": 0.119,
+      "step": 6810
+    },
+    {
+      "epoch": 0.6457110395758379,
+      "grad_norm": 0.3484410345554352,
+      "learning_rate": 1.741715584169665e-05,
+      "loss": 0.1201,
+      "step": 6820
+    },
+    {
+      "epoch": 0.6466578299564476,
+      "grad_norm": 0.2857920229434967,
+      "learning_rate": 1.7413368680174213e-05,
+      "loss": 0.1043,
+      "step": 6830
+    },
+    {
+      "epoch": 0.6476046203370573,
+      "grad_norm": 0.3543633818626404,
+      "learning_rate": 1.7409581518651773e-05,
+      "loss": 0.1148,
+      "step": 6840
+    },
+    {
+      "epoch": 0.6485514107176671,
+      "grad_norm": 0.2742081582546234,
+      "learning_rate": 1.7405794357129333e-05,
+      "loss": 0.1037,
+      "step": 6850
+    },
+    {
+      "epoch": 0.6494982010982768,
+      "grad_norm": 0.36285659670829773,
+      "learning_rate": 1.7402007195606894e-05,
+      "loss": 0.1154,
+      "step": 6860
+    },
+    {
+      "epoch": 0.6504449914788866,
+      "grad_norm": 0.3244501054286957,
+      "learning_rate": 1.7398220034084454e-05,
+      "loss": 0.1115,
+      "step": 6870
+    },
+    {
+      "epoch": 0.6513917818594963,
+      "grad_norm": 0.3813318610191345,
+      "learning_rate": 1.7394432872562017e-05,
+      "loss": 0.1125,
+      "step": 6880
+    },
+    {
+      "epoch": 0.652338572240106,
+      "grad_norm": 0.34014782309532166,
+      "learning_rate": 1.7390645711039577e-05,
+      "loss": 0.1154,
+      "step": 6890
+    },
+    {
+      "epoch": 0.6532853626207158,
+      "grad_norm": 0.31174853444099426,
+      "learning_rate": 1.7386858549517137e-05,
+      "loss": 0.1156,
+      "step": 6900
+    },
+    {
+      "epoch": 0.6542321530013255,
+      "grad_norm": 0.3397250771522522,
+      "learning_rate": 1.7383071387994697e-05,
+      "loss": 0.1113,
+      "step": 6910
+    },
+    {
+      "epoch": 0.6551789433819353,
+      "grad_norm": 0.3041819632053375,
+      "learning_rate": 1.737928422647226e-05,
+      "loss": 0.1235,
+      "step": 6920
+    },
+    {
+      "epoch": 0.656125733762545,
+      "grad_norm": 0.31160038709640503,
+      "learning_rate": 1.737549706494982e-05,
+      "loss": 0.1117,
+      "step": 6930
+    },
+    {
+      "epoch": 0.6570725241431548,
+      "grad_norm": 0.3039339780807495,
+      "learning_rate": 1.7371709903427384e-05,
+      "loss": 0.1164,
+      "step": 6940
+    },
+    {
+      "epoch": 0.6580193145237644,
+      "grad_norm": 0.4804076850414276,
+      "learning_rate": 1.7367922741904944e-05,
+      "loss": 0.1148,
+      "step": 6950
+    },
+    {
+      "epoch": 0.6589661049043741,
+      "grad_norm": 0.31758782267570496,
+      "learning_rate": 1.7364135580382504e-05,
+      "loss": 0.1042,
+      "step": 6960
+    },
+    {
+      "epoch": 0.6599128952849839,
+      "grad_norm": 0.3608994781970978,
+      "learning_rate": 1.7360348418860068e-05,
+      "loss": 0.1173,
+      "step": 6970
+    },
+    {
+      "epoch": 0.6608596856655936,
+      "grad_norm": 0.32902655005455017,
+      "learning_rate": 1.7356561257337628e-05,
+      "loss": 0.1105,
+      "step": 6980
+    },
+    {
+      "epoch": 0.6618064760462034,
+      "grad_norm": 0.29006704688072205,
+      "learning_rate": 1.7352774095815188e-05,
+      "loss": 0.1069,
+      "step": 6990
+    },
+    {
+      "epoch": 0.6627532664268131,
+      "grad_norm": 0.3214852809906006,
+      "learning_rate": 1.7348986934292748e-05,
+      "loss": 0.121,
+      "step": 7000
+    },
+    {
+      "epoch": 0.6637000568074228,
+      "grad_norm": 0.2860215902328491,
+      "learning_rate": 1.7345199772770308e-05,
+      "loss": 0.1135,
+      "step": 7010
+    },
+    {
+      "epoch": 0.6646468471880326,
+      "grad_norm": 0.2932814359664917,
+      "learning_rate": 1.734141261124787e-05,
+      "loss": 0.1047,
+      "step": 7020
+    },
+    {
+      "epoch": 0.6655936375686423,
+      "grad_norm": 0.310357928276062,
+      "learning_rate": 1.733762544972543e-05,
+      "loss": 0.1119,
+      "step": 7030
+    },
+    {
+      "epoch": 0.6665404279492521,
+      "grad_norm": 0.38796934485435486,
+      "learning_rate": 1.7333838288202995e-05,
+      "loss": 0.1136,
+      "step": 7040
+    },
+    {
+      "epoch": 0.6674872183298618,
+      "grad_norm": 0.3642352819442749,
+      "learning_rate": 1.7330051126680555e-05,
+      "loss": 0.1071,
+      "step": 7050
+    },
+    {
+      "epoch": 0.6684340087104715,
+      "grad_norm": 0.40385109186172485,
+      "learning_rate": 1.7326263965158115e-05,
+      "loss": 0.1171,
+      "step": 7060
+    },
+    {
+      "epoch": 0.6693807990910813,
+      "grad_norm": 0.4555074870586395,
+      "learning_rate": 1.7322476803635678e-05,
+      "loss": 0.1205,
+      "step": 7070
+    },
+    {
+      "epoch": 0.670327589471691,
+      "grad_norm": 0.35308781266212463,
+      "learning_rate": 1.731868964211324e-05,
+      "loss": 0.1206,
+      "step": 7080
+    },
+    {
+      "epoch": 0.6712743798523007,
+      "grad_norm": 0.2664026916027069,
+      "learning_rate": 1.73149024805908e-05,
+      "loss": 0.1132,
+      "step": 7090
+    },
+    {
+      "epoch": 0.6722211702329104,
+      "grad_norm": 0.30298927426338196,
+      "learning_rate": 1.731111531906836e-05,
+      "loss": 0.1095,
+      "step": 7100
+    },
+    {
+      "epoch": 0.6731679606135201,
+      "grad_norm": 0.33091816306114197,
+      "learning_rate": 1.7307328157545922e-05,
+      "loss": 0.1047,
+      "step": 7110
+    },
+    {
+      "epoch": 0.6741147509941299,
+      "grad_norm": 0.31363001465797424,
+      "learning_rate": 1.7303540996023482e-05,
+      "loss": 0.1074,
+      "step": 7120
+    },
+    {
+      "epoch": 0.6750615413747396,
+      "grad_norm": 0.3875551223754883,
+      "learning_rate": 1.7299753834501042e-05,
+      "loss": 0.1185,
+      "step": 7130
+    },
+    {
+      "epoch": 0.6760083317553494,
+      "grad_norm": 0.3871995210647583,
+      "learning_rate": 1.7295966672978602e-05,
+      "loss": 0.1232,
+      "step": 7140
+    },
+    {
+      "epoch": 0.6769551221359591,
+      "grad_norm": 0.3625803589820862,
+      "learning_rate": 1.7292179511456165e-05,
+      "loss": 0.1158,
+      "step": 7150
+    },
+    {
+      "epoch": 0.6779019125165688,
+      "grad_norm": 0.4107898473739624,
+      "learning_rate": 1.7288392349933726e-05,
+      "loss": 0.1143,
+      "step": 7160
+    },
+    {
+      "epoch": 0.6788487028971786,
+      "grad_norm": 0.2983371615409851,
+      "learning_rate": 1.728460518841129e-05,
+      "loss": 0.1095,
+      "step": 7170
+    },
+    {
+      "epoch": 0.6797954932777883,
+      "grad_norm": 0.26490136981010437,
+      "learning_rate": 1.728081802688885e-05,
+      "loss": 0.1129,
+      "step": 7180
+    },
+    {
+      "epoch": 0.6807422836583981,
+      "grad_norm": 0.327644020318985,
+      "learning_rate": 1.727703086536641e-05,
+      "loss": 0.1198,
+      "step": 7190
+    },
+    {
+      "epoch": 0.6816890740390078,
+      "grad_norm": 0.37036940455436707,
+      "learning_rate": 1.727324370384397e-05,
+      "loss": 0.127,
+      "step": 7200
+    },
+    {
+      "epoch": 0.6826358644196175,
+      "grad_norm": 0.30712732672691345,
+      "learning_rate": 1.7269456542321533e-05,
+      "loss": 0.1046,
+      "step": 7210
+    },
+    {
+      "epoch": 0.6835826548002273,
+      "grad_norm": 0.30205264687538147,
+      "learning_rate": 1.7265669380799093e-05,
+      "loss": 0.1131,
+      "step": 7220
+    },
+    {
+      "epoch": 0.6845294451808369,
+      "grad_norm": 0.27775585651397705,
+      "learning_rate": 1.7261882219276653e-05,
+      "loss": 0.1251,
+      "step": 7230
+    },
+    {
+      "epoch": 0.6854762355614467,
+      "grad_norm": 0.3181157410144806,
+      "learning_rate": 1.7258095057754213e-05,
+      "loss": 0.1225,
+      "step": 7240
+    },
+    {
+      "epoch": 0.6864230259420564,
+      "grad_norm": 0.3210241496562958,
+      "learning_rate": 1.7254307896231776e-05,
+      "loss": 0.1037,
+      "step": 7250
+    },
+    {
+      "epoch": 0.6873698163226661,
+      "grad_norm": 0.36600542068481445,
+      "learning_rate": 1.7250520734709336e-05,
+      "loss": 0.1176,
+      "step": 7260
+    },
+    {
+      "epoch": 0.6883166067032759,
+      "grad_norm": 0.455118864774704,
+      "learning_rate": 1.7246733573186896e-05,
+      "loss": 0.1241,
+      "step": 7270
+    },
+    {
+      "epoch": 0.6892633970838856,
+      "grad_norm": 0.3404286205768585,
+      "learning_rate": 1.724294641166446e-05,
+      "loss": 0.1216,
+      "step": 7280
+    },
+    {
+      "epoch": 0.6902101874644954,
+      "grad_norm": 0.33997026085853577,
+      "learning_rate": 1.723915925014202e-05,
+      "loss": 0.1202,
+      "step": 7290
+    },
+    {
+      "epoch": 0.6911569778451051,
+      "grad_norm": 0.3330562710762024,
+      "learning_rate": 1.7235372088619583e-05,
+      "loss": 0.1106,
+      "step": 7300
+    },
+    {
+      "epoch": 0.6921037682257148,
+      "grad_norm": 0.29542532563209534,
+      "learning_rate": 1.7231584927097143e-05,
+      "loss": 0.1236,
+      "step": 7310
+    },
+    {
+      "epoch": 0.6930505586063246,
+      "grad_norm": 0.38714075088500977,
+      "learning_rate": 1.7227797765574703e-05,
+      "loss": 0.1203,
+      "step": 7320
+    },
+    {
+      "epoch": 0.6939973489869343,
+      "grad_norm": 0.36225587129592896,
+      "learning_rate": 1.7224010604052263e-05,
+      "loss": 0.1127,
+      "step": 7330
+    },
+    {
+      "epoch": 0.6949441393675441,
+      "grad_norm": 0.3260248899459839,
+      "learning_rate": 1.7220223442529823e-05,
+      "loss": 0.1212,
+      "step": 7340
+    },
+    {
+      "epoch": 0.6958909297481538,
+      "grad_norm": 0.3012619912624359,
+      "learning_rate": 1.7216436281007387e-05,
+      "loss": 0.1196,
+      "step": 7350
+    },
+    {
+      "epoch": 0.6968377201287634,
+      "grad_norm": 0.3051506280899048,
+      "learning_rate": 1.7212649119484947e-05,
+      "loss": 0.1118,
+      "step": 7360
+    },
+    {
+      "epoch": 0.6977845105093732,
+      "grad_norm": 0.33954405784606934,
+      "learning_rate": 1.7208861957962507e-05,
+      "loss": 0.1286,
+      "step": 7370
+    },
+    {
+      "epoch": 0.6987313008899829,
+      "grad_norm": 0.2781881093978882,
+      "learning_rate": 1.720507479644007e-05,
+      "loss": 0.1107,
+      "step": 7380
+    },
+    {
+      "epoch": 0.6996780912705927,
+      "grad_norm": 0.4584430456161499,
+      "learning_rate": 1.720128763491763e-05,
+      "loss": 0.11,
+      "step": 7390
+    },
+    {
+      "epoch": 0.7006248816512024,
+      "grad_norm": 0.30674710869789124,
+      "learning_rate": 1.7197500473395194e-05,
+      "loss": 0.127,
+      "step": 7400
+    },
+    {
+      "epoch": 0.7015716720318121,
+      "grad_norm": 0.2992197573184967,
+      "learning_rate": 1.7193713311872754e-05,
+      "loss": 0.1096,
+      "step": 7410
+    },
+    {
+      "epoch": 0.7025184624124219,
+      "grad_norm": 0.35374733805656433,
+      "learning_rate": 1.7189926150350314e-05,
+      "loss": 0.1053,
+      "step": 7420
+    },
+    {
+      "epoch": 0.7034652527930316,
+      "grad_norm": 0.4027782082557678,
+      "learning_rate": 1.7186138988827874e-05,
+      "loss": 0.1238,
+      "step": 7430
+    },
+    {
+      "epoch": 0.7044120431736414,
+      "grad_norm": 0.2525242269039154,
+      "learning_rate": 1.7182351827305437e-05,
+      "loss": 0.1187,
+      "step": 7440
+    },
+    {
+      "epoch": 0.7053588335542511,
+      "grad_norm": 0.34557586908340454,
+      "learning_rate": 1.7178564665782997e-05,
+      "loss": 0.1149,
+      "step": 7450
+    },
+    {
+      "epoch": 0.7063056239348608,
+      "grad_norm": 0.3434407413005829,
+      "learning_rate": 1.7174777504260558e-05,
+      "loss": 0.1202,
+      "step": 7460
+    },
+    {
+      "epoch": 0.7072524143154706,
+      "grad_norm": 0.3270295262336731,
+      "learning_rate": 1.7170990342738118e-05,
+      "loss": 0.1214,
+      "step": 7470
+    },
+    {
+      "epoch": 0.7081992046960803,
+      "grad_norm": 0.3213960528373718,
+      "learning_rate": 1.7167203181215678e-05,
+      "loss": 0.1166,
+      "step": 7480
+    },
+    {
+      "epoch": 0.7091459950766901,
+      "grad_norm": 0.4006138741970062,
+      "learning_rate": 1.716341601969324e-05,
+      "loss": 0.1221,
+      "step": 7490
+    },
+    {
+      "epoch": 0.7100927854572997,
+      "grad_norm": 0.36110085248947144,
+      "learning_rate": 1.71596288581708e-05,
+      "loss": 0.1167,
+      "step": 7500
+    },
+    {
+      "epoch": 0.7110395758379094,
+      "grad_norm": 0.3229648172855377,
+      "learning_rate": 1.7155841696648365e-05,
+      "loss": 0.1179,
+      "step": 7510
+    },
+    {
+      "epoch": 0.7119863662185192,
+      "grad_norm": 0.3359043300151825,
+      "learning_rate": 1.7152054535125925e-05,
+      "loss": 0.1186,
+      "step": 7520
+    },
+    {
+      "epoch": 0.7129331565991289,
+      "grad_norm": 0.32050198316574097,
+      "learning_rate": 1.7148267373603488e-05,
+      "loss": 0.1256,
+      "step": 7530
+    },
+    {
+      "epoch": 0.7138799469797387,
+      "grad_norm": 0.31654369831085205,
+      "learning_rate": 1.7144480212081048e-05,
+      "loss": 0.1185,
+      "step": 7540
+    },
+    {
+      "epoch": 0.7148267373603484,
+      "grad_norm": 0.35290828347206116,
+      "learning_rate": 1.7140693050558608e-05,
+      "loss": 0.1112,
+      "step": 7550
+    },
+    {
+      "epoch": 0.7157735277409581,
+      "grad_norm": 0.3996545076370239,
+      "learning_rate": 1.7136905889036168e-05,
+      "loss": 0.1186,
+      "step": 7560
+    },
+    {
+      "epoch": 0.7167203181215679,
+      "grad_norm": 0.33729514479637146,
+      "learning_rate": 1.7133118727513728e-05,
+      "loss": 0.1138,
+      "step": 7570
+    },
+    {
+      "epoch": 0.7176671085021776,
+      "grad_norm": 0.2941891849040985,
+      "learning_rate": 1.712933156599129e-05,
+      "loss": 0.113,
+      "step": 7580
+    },
+    {
+      "epoch": 0.7186138988827874,
+      "grad_norm": 0.28606683015823364,
+      "learning_rate": 1.7125544404468852e-05,
+      "loss": 0.1114,
+      "step": 7590
+    },
+    {
+      "epoch": 0.7195606892633971,
+      "grad_norm": 0.4108836054801941,
+      "learning_rate": 1.7121757242946412e-05,
+      "loss": 0.1291,
+      "step": 7600
+    },
+    {
+      "epoch": 0.7205074796440069,
+      "grad_norm": 0.3045104742050171,
+      "learning_rate": 1.7117970081423975e-05,
+      "loss": 0.1146,
+      "step": 7610
+    },
+    {
+      "epoch": 0.7214542700246166,
+      "grad_norm": 0.3012486398220062,
+      "learning_rate": 1.7114182919901535e-05,
+      "loss": 0.1104,
+      "step": 7620
+    },
+    {
+      "epoch": 0.7224010604052263,
+      "grad_norm": 0.28472423553466797,
+      "learning_rate": 1.7110395758379095e-05,
+      "loss": 0.0998,
+      "step": 7630
+    },
+    {
+      "epoch": 0.723347850785836,
+      "grad_norm": 0.3129308223724365,
+      "learning_rate": 1.710660859685666e-05,
+      "loss": 0.1154,
+      "step": 7640
+    },
+    {
+      "epoch": 0.7242946411664457,
+      "grad_norm": 0.33357781171798706,
+      "learning_rate": 1.710282143533422e-05,
+      "loss": 0.1186,
+      "step": 7650
+    },
+    {
+      "epoch": 0.7252414315470554,
+      "grad_norm": 0.300777792930603,
+      "learning_rate": 1.709903427381178e-05,
+      "loss": 0.1217,
+      "step": 7660
+    },
+    {
+      "epoch": 0.7261882219276652,
+      "grad_norm": 0.3664512634277344,
+      "learning_rate": 1.7095247112289342e-05,
+      "loss": 0.1069,
+      "step": 7670
+    },
+    {
+      "epoch": 0.7271350123082749,
+      "grad_norm": 0.3453938364982605,
+      "learning_rate": 1.7091459950766902e-05,
+      "loss": 0.1149,
+      "step": 7680
+    },
+    {
+      "epoch": 0.7280818026888847,
+      "grad_norm": 0.24504321813583374,
+      "learning_rate": 1.7087672789244462e-05,
+      "loss": 0.111,
+      "step": 7690
+    },
+    {
+      "epoch": 0.7290285930694944,
+      "grad_norm": 0.38600441813468933,
+      "learning_rate": 1.7083885627722022e-05,
+      "loss": 0.1129,
+      "step": 7700
+    },
+    {
+      "epoch": 0.7299753834501042,
+      "grad_norm": 0.35876065492630005,
+      "learning_rate": 1.7080098466199582e-05,
+      "loss": 0.1144,
+      "step": 7710
+    },
+    {
+      "epoch": 0.7309221738307139,
+      "grad_norm": 0.3433103859424591,
+      "learning_rate": 1.7076311304677146e-05,
+      "loss": 0.1145,
+      "step": 7720
+    },
+    {
+      "epoch": 0.7318689642113236,
+      "grad_norm": 0.4363567531108856,
+      "learning_rate": 1.7072524143154706e-05,
+      "loss": 0.1068,
+      "step": 7730
+    },
+    {
+      "epoch": 0.7328157545919334,
+      "grad_norm": 0.2609582841396332,
+      "learning_rate": 1.706873698163227e-05,
+      "loss": 0.1039,
+      "step": 7740
+    },
+    {
+      "epoch": 0.7337625449725431,
+      "grad_norm": 0.251044899225235,
+      "learning_rate": 1.706494982010983e-05,
+      "loss": 0.1292,
+      "step": 7750
+    },
+    {
+      "epoch": 0.7347093353531529,
+      "grad_norm": 0.3957395553588867,
+      "learning_rate": 1.706116265858739e-05,
+      "loss": 0.1105,
+      "step": 7760
+    },
+    {
+      "epoch": 0.7356561257337626,
+      "grad_norm": 0.3416815996170044,
+      "learning_rate": 1.7057375497064953e-05,
+      "loss": 0.1149,
+      "step": 7770
+    },
+    {
+      "epoch": 0.7366029161143722,
+      "grad_norm": 0.3129086196422577,
+      "learning_rate": 1.7053588335542513e-05,
+      "loss": 0.1176,
+      "step": 7780
+    },
+    {
+      "epoch": 0.737549706494982,
+      "grad_norm": 0.3613971471786499,
+      "learning_rate": 1.7049801174020073e-05,
+      "loss": 0.1061,
+      "step": 7790
+    },
+    {
+      "epoch": 0.7384964968755917,
+      "grad_norm": 0.32689687609672546,
+      "learning_rate": 1.7046014012497633e-05,
+      "loss": 0.1111,
+      "step": 7800
+    },
+    {
+      "epoch": 0.7394432872562015,
+      "grad_norm": 0.30839022994041443,
+      "learning_rate": 1.7042226850975197e-05,
+      "loss": 0.127,
+      "step": 7810
+    },
+    {
+      "epoch": 0.7403900776368112,
+      "grad_norm": 0.481495201587677,
+      "learning_rate": 1.7038439689452757e-05,
+      "loss": 0.1147,
+      "step": 7820
+    },
+    {
+      "epoch": 0.7413368680174209,
+      "grad_norm": 0.27305448055267334,
+      "learning_rate": 1.7034652527930317e-05,
+      "loss": 0.1233,
+      "step": 7830
+    },
+    {
+      "epoch": 0.7422836583980307,
+      "grad_norm": 0.35240069031715393,
+      "learning_rate": 1.7030865366407877e-05,
+      "loss": 0.114,
+      "step": 7840
+    },
+    {
+      "epoch": 0.7432304487786404,
+      "grad_norm": 0.39616870880126953,
+      "learning_rate": 1.702707820488544e-05,
+      "loss": 0.1218,
+      "step": 7850
+    },
+    {
+      "epoch": 0.7441772391592502,
+      "grad_norm": 0.2716734707355499,
+      "learning_rate": 1.7023291043363e-05,
+      "loss": 0.1211,
+      "step": 7860
+    },
+    {
+      "epoch": 0.7451240295398599,
+      "grad_norm": 0.26934340596199036,
+      "learning_rate": 1.7019503881840564e-05,
+      "loss": 0.1042,
+      "step": 7870
+    },
+    {
+      "epoch": 0.7460708199204696,
+      "grad_norm": 0.27856406569480896,
+      "learning_rate": 1.7015716720318124e-05,
+      "loss": 0.1126,
+      "step": 7880
+    },
+    {
+      "epoch": 0.7470176103010794,
+      "grad_norm": 0.2654203772544861,
+      "learning_rate": 1.7011929558795684e-05,
+      "loss": 0.1286,
+      "step": 7890
+    },
+    {
+      "epoch": 0.7479644006816891,
+      "grad_norm": 0.29593995213508606,
+      "learning_rate": 1.7008142397273244e-05,
+      "loss": 0.1136,
+      "step": 7900
+    },
+    {
+      "epoch": 0.7489111910622988,
+      "grad_norm": 0.32165074348449707,
+      "learning_rate": 1.7004355235750807e-05,
+      "loss": 0.1229,
+      "step": 7910
+    },
+    {
+      "epoch": 0.7498579814429085,
+      "grad_norm": 0.2874644696712494,
+      "learning_rate": 1.7000568074228367e-05,
+      "loss": 0.1071,
+      "step": 7920
+    },
+    {
+      "epoch": 0.7508047718235182,
+      "grad_norm": 0.3464849293231964,
+      "learning_rate": 1.6996780912705927e-05,
+      "loss": 0.1043,
+      "step": 7930
+    },
+    {
+      "epoch": 0.751751562204128,
+      "grad_norm": 0.27734798192977905,
+      "learning_rate": 1.6992993751183487e-05,
+      "loss": 0.1211,
+      "step": 7940
+    },
+    {
+      "epoch": 0.7526983525847377,
+      "grad_norm": 0.2745431065559387,
+      "learning_rate": 1.698920658966105e-05,
+      "loss": 0.1188,
+      "step": 7950
+    },
+    {
+      "epoch": 0.7536451429653475,
+      "grad_norm": 0.3578427731990814,
+      "learning_rate": 1.698541942813861e-05,
+      "loss": 0.1201,
+      "step": 7960
+    },
+    {
+      "epoch": 0.7545919333459572,
+      "grad_norm": 0.26581892371177673,
+      "learning_rate": 1.6981632266616174e-05,
+      "loss": 0.1136,
+      "step": 7970
+    },
+    {
+      "epoch": 0.7555387237265669,
+      "grad_norm": 0.32221293449401855,
+      "learning_rate": 1.6977845105093734e-05,
+      "loss": 0.11,
+      "step": 7980
+    },
+    {
+      "epoch": 0.7564855141071767,
+      "grad_norm": 0.2992156445980072,
+      "learning_rate": 1.6974057943571294e-05,
+      "loss": 0.1014,
+      "step": 7990
+    },
+    {
+      "epoch": 0.7574323044877864,
+      "grad_norm": 0.3733976483345032,
+      "learning_rate": 1.6970270782048858e-05,
+      "loss": 0.1099,
+      "step": 8000
+    },
+    {
+      "epoch": 0.7583790948683962,
+      "grad_norm": 0.25940683484077454,
+      "learning_rate": 1.6966483620526418e-05,
+      "loss": 0.105,
+      "step": 8010
+    },
+    {
+      "epoch": 0.7593258852490059,
+      "grad_norm": 0.3570631742477417,
+      "learning_rate": 1.6962696459003978e-05,
+      "loss": 0.1009,
+      "step": 8020
+    },
+    {
+      "epoch": 0.7602726756296156,
+      "grad_norm": 0.30129891633987427,
+      "learning_rate": 1.6958909297481538e-05,
+      "loss": 0.1156,
+      "step": 8030
+    },
+    {
+      "epoch": 0.7612194660102254,
+      "grad_norm": 0.3454397916793823,
+      "learning_rate": 1.6955122135959098e-05,
+      "loss": 0.1191,
+      "step": 8040
+    },
+    {
+      "epoch": 0.762166256390835,
+      "grad_norm": 0.2947542071342468,
+      "learning_rate": 1.695133497443666e-05,
+      "loss": 0.1053,
+      "step": 8050
+    },
+    {
+      "epoch": 0.7631130467714448,
+      "grad_norm": 0.3459526300430298,
+      "learning_rate": 1.694754781291422e-05,
+      "loss": 0.1187,
+      "step": 8060
+    },
+    {
+      "epoch": 0.7640598371520545,
+      "grad_norm": 0.3022899031639099,
+      "learning_rate": 1.694376065139178e-05,
+      "loss": 0.1118,
+      "step": 8070
+    },
+    {
+      "epoch": 0.7650066275326642,
+      "grad_norm": 0.3002983331680298,
+      "learning_rate": 1.6939973489869345e-05,
+      "loss": 0.121,
+      "step": 8080
+    },
+    {
+      "epoch": 0.765953417913274,
+      "grad_norm": 0.2755148410797119,
+      "learning_rate": 1.6936186328346905e-05,
+      "loss": 0.1182,
+      "step": 8090
+    },
+    {
+      "epoch": 0.7669002082938837,
+      "grad_norm": 0.34115657210350037,
+      "learning_rate": 1.693239916682447e-05,
+      "loss": 0.1171,
+      "step": 8100
+    },
+    {
+      "epoch": 0.7678469986744935,
+      "grad_norm": 0.24916787445545197,
+      "learning_rate": 1.692861200530203e-05,
+      "loss": 0.1161,
+      "step": 8110
+    },
+    {
+      "epoch": 0.7687937890551032,
+      "grad_norm": 0.27540966868400574,
+      "learning_rate": 1.692482484377959e-05,
+      "loss": 0.1125,
+      "step": 8120
+    },
+    {
+      "epoch": 0.769740579435713,
+      "grad_norm": 0.3216499090194702,
+      "learning_rate": 1.692103768225715e-05,
+      "loss": 0.1069,
+      "step": 8130
+    },
+    {
+      "epoch": 0.7706873698163227,
+      "grad_norm": 0.33225181698799133,
+      "learning_rate": 1.6917250520734712e-05,
+      "loss": 0.1101,
+      "step": 8140
+    },
+    {
+      "epoch": 0.7716341601969324,
+      "grad_norm": 0.3109791576862335,
+      "learning_rate": 1.6913463359212272e-05,
+      "loss": 0.1137,
+      "step": 8150
+    },
+    {
+      "epoch": 0.7725809505775422,
+      "grad_norm": 0.3619652986526489,
+      "learning_rate": 1.6909676197689832e-05,
+      "loss": 0.1202,
+      "step": 8160
+    },
+    {
+      "epoch": 0.7735277409581519,
+      "grad_norm": 0.3975580930709839,
+      "learning_rate": 1.6905889036167392e-05,
+      "loss": 0.1161,
+      "step": 8170
+    },
+    {
+      "epoch": 0.7744745313387617,
+      "grad_norm": 0.35568127036094666,
+      "learning_rate": 1.6902101874644952e-05,
+      "loss": 0.1113,
+      "step": 8180
+    },
+    {
+      "epoch": 0.7754213217193713,
+      "grad_norm": 0.356973260641098,
+      "learning_rate": 1.6898314713122516e-05,
+      "loss": 0.1047,
+      "step": 8190
+    },
+    {
+      "epoch": 0.776368112099981,
+      "grad_norm": 0.4212486445903778,
+      "learning_rate": 1.6894527551600076e-05,
+      "loss": 0.1096,
+      "step": 8200
+    },
+    {
+      "epoch": 0.7773149024805908,
+      "grad_norm": 0.34463343024253845,
+      "learning_rate": 1.689074039007764e-05,
+      "loss": 0.1244,
+      "step": 8210
+    },
+    {
+      "epoch": 0.7782616928612005,
+      "grad_norm": 0.3768672049045563,
+      "learning_rate": 1.68869532285552e-05,
+      "loss": 0.1158,
+      "step": 8220
+    },
+    {
+      "epoch": 0.7792084832418102,
+      "grad_norm": 0.36685991287231445,
+      "learning_rate": 1.6883166067032763e-05,
+      "loss": 0.1141,
+      "step": 8230
+    },
+    {
+      "epoch": 0.78015527362242,
+      "grad_norm": 0.3462464213371277,
+      "learning_rate": 1.6879378905510323e-05,
+      "loss": 0.1173,
+      "step": 8240
+    },
+    {
+      "epoch": 0.7811020640030297,
+      "grad_norm": 0.38483893871307373,
+      "learning_rate": 1.6875591743987883e-05,
+      "loss": 0.1107,
+      "step": 8250
+    },
+    {
+      "epoch": 0.7820488543836395,
+      "grad_norm": 0.3127537667751312,
+      "learning_rate": 1.6871804582465443e-05,
+      "loss": 0.1179,
+      "step": 8260
+    },
+    {
+      "epoch": 0.7829956447642492,
+      "grad_norm": 0.39382901787757874,
+      "learning_rate": 1.6868017420943003e-05,
+      "loss": 0.121,
+      "step": 8270
+    },
+    {
+      "epoch": 0.783942435144859,
+      "grad_norm": 0.35741671919822693,
+      "learning_rate": 1.6864230259420566e-05,
+      "loss": 0.1101,
+      "step": 8280
+    },
+    {
+      "epoch": 0.7848892255254687,
+      "grad_norm": 0.3009890615940094,
+      "learning_rate": 1.6860443097898126e-05,
+      "loss": 0.1231,
+      "step": 8290
+    },
+    {
+      "epoch": 0.7858360159060784,
+      "grad_norm": 0.34346726536750793,
+      "learning_rate": 1.6856655936375686e-05,
+      "loss": 0.1222,
+      "step": 8300
+    },
+    {
+      "epoch": 0.7867828062866882,
+      "grad_norm": 0.43910688161849976,
+      "learning_rate": 1.685286877485325e-05,
+      "loss": 0.1129,
+      "step": 8310
+    },
+    {
+      "epoch": 0.7877295966672979,
+      "grad_norm": 0.36337924003601074,
+      "learning_rate": 1.684908161333081e-05,
+      "loss": 0.1164,
+      "step": 8320
+    },
+    {
+      "epoch": 0.7886763870479075,
+      "grad_norm": 0.3842446208000183,
+      "learning_rate": 1.6845294451808373e-05,
+      "loss": 0.12,
+      "step": 8330
+    },
+    {
+      "epoch": 0.7896231774285173,
+      "grad_norm": 0.4252307713031769,
+      "learning_rate": 1.6841507290285933e-05,
+      "loss": 0.1145,
+      "step": 8340
+    },
+    {
+      "epoch": 0.790569967809127,
+      "grad_norm": 0.3211591839790344,
+      "learning_rate": 1.6837720128763493e-05,
+      "loss": 0.1016,
+      "step": 8350
+    },
+    {
+      "epoch": 0.7915167581897368,
+      "grad_norm": 0.35847267508506775,
+      "learning_rate": 1.6833932967241053e-05,
+      "loss": 0.1064,
+      "step": 8360
+    },
+    {
+      "epoch": 0.7924635485703465,
+      "grad_norm": 0.3097435534000397,
+      "learning_rate": 1.6830145805718617e-05,
+      "loss": 0.1195,
+      "step": 8370
+    },
+    {
+      "epoch": 0.7934103389509563,
+      "grad_norm": 0.43024635314941406,
+      "learning_rate": 1.6826358644196177e-05,
+      "loss": 0.106,
+      "step": 8380
+    },
+    {
+      "epoch": 0.794357129331566,
+      "grad_norm": 0.23815903067588806,
+      "learning_rate": 1.6822571482673737e-05,
+      "loss": 0.1055,
+      "step": 8390
+    },
+    {
+      "epoch": 0.7953039197121757,
+      "grad_norm": 0.32491034269332886,
+      "learning_rate": 1.6818784321151297e-05,
+      "loss": 0.1108,
+      "step": 8400
+    },
+    {
+      "epoch": 0.7962507100927855,
+      "grad_norm": 0.3355867862701416,
+      "learning_rate": 1.6814997159628857e-05,
+      "loss": 0.1172,
+      "step": 8410
+    },
+    {
+      "epoch": 0.7971975004733952,
+      "grad_norm": 0.4188220500946045,
+      "learning_rate": 1.681120999810642e-05,
+      "loss": 0.1208,
+      "step": 8420
+    },
+    {
+      "epoch": 0.798144290854005,
+      "grad_norm": 0.4368966817855835,
+      "learning_rate": 1.680742283658398e-05,
+      "loss": 0.1239,
+      "step": 8430
+    },
+    {
+      "epoch": 0.7990910812346147,
+      "grad_norm": 0.28890368342399597,
+      "learning_rate": 1.6803635675061544e-05,
+      "loss": 0.1179,
+      "step": 8440
+    },
+    {
+      "epoch": 0.8000378716152244,
+      "grad_norm": 0.28585273027420044,
+      "learning_rate": 1.6799848513539104e-05,
+      "loss": 0.1128,
+      "step": 8450
+    },
+    {
+      "epoch": 0.8009846619958341,
+      "grad_norm": 0.32358139753341675,
+      "learning_rate": 1.6796061352016667e-05,
+      "loss": 0.1168,
+      "step": 8460
+    },
+    {
+      "epoch": 0.8019314523764438,
+      "grad_norm": 0.28638800978660583,
+      "learning_rate": 1.6792274190494228e-05,
+      "loss": 0.1167,
+      "step": 8470
+    },
+    {
+      "epoch": 0.8028782427570536,
+      "grad_norm": 0.2773953974246979,
+      "learning_rate": 1.6788487028971788e-05,
+      "loss": 0.1208,
+      "step": 8480
+    },
+    {
+      "epoch": 0.8038250331376633,
+      "grad_norm": 0.34793245792388916,
+      "learning_rate": 1.6784699867449348e-05,
+      "loss": 0.1176,
+      "step": 8490
+    },
+    {
+      "epoch": 0.804771823518273,
+      "grad_norm": 0.3978104591369629,
+      "learning_rate": 1.6780912705926908e-05,
+      "loss": 0.1153,
+      "step": 8500
+    },
+    {
+      "epoch": 0.8057186138988828,
+      "grad_norm": 0.2925117313861847,
+      "learning_rate": 1.677712554440447e-05,
+      "loss": 0.1156,
+      "step": 8510
+    },
+    {
+      "epoch": 0.8066654042794925,
+      "grad_norm": 0.30174776911735535,
+      "learning_rate": 1.677333838288203e-05,
+      "loss": 0.1177,
+      "step": 8520
+    },
+    {
+      "epoch": 0.8076121946601023,
+      "grad_norm": 0.27571502327919006,
+      "learning_rate": 1.676955122135959e-05,
+      "loss": 0.1082,
+      "step": 8530
+    },
+    {
+      "epoch": 0.808558985040712,
+      "grad_norm": 0.34001198410987854,
+      "learning_rate": 1.676576405983715e-05,
+      "loss": 0.1208,
+      "step": 8540
+    },
+    {
+      "epoch": 0.8095057754213217,
+      "grad_norm": 0.3395918309688568,
+      "learning_rate": 1.6761976898314715e-05,
+      "loss": 0.1197,
+      "step": 8550
+    },
+    {
+      "epoch": 0.8104525658019315,
+      "grad_norm": 0.31503212451934814,
+      "learning_rate": 1.6758189736792275e-05,
+      "loss": 0.1151,
+      "step": 8560
+    },
+    {
+      "epoch": 0.8113993561825412,
+      "grad_norm": 0.2970471680164337,
+      "learning_rate": 1.6754402575269838e-05,
+      "loss": 0.1102,
+      "step": 8570
+    },
+    {
+      "epoch": 0.812346146563151,
+      "grad_norm": 0.2834688425064087,
+      "learning_rate": 1.6750615413747398e-05,
+      "loss": 0.1163,
+      "step": 8580
+    },
+    {
+      "epoch": 0.8132929369437607,
+      "grad_norm": 0.4195939600467682,
+      "learning_rate": 1.6746828252224958e-05,
+      "loss": 0.1146,
+      "step": 8590
+    },
+    {
+      "epoch": 0.8142397273243703,
+      "grad_norm": 0.2607402801513672,
+      "learning_rate": 1.6743041090702522e-05,
+      "loss": 0.1109,
+      "step": 8600
+    },
+    {
+      "epoch": 0.8151865177049801,
+      "grad_norm": 0.2828407883644104,
+      "learning_rate": 1.6739253929180082e-05,
+      "loss": 0.1215,
+      "step": 8610
+    },
+    {
+      "epoch": 0.8161333080855898,
+      "grad_norm": 0.2749273180961609,
+      "learning_rate": 1.6735466767657642e-05,
+      "loss": 0.1039,
+      "step": 8620
+    },
+    {
+      "epoch": 0.8170800984661996,
+      "grad_norm": 0.3468409478664398,
+      "learning_rate": 1.6731679606135202e-05,
+      "loss": 0.1093,
+      "step": 8630
+    },
+    {
+      "epoch": 0.8180268888468093,
+      "grad_norm": 0.30343708395957947,
+      "learning_rate": 1.6727892444612762e-05,
+      "loss": 0.1089,
+      "step": 8640
+    },
+    {
+      "epoch": 0.818973679227419,
+      "grad_norm": 0.33821362257003784,
+      "learning_rate": 1.6724105283090325e-05,
+      "loss": 0.1059,
+      "step": 8650
+    },
+    {
+      "epoch": 0.8199204696080288,
+      "grad_norm": 0.3864172101020813,
+      "learning_rate": 1.6720318121567885e-05,
+      "loss": 0.1176,
+      "step": 8660
+    },
+    {
+      "epoch": 0.8208672599886385,
+      "grad_norm": 0.4220865070819855,
+      "learning_rate": 1.671653096004545e-05,
+      "loss": 0.1213,
+      "step": 8670
+    },
+    {
+      "epoch": 0.8218140503692483,
+      "grad_norm": 0.2996737062931061,
+      "learning_rate": 1.671274379852301e-05,
+      "loss": 0.11,
+      "step": 8680
+    },
+    {
+      "epoch": 0.822760840749858,
+      "grad_norm": 0.3411973714828491,
+      "learning_rate": 1.670895663700057e-05,
+      "loss": 0.1227,
+      "step": 8690
+    },
+    {
+      "epoch": 0.8237076311304677,
+      "grad_norm": 0.34379059076309204,
+      "learning_rate": 1.6705169475478132e-05,
+      "loss": 0.1144,
+      "step": 8700
+    },
+    {
+      "epoch": 0.8246544215110775,
+      "grad_norm": 0.3353620171546936,
+      "learning_rate": 1.6701382313955692e-05,
+      "loss": 0.1053,
+      "step": 8710
+    },
+    {
+      "epoch": 0.8256012118916872,
+      "grad_norm": 0.3999831974506378,
+      "learning_rate": 1.6697595152433252e-05,
+      "loss": 0.1205,
+      "step": 8720
+    },
+    {
+      "epoch": 0.826548002272297,
+      "grad_norm": 0.34344691038131714,
+      "learning_rate": 1.6693807990910813e-05,
+      "loss": 0.1145,
+      "step": 8730
+    },
+    {
+      "epoch": 0.8274947926529066,
+      "grad_norm": 0.3627908229827881,
+      "learning_rate": 1.6690020829388376e-05,
+      "loss": 0.1125,
+      "step": 8740
+    },
+    {
+      "epoch": 0.8284415830335163,
+      "grad_norm": 0.40607041120529175,
+      "learning_rate": 1.6686233667865936e-05,
+      "loss": 0.1202,
+      "step": 8750
+    },
+    {
+      "epoch": 0.8293883734141261,
+      "grad_norm": 0.278888076543808,
+      "learning_rate": 1.6682446506343496e-05,
+      "loss": 0.1136,
+      "step": 8760
+    },
+    {
+      "epoch": 0.8303351637947358,
+      "grad_norm": 0.32712674140930176,
+      "learning_rate": 1.6678659344821056e-05,
+      "loss": 0.1088,
+      "step": 8770
+    },
+    {
+      "epoch": 0.8312819541753456,
+      "grad_norm": 0.44736137986183167,
+      "learning_rate": 1.667487218329862e-05,
+      "loss": 0.1121,
+      "step": 8780
+    },
+    {
+      "epoch": 0.8322287445559553,
+      "grad_norm": 0.3264457583427429,
+      "learning_rate": 1.667108502177618e-05,
+      "loss": 0.1052,
+      "step": 8790
+    },
+    {
+      "epoch": 0.833175534936565,
+      "grad_norm": 0.3332121670246124,
+      "learning_rate": 1.6667297860253743e-05,
+      "loss": 0.1184,
+      "step": 8800
+    },
+    {
+      "epoch": 0.8341223253171748,
+      "grad_norm": 0.3029613196849823,
+      "learning_rate": 1.6663510698731303e-05,
+      "loss": 0.1133,
+      "step": 8810
+    },
+    {
+      "epoch": 0.8350691156977845,
+      "grad_norm": 0.33913642168045044,
+      "learning_rate": 1.6659723537208863e-05,
+      "loss": 0.1086,
+      "step": 8820
+    },
+    {
+      "epoch": 0.8360159060783943,
+      "grad_norm": 0.2835783362388611,
+      "learning_rate": 1.6655936375686423e-05,
+      "loss": 0.1174,
+      "step": 8830
+    },
+    {
+      "epoch": 0.836962696459004,
+      "grad_norm": 0.3837894797325134,
+      "learning_rate": 1.6652149214163987e-05,
+      "loss": 0.1116,
+      "step": 8840
+    },
+    {
+      "epoch": 0.8379094868396137,
+      "grad_norm": 0.29306384921073914,
+      "learning_rate": 1.6648362052641547e-05,
+      "loss": 0.1149,
+      "step": 8850
+    },
+    {
+      "epoch": 0.8388562772202235,
+      "grad_norm": 0.3184460997581482,
+      "learning_rate": 1.6644574891119107e-05,
+      "loss": 0.1169,
+      "step": 8860
+    },
+    {
+      "epoch": 0.8398030676008332,
+      "grad_norm": 0.3720015287399292,
+      "learning_rate": 1.6640787729596667e-05,
+      "loss": 0.1211,
+      "step": 8870
+    },
+    {
+      "epoch": 0.8407498579814429,
+      "grad_norm": 0.423492431640625,
+      "learning_rate": 1.663700056807423e-05,
+      "loss": 0.1099,
+      "step": 8880
+    },
+    {
+      "epoch": 0.8416966483620526,
+      "grad_norm": 0.32279732823371887,
+      "learning_rate": 1.663321340655179e-05,
+      "loss": 0.1116,
+      "step": 8890
+    },
+    {
+      "epoch": 0.8426434387426623,
+      "grad_norm": 0.26023128628730774,
+      "learning_rate": 1.662942624502935e-05,
+      "loss": 0.101,
+      "step": 8900
+    },
+    {
+      "epoch": 0.8435902291232721,
+      "grad_norm": 0.34170594811439514,
+      "learning_rate": 1.6625639083506914e-05,
+      "loss": 0.1267,
+      "step": 8910
+    },
+    {
+      "epoch": 0.8445370195038818,
+      "grad_norm": 0.27268069982528687,
+      "learning_rate": 1.6621851921984474e-05,
+      "loss": 0.1104,
+      "step": 8920
+    },
+    {
+      "epoch": 0.8454838098844916,
+      "grad_norm": 0.27401381731033325,
+      "learning_rate": 1.6618064760462037e-05,
+      "loss": 0.1003,
+      "step": 8930
+    },
+    {
+      "epoch": 0.8464306002651013,
+      "grad_norm": 0.3447090983390808,
+      "learning_rate": 1.6614277598939597e-05,
+      "loss": 0.1209,
+      "step": 8940
+    },
+    {
+      "epoch": 0.847377390645711,
+      "grad_norm": 0.3804704546928406,
+      "learning_rate": 1.6610490437417157e-05,
+      "loss": 0.1217,
+      "step": 8950
+    },
+    {
+      "epoch": 0.8483241810263208,
+      "grad_norm": 0.47949594259262085,
+      "learning_rate": 1.6606703275894717e-05,
+      "loss": 0.1173,
+      "step": 8960
+    },
+    {
+      "epoch": 0.8492709714069305,
+      "grad_norm": 0.3841295838356018,
+      "learning_rate": 1.6602916114372277e-05,
+      "loss": 0.1079,
+      "step": 8970
+    },
+    {
+      "epoch": 0.8502177617875403,
+      "grad_norm": 0.29418498277664185,
+      "learning_rate": 1.659912895284984e-05,
+      "loss": 0.111,
+      "step": 8980
+    },
+    {
+      "epoch": 0.85116455216815,
+      "grad_norm": 0.30468037724494934,
+      "learning_rate": 1.65953417913274e-05,
+      "loss": 0.116,
+      "step": 8990
+    },
+    {
+      "epoch": 0.8521113425487598,
+      "grad_norm": 0.2763366997241974,
+      "learning_rate": 1.659155462980496e-05,
+      "loss": 0.116,
+      "step": 9000
+    },
+    {
+      "epoch": 0.8530581329293694,
+      "grad_norm": 0.30675357580184937,
+      "learning_rate": 1.6587767468282524e-05,
+      "loss": 0.11,
+      "step": 9010
+    },
+    {
+      "epoch": 0.8540049233099791,
+      "grad_norm": 0.31186580657958984,
+      "learning_rate": 1.6583980306760084e-05,
+      "loss": 0.115,
+      "step": 9020
+    },
+    {
+      "epoch": 0.8549517136905889,
+      "grad_norm": 0.4842294156551361,
+      "learning_rate": 1.6580193145237648e-05,
+      "loss": 0.1288,
+      "step": 9030
+    },
+    {
+      "epoch": 0.8558985040711986,
+      "grad_norm": 0.3168499767780304,
+      "learning_rate": 1.6576405983715208e-05,
+      "loss": 0.1196,
+      "step": 9040
+    },
+    {
+      "epoch": 0.8568452944518083,
+      "grad_norm": 0.2683272361755371,
+      "learning_rate": 1.6572618822192768e-05,
+      "loss": 0.1193,
+      "step": 9050
+    },
+    {
+      "epoch": 0.8577920848324181,
+      "grad_norm": 0.3049313724040985,
+      "learning_rate": 1.6568831660670328e-05,
+      "loss": 0.1079,
+      "step": 9060
+    },
+    {
+      "epoch": 0.8587388752130278,
+      "grad_norm": 0.31613585352897644,
+      "learning_rate": 1.656504449914789e-05,
+      "loss": 0.1108,
+      "step": 9070
+    },
+    {
+      "epoch": 0.8596856655936376,
+      "grad_norm": 0.34152960777282715,
+      "learning_rate": 1.656125733762545e-05,
+      "loss": 0.113,
+      "step": 9080
+    },
+    {
+      "epoch": 0.8606324559742473,
+      "grad_norm": 0.3691774010658264,
+      "learning_rate": 1.655747017610301e-05,
+      "loss": 0.1079,
+      "step": 9090
+    },
+    {
+      "epoch": 0.861579246354857,
+      "grad_norm": 0.3918764591217041,
+      "learning_rate": 1.655368301458057e-05,
+      "loss": 0.1142,
+      "step": 9100
+    },
+    {
+      "epoch": 0.8625260367354668,
+      "grad_norm": 0.3079935610294342,
+      "learning_rate": 1.654989585305813e-05,
+      "loss": 0.114,
+      "step": 9110
+    },
+    {
+      "epoch": 0.8634728271160765,
+      "grad_norm": 0.31950023770332336,
+      "learning_rate": 1.6546108691535695e-05,
+      "loss": 0.103,
+      "step": 9120
+    },
+    {
+      "epoch": 0.8644196174966863,
+      "grad_norm": 0.2867056429386139,
+      "learning_rate": 1.6542321530013255e-05,
+      "loss": 0.1095,
+      "step": 9130
+    },
+    {
+      "epoch": 0.865366407877296,
+      "grad_norm": 0.4084359109401703,
+      "learning_rate": 1.653853436849082e-05,
+      "loss": 0.1185,
+      "step": 9140
+    },
+    {
+      "epoch": 0.8663131982579056,
+      "grad_norm": 0.3682761490345001,
+      "learning_rate": 1.653474720696838e-05,
+      "loss": 0.1198,
+      "step": 9150
+    },
+    {
+      "epoch": 0.8672599886385154,
+      "grad_norm": 0.3352651596069336,
+      "learning_rate": 1.6530960045445942e-05,
+      "loss": 0.1118,
+      "step": 9160
+    },
+    {
+      "epoch": 0.8682067790191251,
+      "grad_norm": 0.300908625125885,
+      "learning_rate": 1.6527172883923502e-05,
+      "loss": 0.1216,
+      "step": 9170
+    },
+    {
+      "epoch": 0.8691535693997349,
+      "grad_norm": 0.2699333131313324,
+      "learning_rate": 1.6523385722401062e-05,
+      "loss": 0.1061,
+      "step": 9180
+    },
+    {
+      "epoch": 0.8701003597803446,
+      "grad_norm": 0.34169745445251465,
+      "learning_rate": 1.6519598560878622e-05,
+      "loss": 0.1124,
+      "step": 9190
+    },
+    {
+      "epoch": 0.8710471501609544,
+      "grad_norm": 0.29521581530570984,
+      "learning_rate": 1.6515811399356182e-05,
+      "loss": 0.1248,
+      "step": 9200
+    },
+    {
+      "epoch": 0.8719939405415641,
+      "grad_norm": 0.2916036546230316,
+      "learning_rate": 1.6512024237833746e-05,
+      "loss": 0.1189,
+      "step": 9210
+    },
+    {
+      "epoch": 0.8729407309221738,
+      "grad_norm": 0.31541579961776733,
+      "learning_rate": 1.6508237076311306e-05,
+      "loss": 0.1112,
+      "step": 9220
+    },
+    {
+      "epoch": 0.8738875213027836,
+      "grad_norm": 0.9175082445144653,
+      "learning_rate": 1.6504449914788866e-05,
+      "loss": 0.1219,
+      "step": 9230
+    },
+    {
+      "epoch": 0.8748343116833933,
+      "grad_norm": 0.3295627236366272,
+      "learning_rate": 1.650066275326643e-05,
+      "loss": 0.1114,
+      "step": 9240
+    },
+    {
+      "epoch": 0.8757811020640031,
+      "grad_norm": 0.3011680245399475,
+      "learning_rate": 1.649687559174399e-05,
+      "loss": 0.1095,
+      "step": 9250
+    },
+    {
+      "epoch": 0.8767278924446128,
+      "grad_norm": 0.28221869468688965,
+      "learning_rate": 1.649308843022155e-05,
+      "loss": 0.1134,
+      "step": 9260
+    },
+    {
+      "epoch": 0.8776746828252225,
+      "grad_norm": 0.2599260210990906,
+      "learning_rate": 1.6489301268699113e-05,
+      "loss": 0.1154,
+      "step": 9270
+    },
+    {
+      "epoch": 0.8786214732058323,
+      "grad_norm": 0.3351801931858063,
+      "learning_rate": 1.6485514107176673e-05,
+      "loss": 0.1115,
+      "step": 9280
+    },
+    {
+      "epoch": 0.8795682635864419,
+      "grad_norm": 0.256975382566452,
+      "learning_rate": 1.6481726945654233e-05,
+      "loss": 0.1101,
+      "step": 9290
+    },
+    {
+      "epoch": 0.8805150539670517,
+      "grad_norm": 0.33674198389053345,
+      "learning_rate": 1.6477939784131796e-05,
+      "loss": 0.1193,
+      "step": 9300
+    },
+    {
+      "epoch": 0.8814618443476614,
+      "grad_norm": 0.3078038692474365,
+      "learning_rate": 1.6474152622609356e-05,
+      "loss": 0.1195,
+      "step": 9310
+    },
+    {
+      "epoch": 0.8824086347282711,
+      "grad_norm": 0.34881317615509033,
+      "learning_rate": 1.6470365461086916e-05,
+      "loss": 0.1133,
+      "step": 9320
+    },
+    {
+      "epoch": 0.8833554251088809,
+      "grad_norm": 0.32375583052635193,
+      "learning_rate": 1.6466578299564476e-05,
+      "loss": 0.1074,
+      "step": 9330
+    },
+    {
+      "epoch": 0.8843022154894906,
+      "grad_norm": 0.40294790267944336,
+      "learning_rate": 1.6462791138042037e-05,
+      "loss": 0.1231,
+      "step": 9340
+    },
+    {
+      "epoch": 0.8852490058701004,
+      "grad_norm": 0.3399127721786499,
+      "learning_rate": 1.64590039765196e-05,
+      "loss": 0.1075,
+      "step": 9350
+    },
+    {
+      "epoch": 0.8861957962507101,
+      "grad_norm": 0.4025484323501587,
+      "learning_rate": 1.645521681499716e-05,
+      "loss": 0.1187,
+      "step": 9360
+    },
+    {
+      "epoch": 0.8871425866313198,
+      "grad_norm": 0.30942675471305847,
+      "learning_rate": 1.6451429653474723e-05,
+      "loss": 0.1056,
+      "step": 9370
+    },
+    {
+      "epoch": 0.8880893770119296,
+      "grad_norm": 0.28850677609443665,
+      "learning_rate": 1.6447642491952284e-05,
+      "loss": 0.1099,
+      "step": 9380
+    },
+    {
+      "epoch": 0.8890361673925393,
+      "grad_norm": 0.4062662422657013,
+      "learning_rate": 1.6443855330429844e-05,
+      "loss": 0.1115,
+      "step": 9390
+    },
+    {
+      "epoch": 0.8899829577731491,
+      "grad_norm": 0.28927358984947205,
+      "learning_rate": 1.6440068168907407e-05,
+      "loss": 0.1236,
+      "step": 9400
+    },
+    {
+      "epoch": 0.8909297481537588,
+      "grad_norm": 0.3105138838291168,
+      "learning_rate": 1.6436281007384967e-05,
+      "loss": 0.1183,
+      "step": 9410
+    },
+    {
+      "epoch": 0.8918765385343685,
+      "grad_norm": 0.3197774589061737,
+      "learning_rate": 1.6432493845862527e-05,
+      "loss": 0.1086,
+      "step": 9420
+    },
+    {
+      "epoch": 0.8928233289149782,
+      "grad_norm": 0.32746583223342896,
+      "learning_rate": 1.6428706684340087e-05,
+      "loss": 0.1059,
+      "step": 9430
+    },
+    {
+      "epoch": 0.8937701192955879,
+      "grad_norm": 0.37387615442276,
+      "learning_rate": 1.642491952281765e-05,
+      "loss": 0.113,
+      "step": 9440
+    },
+    {
+      "epoch": 0.8947169096761977,
+      "grad_norm": 0.2911737263202667,
+      "learning_rate": 1.642113236129521e-05,
+      "loss": 0.1192,
+      "step": 9450
+    },
+    {
+      "epoch": 0.8956637000568074,
+      "grad_norm": 0.25878214836120605,
+      "learning_rate": 1.641734519977277e-05,
+      "loss": 0.112,
+      "step": 9460
+    },
+    {
+      "epoch": 0.8966104904374171,
+      "grad_norm": 0.3366353511810303,
+      "learning_rate": 1.641355803825033e-05,
+      "loss": 0.1107,
+      "step": 9470
+    },
+    {
+      "epoch": 0.8975572808180269,
+      "grad_norm": 0.3223920166492462,
+      "learning_rate": 1.6409770876727894e-05,
+      "loss": 0.117,
+      "step": 9480
+    },
+    {
+      "epoch": 0.8985040711986366,
+      "grad_norm": 0.32262280583381653,
+      "learning_rate": 1.6405983715205454e-05,
+      "loss": 0.1127,
+      "step": 9490
+    },
+    {
+      "epoch": 0.8994508615792464,
+      "grad_norm": 0.344641774892807,
+      "learning_rate": 1.6402196553683018e-05,
+      "loss": 0.1107,
+      "step": 9500
+    },
+    {
+      "epoch": 0.9003976519598561,
+      "grad_norm": 0.30609458684921265,
+      "learning_rate": 1.6398409392160578e-05,
+      "loss": 0.1189,
+      "step": 9510
+    },
+    {
+      "epoch": 0.9013444423404658,
+      "grad_norm": 0.27467280626296997,
+      "learning_rate": 1.6394622230638138e-05,
+      "loss": 0.1277,
+      "step": 9520
+    },
+    {
+      "epoch": 0.9022912327210756,
+      "grad_norm": 0.31600040197372437,
+      "learning_rate": 1.6390835069115698e-05,
+      "loss": 0.1095,
+      "step": 9530
+    },
+    {
+      "epoch": 0.9032380231016853,
+      "grad_norm": 0.2679688036441803,
+      "learning_rate": 1.638704790759326e-05,
+      "loss": 0.1054,
+      "step": 9540
+    },
+    {
+      "epoch": 0.9041848134822951,
+      "grad_norm": 0.3445126414299011,
+      "learning_rate": 1.638326074607082e-05,
+      "loss": 0.1128,
+      "step": 9550
+    },
+    {
+      "epoch": 0.9051316038629047,
+      "grad_norm": 0.3073931634426117,
+      "learning_rate": 1.637947358454838e-05,
+      "loss": 0.116,
+      "step": 9560
+    },
+    {
+      "epoch": 0.9060783942435144,
+      "grad_norm": 0.36402496695518494,
+      "learning_rate": 1.637568642302594e-05,
+      "loss": 0.1106,
+      "step": 9570
+    },
+    {
+      "epoch": 0.9070251846241242,
+      "grad_norm": 0.34014105796813965,
+      "learning_rate": 1.6371899261503505e-05,
+      "loss": 0.1166,
+      "step": 9580
+    },
+    {
+      "epoch": 0.9079719750047339,
+      "grad_norm": 0.3127152919769287,
+      "learning_rate": 1.6368112099981065e-05,
+      "loss": 0.1146,
+      "step": 9590
+    },
+    {
+      "epoch": 0.9089187653853437,
+      "grad_norm": 0.252822607755661,
+      "learning_rate": 1.636432493845863e-05,
+      "loss": 0.1127,
+      "step": 9600
+    },
+    {
+      "epoch": 0.9098655557659534,
+      "grad_norm": 0.2900446653366089,
+      "learning_rate": 1.636053777693619e-05,
+      "loss": 0.1093,
+      "step": 9610
+    },
+    {
+      "epoch": 0.9108123461465631,
+      "grad_norm": 0.2973507046699524,
+      "learning_rate": 1.635675061541375e-05,
+      "loss": 0.1063,
+      "step": 9620
+    },
+    {
+      "epoch": 0.9117591365271729,
+      "grad_norm": 0.2629620134830475,
+      "learning_rate": 1.6352963453891312e-05,
+      "loss": 0.1135,
+      "step": 9630
+    },
+    {
+      "epoch": 0.9127059269077826,
+      "grad_norm": 0.2994817793369293,
+      "learning_rate": 1.6349176292368872e-05,
+      "loss": 0.118,
+      "step": 9640
+    },
+    {
+      "epoch": 0.9136527172883924,
+      "grad_norm": 0.39161399006843567,
+      "learning_rate": 1.6345389130846432e-05,
+      "loss": 0.1134,
+      "step": 9650
+    },
+    {
+      "epoch": 0.9145995076690021,
+      "grad_norm": 0.3175990581512451,
+      "learning_rate": 1.6341601969323992e-05,
+      "loss": 0.1145,
+      "step": 9660
+    },
+    {
+      "epoch": 0.9155462980496119,
+      "grad_norm": 0.31118014454841614,
+      "learning_rate": 1.6337814807801552e-05,
+      "loss": 0.1245,
+      "step": 9670
+    },
+    {
+      "epoch": 0.9164930884302216,
+      "grad_norm": 0.29309409856796265,
+      "learning_rate": 1.6334027646279115e-05,
+      "loss": 0.1103,
+      "step": 9680
+    },
+    {
+      "epoch": 0.9174398788108313,
+      "grad_norm": 0.2814522385597229,
+      "learning_rate": 1.6330240484756676e-05,
+      "loss": 0.1055,
+      "step": 9690
+    },
+    {
+      "epoch": 0.918386669191441,
+      "grad_norm": 0.2632008492946625,
+      "learning_rate": 1.6326453323234236e-05,
+      "loss": 0.0977,
+      "step": 9700
+    },
+    {
+      "epoch": 0.9193334595720507,
+      "grad_norm": 0.3142138123512268,
+      "learning_rate": 1.63226661617118e-05,
+      "loss": 0.1192,
+      "step": 9710
+    },
+    {
+      "epoch": 0.9202802499526604,
+      "grad_norm": 0.29596778750419617,
+      "learning_rate": 1.631887900018936e-05,
+      "loss": 0.1105,
+      "step": 9720
+    },
+    {
+      "epoch": 0.9212270403332702,
+      "grad_norm": 0.31886908411979675,
+      "learning_rate": 1.6315091838666923e-05,
+      "loss": 0.1108,
+      "step": 9730
+    },
+    {
+      "epoch": 0.9221738307138799,
+      "grad_norm": 0.32292440533638,
+      "learning_rate": 1.6311304677144483e-05,
+      "loss": 0.122,
+      "step": 9740
+    },
+    {
+      "epoch": 0.9231206210944897,
+      "grad_norm": 0.34209519624710083,
+      "learning_rate": 1.6307517515622043e-05,
+      "loss": 0.1117,
+      "step": 9750
+    },
+    {
+      "epoch": 0.9240674114750994,
+      "grad_norm": 0.40027058124542236,
+      "learning_rate": 1.6303730354099603e-05,
+      "loss": 0.1177,
+      "step": 9760
+    },
+    {
+      "epoch": 0.9250142018557092,
+      "grad_norm": 0.3353174924850464,
+      "learning_rate": 1.6299943192577166e-05,
+      "loss": 0.1147,
+      "step": 9770
+    },
+    {
+      "epoch": 0.9259609922363189,
+      "grad_norm": 0.31666451692581177,
+      "learning_rate": 1.6296156031054726e-05,
+      "loss": 0.1182,
+      "step": 9780
+    },
+    {
+      "epoch": 0.9269077826169286,
+      "grad_norm": 0.33596381545066833,
+      "learning_rate": 1.6292368869532286e-05,
+      "loss": 0.126,
+      "step": 9790
+    },
+    {
+      "epoch": 0.9278545729975384,
+      "grad_norm": 0.30454882979393005,
+      "learning_rate": 1.6288581708009846e-05,
+      "loss": 0.1176,
+      "step": 9800
+    },
+    {
+      "epoch": 0.9288013633781481,
+      "grad_norm": 0.3319470286369324,
+      "learning_rate": 1.6284794546487406e-05,
+      "loss": 0.117,
+      "step": 9810
+    },
+    {
+      "epoch": 0.9297481537587579,
+      "grad_norm": 0.30625036358833313,
+      "learning_rate": 1.628100738496497e-05,
+      "loss": 0.1104,
+      "step": 9820
+    },
+    {
+      "epoch": 0.9306949441393676,
+      "grad_norm": 0.2689434885978699,
+      "learning_rate": 1.627722022344253e-05,
+      "loss": 0.1075,
+      "step": 9830
+    },
+    {
+      "epoch": 0.9316417345199772,
+      "grad_norm": 0.34801292419433594,
+      "learning_rate": 1.6273433061920093e-05,
+      "loss": 0.1101,
+      "step": 9840
+    },
+    {
+      "epoch": 0.932588524900587,
+      "grad_norm": 0.379484087228775,
+      "learning_rate": 1.6269645900397653e-05,
+      "loss": 0.1106,
+      "step": 9850
+    },
+    {
+      "epoch": 0.9335353152811967,
+      "grad_norm": 0.24420376121997833,
+      "learning_rate": 1.6265858738875217e-05,
+      "loss": 0.1034,
+      "step": 9860
+    },
+    {
+      "epoch": 0.9344821056618065,
+      "grad_norm": 0.3154924213886261,
+      "learning_rate": 1.6262071577352777e-05,
+      "loss": 0.1125,
+      "step": 9870
+    },
+    {
+      "epoch": 0.9354288960424162,
+      "grad_norm": 0.3439418375492096,
+      "learning_rate": 1.6258284415830337e-05,
+      "loss": 0.1131,
+      "step": 9880
+    },
+    {
+      "epoch": 0.9363756864230259,
+      "grad_norm": 0.3413674533367157,
+      "learning_rate": 1.6254497254307897e-05,
+      "loss": 0.1227,
+      "step": 9890
+    },
+    {
+      "epoch": 0.9373224768036357,
+      "grad_norm": 0.3342759311199188,
+      "learning_rate": 1.6250710092785457e-05,
+      "loss": 0.1107,
+      "step": 9900
+    },
+    {
+      "epoch": 0.9382692671842454,
+      "grad_norm": 0.28516536951065063,
+      "learning_rate": 1.624692293126302e-05,
+      "loss": 0.1141,
+      "step": 9910
+    },
+    {
+      "epoch": 0.9392160575648552,
+      "grad_norm": 0.33588477969169617,
+      "learning_rate": 1.624313576974058e-05,
+      "loss": 0.1111,
+      "step": 9920
+    },
+    {
+      "epoch": 0.9401628479454649,
+      "grad_norm": 0.23833739757537842,
+      "learning_rate": 1.623934860821814e-05,
+      "loss": 0.1101,
+      "step": 9930
+    },
+    {
+      "epoch": 0.9411096383260746,
+      "grad_norm": 0.2724970579147339,
+      "learning_rate": 1.6235561446695704e-05,
+      "loss": 0.1149,
+      "step": 9940
+    },
+    {
+      "epoch": 0.9420564287066844,
+      "grad_norm": 0.3115902543067932,
+      "learning_rate": 1.6231774285173264e-05,
+      "loss": 0.1131,
+      "step": 9950
+    },
+    {
+      "epoch": 0.9430032190872941,
+      "grad_norm": 0.2824494242668152,
+      "learning_rate": 1.6227987123650827e-05,
+      "loss": 0.1086,
+      "step": 9960
+    },
+    {
+      "epoch": 0.9439500094679039,
+      "grad_norm": 0.2815304398536682,
+      "learning_rate": 1.6224199962128387e-05,
+      "loss": 0.1151,
+      "step": 9970
+    },
+    {
+      "epoch": 0.9448967998485135,
+      "grad_norm": 0.24285712838172913,
+      "learning_rate": 1.6220412800605947e-05,
+      "loss": 0.1058,
+      "step": 9980
+    },
+    {
+      "epoch": 0.9458435902291232,
+      "grad_norm": 0.35544589161872864,
+      "learning_rate": 1.6216625639083508e-05,
+      "loss": 0.1113,
+      "step": 9990
+    },
+    {
+      "epoch": 0.946790380609733,
+      "grad_norm": 0.3292807638645172,
+      "learning_rate": 1.621283847756107e-05,
+      "loss": 0.13,
+      "step": 10000
+    },
+    {
+      "epoch": 0.9477371709903427,
+      "grad_norm": 0.3442451059818268,
+      "learning_rate": 1.620905131603863e-05,
+      "loss": 0.1217,
+      "step": 10010
+    },
+    {
+      "epoch": 0.9486839613709525,
+      "grad_norm": 0.2950438857078552,
+      "learning_rate": 1.620526415451619e-05,
+      "loss": 0.1069,
+      "step": 10020
+    },
+    {
+      "epoch": 0.9496307517515622,
+      "grad_norm": 0.29165175557136536,
+      "learning_rate": 1.620147699299375e-05,
+      "loss": 0.111,
+      "step": 10030
+    },
+    {
+      "epoch": 0.9505775421321719,
+      "grad_norm": 0.3380615711212158,
+      "learning_rate": 1.619768983147131e-05,
+      "loss": 0.1299,
+      "step": 10040
+    },
+    {
+      "epoch": 0.9515243325127817,
+      "grad_norm": 0.28867748379707336,
+      "learning_rate": 1.6193902669948875e-05,
+      "loss": 0.1143,
+      "step": 10050
+    },
+    {
+      "epoch": 0.9524711228933914,
+      "grad_norm": 0.3240725100040436,
+      "learning_rate": 1.6190115508426435e-05,
+      "loss": 0.1108,
+      "step": 10060
+    },
+    {
+      "epoch": 0.9534179132740012,
+      "grad_norm": 0.2736086845397949,
+      "learning_rate": 1.6186328346903998e-05,
+      "loss": 0.1127,
+      "step": 10070
+    },
+    {
+      "epoch": 0.9543647036546109,
+      "grad_norm": 0.3198991119861603,
+      "learning_rate": 1.6182541185381558e-05,
+      "loss": 0.1077,
+      "step": 10080
+    },
+    {
+      "epoch": 0.9553114940352206,
+      "grad_norm": 0.32814446091651917,
+      "learning_rate": 1.6178754023859118e-05,
+      "loss": 0.1157,
+      "step": 10090
+    },
+    {
+      "epoch": 0.9562582844158304,
+      "grad_norm": 0.3374975919723511,
+      "learning_rate": 1.617496686233668e-05,
+      "loss": 0.1134,
+      "step": 10100
+    },
+    {
+      "epoch": 0.95720507479644,
+      "grad_norm": 0.24537509679794312,
+      "learning_rate": 1.617117970081424e-05,
+      "loss": 0.1169,
+      "step": 10110
+    },
+    {
+      "epoch": 0.9581518651770498,
+      "grad_norm": 0.29897022247314453,
+      "learning_rate": 1.6167392539291802e-05,
+      "loss": 0.1153,
+      "step": 10120
+    },
+    {
+      "epoch": 0.9590986555576595,
+      "grad_norm": 0.3204127550125122,
+      "learning_rate": 1.6163605377769362e-05,
+      "loss": 0.1089,
+      "step": 10130
+    },
+    {
+      "epoch": 0.9600454459382692,
+      "grad_norm": 0.2863658666610718,
+      "learning_rate": 1.6159818216246925e-05,
+      "loss": 0.121,
+      "step": 10140
+    },
+    {
+      "epoch": 0.960992236318879,
+      "grad_norm": 0.3890479803085327,
+      "learning_rate": 1.6156031054724485e-05,
+      "loss": 0.1148,
+      "step": 10150
+    },
+    {
+      "epoch": 0.9619390266994887,
+      "grad_norm": 0.3382028341293335,
+      "learning_rate": 1.6152243893202045e-05,
+      "loss": 0.1195,
+      "step": 10160
+    },
+    {
+      "epoch": 0.9628858170800985,
+      "grad_norm": 0.27825555205345154,
+      "learning_rate": 1.6148456731679605e-05,
+      "loss": 0.1097,
+      "step": 10170
+    },
+    {
+      "epoch": 0.9638326074607082,
+      "grad_norm": 0.292385995388031,
+      "learning_rate": 1.614466957015717e-05,
+      "loss": 0.1158,
+      "step": 10180
+    },
+    {
+      "epoch": 0.9647793978413179,
+      "grad_norm": 0.3470708429813385,
+      "learning_rate": 1.614088240863473e-05,
+      "loss": 0.1291,
+      "step": 10190
+    },
+    {
+      "epoch": 0.9657261882219277,
+      "grad_norm": 0.34543946385383606,
+      "learning_rate": 1.6137095247112292e-05,
+      "loss": 0.122,
+      "step": 10200
+    },
+    {
+      "epoch": 0.9666729786025374,
+      "grad_norm": 0.2738993465900421,
+      "learning_rate": 1.6133308085589852e-05,
+      "loss": 0.1215,
+      "step": 10210
+    },
+    {
+      "epoch": 0.9676197689831472,
+      "grad_norm": 0.28177791833877563,
+      "learning_rate": 1.6129520924067412e-05,
+      "loss": 0.1255,
+      "step": 10220
+    },
+    {
+      "epoch": 0.9685665593637569,
+      "grad_norm": 0.2852981984615326,
+      "learning_rate": 1.6125733762544972e-05,
+      "loss": 0.1039,
+      "step": 10230
+    },
+    {
+      "epoch": 0.9695133497443666,
+      "grad_norm": 0.3322448432445526,
+      "learning_rate": 1.6121946601022536e-05,
+      "loss": 0.1114,
+      "step": 10240
+    },
+    {
+      "epoch": 0.9704601401249763,
+      "grad_norm": 0.3602607548236847,
+      "learning_rate": 1.6118159439500096e-05,
+      "loss": 0.1167,
+      "step": 10250
+    },
+    {
+      "epoch": 0.971406930505586,
+      "grad_norm": 0.33711227774620056,
+      "learning_rate": 1.6114372277977656e-05,
+      "loss": 0.1051,
+      "step": 10260
+    },
+    {
+      "epoch": 0.9723537208861958,
+      "grad_norm": 0.39165088534355164,
+      "learning_rate": 1.6110585116455216e-05,
+      "loss": 0.1157,
+      "step": 10270
+    },
+    {
+      "epoch": 0.9733005112668055,
+      "grad_norm": 0.32423534989356995,
+      "learning_rate": 1.610679795493278e-05,
+      "loss": 0.1048,
+      "step": 10280
+    },
+    {
+      "epoch": 0.9742473016474152,
+      "grad_norm": 0.31623855233192444,
+      "learning_rate": 1.610301079341034e-05,
+      "loss": 0.1123,
+      "step": 10290
+    },
+    {
+      "epoch": 0.975194092028025,
+      "grad_norm": 0.3637257516384125,
+      "learning_rate": 1.6099223631887903e-05,
+      "loss": 0.1087,
+      "step": 10300
+    },
+    {
+      "epoch": 0.9761408824086347,
+      "grad_norm": 0.5228608846664429,
+      "learning_rate": 1.6095436470365463e-05,
+      "loss": 0.1167,
+      "step": 10310
+    },
+    {
+      "epoch": 0.9770876727892445,
+      "grad_norm": 0.3268279731273651,
+      "learning_rate": 1.6091649308843023e-05,
+      "loss": 0.1231,
+      "step": 10320
+    },
+    {
+      "epoch": 0.9780344631698542,
+      "grad_norm": 0.2953413426876068,
+      "learning_rate": 1.6087862147320586e-05,
+      "loss": 0.1084,
+      "step": 10330
+    },
+    {
+      "epoch": 0.978981253550464,
+      "grad_norm": 0.26272329688072205,
+      "learning_rate": 1.6084074985798147e-05,
+      "loss": 0.109,
+      "step": 10340
+    },
+    {
+      "epoch": 0.9799280439310737,
+      "grad_norm": 0.28040027618408203,
+      "learning_rate": 1.6080287824275707e-05,
+      "loss": 0.1086,
+      "step": 10350
+    },
+    {
+      "epoch": 0.9808748343116834,
+      "grad_norm": 0.28157317638397217,
+      "learning_rate": 1.6076500662753267e-05,
+      "loss": 0.1056,
+      "step": 10360
+    },
+    {
+      "epoch": 0.9818216246922932,
+      "grad_norm": 0.2670665979385376,
+      "learning_rate": 1.6072713501230827e-05,
+      "loss": 0.116,
+      "step": 10370
+    },
+    {
+      "epoch": 0.9827684150729029,
+      "grad_norm": 0.36410754919052124,
+      "learning_rate": 1.606892633970839e-05,
+      "loss": 0.1142,
+      "step": 10380
+    },
+    {
+      "epoch": 0.9837152054535125,
+      "grad_norm": 0.32198601961135864,
+      "learning_rate": 1.606513917818595e-05,
+      "loss": 0.1098,
+      "step": 10390
+    },
+    {
+      "epoch": 0.9846619958341223,
+      "grad_norm": 0.3849109709262848,
+      "learning_rate": 1.606135201666351e-05,
+      "loss": 0.1116,
+      "step": 10400
+    },
+    {
+      "epoch": 0.985608786214732,
+      "grad_norm": 0.344265878200531,
+      "learning_rate": 1.6057564855141074e-05,
+      "loss": 0.1124,
+      "step": 10410
+    },
+    {
+      "epoch": 0.9865555765953418,
+      "grad_norm": 0.33011743426322937,
+      "learning_rate": 1.6053777693618634e-05,
+      "loss": 0.1032,
+      "step": 10420
+    },
+    {
+      "epoch": 0.9875023669759515,
+      "grad_norm": 0.4829758405685425,
+      "learning_rate": 1.6049990532096197e-05,
+      "loss": 0.1041,
+      "step": 10430
+    },
+    {
+      "epoch": 0.9884491573565612,
+      "grad_norm": 0.3808184862136841,
+      "learning_rate": 1.6046203370573757e-05,
+      "loss": 0.1116,
+      "step": 10440
+    },
+    {
+      "epoch": 0.989395947737171,
+      "grad_norm": 0.31177645921707153,
+      "learning_rate": 1.6042416209051317e-05,
+      "loss": 0.1195,
+      "step": 10450
+    },
+    {
+      "epoch": 0.9903427381177807,
+      "grad_norm": 0.30874019861221313,
+      "learning_rate": 1.6038629047528877e-05,
+      "loss": 0.1122,
+      "step": 10460
+    },
+    {
+      "epoch": 0.9912895284983905,
+      "grad_norm": 0.286725789308548,
+      "learning_rate": 1.603484188600644e-05,
+      "loss": 0.1127,
+      "step": 10470
+    },
+    {
+      "epoch": 0.9922363188790002,
+      "grad_norm": 0.30968090891838074,
+      "learning_rate": 1.6031054724484e-05,
+      "loss": 0.1119,
+      "step": 10480
+    },
+    {
+      "epoch": 0.99318310925961,
+      "grad_norm": 0.29032912850379944,
+      "learning_rate": 1.602726756296156e-05,
+      "loss": 0.1139,
+      "step": 10490
+    },
+    {
+      "epoch": 0.9941298996402197,
+      "grad_norm": 0.3562324643135071,
+      "learning_rate": 1.602348040143912e-05,
+      "loss": 0.1163,
+      "step": 10500
+    },
+    {
+      "epoch": 0.9950766900208294,
+      "grad_norm": 0.28674376010894775,
+      "learning_rate": 1.6019693239916684e-05,
+      "loss": 0.1055,
+      "step": 10510
+    },
+    {
+      "epoch": 0.9960234804014392,
+      "grad_norm": 0.2935590147972107,
+      "learning_rate": 1.6015906078394244e-05,
+      "loss": 0.1043,
+      "step": 10520
+    },
+    {
+      "epoch": 0.9969702707820488,
+      "grad_norm": 0.3492721617221832,
+      "learning_rate": 1.6012118916871804e-05,
+      "loss": 0.1146,
+      "step": 10530
+    },
+    {
+      "epoch": 0.9979170611626585,
+      "grad_norm": 0.30205583572387695,
+      "learning_rate": 1.6008331755349368e-05,
+      "loss": 0.107,
+      "step": 10540
+    },
+    {
+      "epoch": 0.9988638515432683,
+      "grad_norm": 0.3231695294380188,
+      "learning_rate": 1.6004544593826928e-05,
+      "loss": 0.1152,
+      "step": 10550
+    },
+    {
+      "epoch": 0.999810641923878,
+      "grad_norm": 0.31855645775794983,
+      "learning_rate": 1.600075743230449e-05,
+      "loss": 0.1111,
+      "step": 10560
+    },
+    {
+      "epoch": 1.0,
+      "eval_f1_micro": 0.3363452934467936,
+      "eval_loss": 0.11279743909835815,
+      "eval_precision": 0.6191527157237444,
+      "eval_recall": 0.23088503924363168,
+      "eval_runtime": 345.408,
+      "eval_samples_per_second": 122.305,
+      "eval_steps_per_second": 7.646,
+      "step": 10562
+    },
+    {
+      "epoch": 1.0007574323044879,
+      "grad_norm": 0.3501763641834259,
+      "learning_rate": 1.599697027078205e-05,
+      "loss": 0.1054,
+      "step": 10570
+    },
+    {
+      "epoch": 1.0017042226850976,
+      "grad_norm": 0.2885648012161255,
+      "learning_rate": 1.599318310925961e-05,
+      "loss": 0.1037,
+      "step": 10580
+    },
+    {
+      "epoch": 1.0026510130657071,
+      "grad_norm": 0.24593491852283478,
+      "learning_rate": 1.598939594773717e-05,
+      "loss": 0.0999,
+      "step": 10590
+    },
+    {
+      "epoch": 1.0035978034463169,
+      "grad_norm": 0.2742719054222107,
+      "learning_rate": 1.598560878621473e-05,
+      "loss": 0.1129,
+      "step": 10600
+    },
+    {
+      "epoch": 1.0045445938269266,
+      "grad_norm": 0.221815824508667,
+      "learning_rate": 1.5981821624692295e-05,
+      "loss": 0.1023,
+      "step": 10610
+    },
+    {
+      "epoch": 1.0054913842075364,
+      "grad_norm": 0.2318876087665558,
+      "learning_rate": 1.5978034463169855e-05,
+      "loss": 0.1098,
+      "step": 10620
+    },
+    {
+      "epoch": 1.006438174588146,
+      "grad_norm": 0.25267088413238525,
+      "learning_rate": 1.5974247301647415e-05,
+      "loss": 0.1091,
+      "step": 10630
+    },
+    {
+      "epoch": 1.0073849649687558,
+      "grad_norm": 0.3592093884944916,
+      "learning_rate": 1.597046014012498e-05,
+      "loss": 0.1121,
+      "step": 10640
+    },
+    {
+      "epoch": 1.0083317553493656,
+      "grad_norm": 0.2729831337928772,
+      "learning_rate": 1.596667297860254e-05,
+      "loss": 0.1002,
+      "step": 10650
+    },
+    {
+      "epoch": 1.0092785457299753,
+      "grad_norm": 0.24247096478939056,
+      "learning_rate": 1.5962885817080102e-05,
+      "loss": 0.1015,
+      "step": 10660
+    },
+    {
+      "epoch": 1.010225336110585,
+      "grad_norm": 0.2638133466243744,
+      "learning_rate": 1.5959098655557662e-05,
+      "loss": 0.1011,
+      "step": 10670
+    },
+    {
+      "epoch": 1.0111721264911948,
+      "grad_norm": 0.377765029668808,
+      "learning_rate": 1.5955311494035222e-05,
+      "loss": 0.1101,
+      "step": 10680
+    },
+    {
+      "epoch": 1.0121189168718046,
+      "grad_norm": 0.3203154504299164,
+      "learning_rate": 1.5951524332512782e-05,
+      "loss": 0.1197,
+      "step": 10690
+    },
+    {
+      "epoch": 1.0130657072524143,
+      "grad_norm": 0.35998988151550293,
+      "learning_rate": 1.5947737170990346e-05,
+      "loss": 0.1005,
+      "step": 10700
+    },
+    {
+      "epoch": 1.014012497633024,
+      "grad_norm": 0.28926974534988403,
+      "learning_rate": 1.5943950009467906e-05,
+      "loss": 0.1066,
+      "step": 10710
+    },
+    {
+      "epoch": 1.0149592880136338,
+      "grad_norm": 0.3111324608325958,
+      "learning_rate": 1.5940162847945466e-05,
+      "loss": 0.1067,
+      "step": 10720
+    },
+    {
+      "epoch": 1.0159060783942435,
+      "grad_norm": 0.27141204476356506,
+      "learning_rate": 1.5936375686423026e-05,
+      "loss": 0.1054,
+      "step": 10730
+    },
+    {
+      "epoch": 1.0168528687748533,
+      "grad_norm": 0.3853040039539337,
+      "learning_rate": 1.5932588524900586e-05,
+      "loss": 0.113,
+      "step": 10740
+    },
+    {
+      "epoch": 1.017799659155463,
+      "grad_norm": 0.3320479094982147,
+      "learning_rate": 1.592880136337815e-05,
+      "loss": 0.1027,
+      "step": 10750
+    },
+    {
+      "epoch": 1.0187464495360727,
+      "grad_norm": 0.38209497928619385,
+      "learning_rate": 1.592501420185571e-05,
+      "loss": 0.1289,
+      "step": 10760
+    },
+    {
+      "epoch": 1.0196932399166825,
+      "grad_norm": 0.42024433612823486,
+      "learning_rate": 1.5921227040333273e-05,
+      "loss": 0.1168,
+      "step": 10770
+    },
+    {
+      "epoch": 1.0206400302972922,
+      "grad_norm": 0.2727624177932739,
+      "learning_rate": 1.5917439878810833e-05,
+      "loss": 0.1107,
+      "step": 10780
+    },
+    {
+      "epoch": 1.021586820677902,
+      "grad_norm": 0.39503511786460876,
+      "learning_rate": 1.5913652717288393e-05,
+      "loss": 0.1126,
+      "step": 10790
+    },
+    {
+      "epoch": 1.0225336110585117,
+      "grad_norm": 0.27807143330574036,
+      "learning_rate": 1.5909865555765956e-05,
+      "loss": 0.1132,
+      "step": 10800
+    },
+    {
+      "epoch": 1.0234804014391214,
+      "grad_norm": 0.3399743139743805,
+      "learning_rate": 1.5906078394243516e-05,
+      "loss": 0.1136,
+      "step": 10810
+    },
+    {
+      "epoch": 1.0244271918197312,
+      "grad_norm": 0.3346470296382904,
+      "learning_rate": 1.5902291232721076e-05,
+      "loss": 0.1144,
+      "step": 10820
+    },
+    {
+      "epoch": 1.025373982200341,
+      "grad_norm": 0.33610638976097107,
+      "learning_rate": 1.5898504071198636e-05,
+      "loss": 0.1088,
+      "step": 10830
+    },
+    {
+      "epoch": 1.0263207725809507,
+      "grad_norm": 0.33162784576416016,
+      "learning_rate": 1.58947169096762e-05,
+      "loss": 0.1026,
+      "step": 10840
+    },
+    {
+      "epoch": 1.0272675629615604,
+      "grad_norm": 0.3138902485370636,
+      "learning_rate": 1.589092974815376e-05,
+      "loss": 0.1161,
+      "step": 10850
+    },
+    {
+      "epoch": 1.0282143533421702,
+      "grad_norm": 0.3033609688282013,
+      "learning_rate": 1.588714258663132e-05,
+      "loss": 0.0992,
+      "step": 10860
+    },
+    {
+      "epoch": 1.0291611437227797,
+      "grad_norm": 0.3374888300895691,
+      "learning_rate": 1.5883355425108883e-05,
+      "loss": 0.122,
+      "step": 10870
+    },
+    {
+      "epoch": 1.0301079341033894,
+      "grad_norm": 0.294697642326355,
+      "learning_rate": 1.5879568263586443e-05,
+      "loss": 0.1126,
+      "step": 10880
+    },
+    {
+      "epoch": 1.0310547244839992,
+      "grad_norm": 0.3098808526992798,
+      "learning_rate": 1.5875781102064003e-05,
+      "loss": 0.1078,
+      "step": 10890
+    },
+    {
+      "epoch": 1.032001514864609,
+      "grad_norm": 0.23213276267051697,
+      "learning_rate": 1.5871993940541567e-05,
+      "loss": 0.1113,
+      "step": 10900
+    },
+    {
+      "epoch": 1.0329483052452186,
+      "grad_norm": 0.30921486020088196,
+      "learning_rate": 1.5868206779019127e-05,
+      "loss": 0.118,
+      "step": 10910
+    },
+    {
+      "epoch": 1.0338950956258284,
+      "grad_norm": 0.29103872179985046,
+      "learning_rate": 1.5864419617496687e-05,
+      "loss": 0.1108,
+      "step": 10920
+    },
+    {
+      "epoch": 1.0348418860064381,
+      "grad_norm": 0.34007346630096436,
+      "learning_rate": 1.5860632455974247e-05,
+      "loss": 0.1153,
+      "step": 10930
+    },
+    {
+      "epoch": 1.0357886763870479,
+      "grad_norm": 0.3389047682285309,
+      "learning_rate": 1.585684529445181e-05,
+      "loss": 0.1126,
+      "step": 10940
+    },
+    {
+      "epoch": 1.0367354667676576,
+      "grad_norm": 0.29568174481391907,
+      "learning_rate": 1.585305813292937e-05,
+      "loss": 0.1059,
+      "step": 10950
+    },
+    {
+      "epoch": 1.0376822571482673,
+      "grad_norm": 0.38442787528038025,
+      "learning_rate": 1.584927097140693e-05,
+      "loss": 0.1051,
+      "step": 10960
+    },
+    {
+      "epoch": 1.038629047528877,
+      "grad_norm": 0.3523216247558594,
+      "learning_rate": 1.584548380988449e-05,
+      "loss": 0.1061,
+      "step": 10970
+    },
+    {
+      "epoch": 1.0395758379094868,
+      "grad_norm": 0.3302572965621948,
+      "learning_rate": 1.5841696648362054e-05,
+      "loss": 0.1105,
+      "step": 10980
+    },
+    {
+      "epoch": 1.0405226282900966,
+      "grad_norm": 0.32812613248825073,
+      "learning_rate": 1.5837909486839614e-05,
+      "loss": 0.1081,
+      "step": 10990
+    },
+    {
+      "epoch": 1.0414694186707063,
+      "grad_norm": 0.33464688062667847,
+      "learning_rate": 1.5834122325317178e-05,
+      "loss": 0.1139,
+      "step": 11000
+    },
+    {
+      "epoch": 1.042416209051316,
+      "grad_norm": 0.29011818766593933,
+      "learning_rate": 1.5830335163794738e-05,
+      "loss": 0.1016,
+      "step": 11010
+    },
+    {
+      "epoch": 1.0433629994319258,
+      "grad_norm": 0.26706334948539734,
+      "learning_rate": 1.5826548002272298e-05,
+      "loss": 0.1128,
+      "step": 11020
+    },
+    {
+      "epoch": 1.0443097898125355,
+      "grad_norm": 0.3417057991027832,
+      "learning_rate": 1.582276084074986e-05,
+      "loss": 0.1046,
+      "step": 11030
+    },
+    {
+      "epoch": 1.0452565801931453,
+      "grad_norm": 0.4278360903263092,
+      "learning_rate": 1.581897367922742e-05,
+      "loss": 0.1078,
+      "step": 11040
+    },
+    {
+      "epoch": 1.046203370573755,
+      "grad_norm": 0.34714770317077637,
+      "learning_rate": 1.581518651770498e-05,
+      "loss": 0.1127,
+      "step": 11050
+    },
+    {
+      "epoch": 1.0471501609543648,
+      "grad_norm": 0.2509910464286804,
+      "learning_rate": 1.581139935618254e-05,
+      "loss": 0.1104,
+      "step": 11060
+    },
+    {
+      "epoch": 1.0480969513349745,
+      "grad_norm": 0.27627503871917725,
+      "learning_rate": 1.58076121946601e-05,
+      "loss": 0.1139,
+      "step": 11070
+    },
+    {
+      "epoch": 1.0490437417155842,
+      "grad_norm": 0.3637837767601013,
+      "learning_rate": 1.5803825033137665e-05,
+      "loss": 0.1117,
+      "step": 11080
+    },
+    {
+      "epoch": 1.049990532096194,
+      "grad_norm": 0.22421805560588837,
+      "learning_rate": 1.5800037871615225e-05,
+      "loss": 0.0998,
+      "step": 11090
+    },
+    {
+      "epoch": 1.0509373224768037,
+      "grad_norm": 0.5345719456672668,
+      "learning_rate": 1.5796250710092785e-05,
+      "loss": 0.1088,
+      "step": 11100
+    },
+    {
+      "epoch": 1.0518841128574135,
+      "grad_norm": 0.29915082454681396,
+      "learning_rate": 1.5792463548570348e-05,
+      "loss": 0.1158,
+      "step": 11110
+    },
+    {
+      "epoch": 1.0528309032380232,
+      "grad_norm": 0.2868350148200989,
+      "learning_rate": 1.578867638704791e-05,
+      "loss": 0.1053,
+      "step": 11120
+    },
+    {
+      "epoch": 1.053777693618633,
+      "grad_norm": 0.3004460632801056,
+      "learning_rate": 1.5784889225525472e-05,
+      "loss": 0.1086,
+      "step": 11130
+    },
+    {
+      "epoch": 1.0547244839992427,
+      "grad_norm": 0.37865251302719116,
+      "learning_rate": 1.5781102064003032e-05,
+      "loss": 0.1094,
+      "step": 11140
+    },
+    {
+      "epoch": 1.0556712743798522,
+      "grad_norm": 0.37065449357032776,
+      "learning_rate": 1.5777314902480592e-05,
+      "loss": 0.1105,
+      "step": 11150
+    },
+    {
+      "epoch": 1.056618064760462,
+      "grad_norm": 0.3341347575187683,
+      "learning_rate": 1.5773527740958152e-05,
+      "loss": 0.1,
+      "step": 11160
+    },
+    {
+      "epoch": 1.0575648551410717,
+      "grad_norm": 0.34143874049186707,
+      "learning_rate": 1.5769740579435715e-05,
+      "loss": 0.104,
+      "step": 11170
+    },
+    {
+      "epoch": 1.0585116455216814,
+      "grad_norm": 0.39380621910095215,
+      "learning_rate": 1.5765953417913275e-05,
+      "loss": 0.1102,
+      "step": 11180
+    },
+    {
+      "epoch": 1.0594584359022912,
+      "grad_norm": 0.2955397367477417,
+      "learning_rate": 1.5762166256390835e-05,
+      "loss": 0.1201,
+      "step": 11190
+    },
+    {
+      "epoch": 1.060405226282901,
+      "grad_norm": 0.33224380016326904,
+      "learning_rate": 1.5758379094868395e-05,
+      "loss": 0.1158,
+      "step": 11200
+    },
+    {
+      "epoch": 1.0613520166635106,
+      "grad_norm": 0.3783450722694397,
+      "learning_rate": 1.575459193334596e-05,
+      "loss": 0.1129,
+      "step": 11210
+    },
+    {
+      "epoch": 1.0622988070441204,
+      "grad_norm": 0.3262997567653656,
+      "learning_rate": 1.575080477182352e-05,
+      "loss": 0.1029,
+      "step": 11220
+    },
+    {
+      "epoch": 1.0632455974247301,
+      "grad_norm": 0.38155102729797363,
+      "learning_rate": 1.5747017610301082e-05,
+      "loss": 0.1021,
+      "step": 11230
+    },
+    {
+      "epoch": 1.0641923878053399,
+      "grad_norm": 0.3473179042339325,
+      "learning_rate": 1.5743230448778642e-05,
+      "loss": 0.1022,
+      "step": 11240
+    },
+    {
+      "epoch": 1.0651391781859496,
+      "grad_norm": 0.3675661087036133,
+      "learning_rate": 1.5739443287256203e-05,
+      "loss": 0.1015,
+      "step": 11250
+    },
+    {
+      "epoch": 1.0660859685665593,
+      "grad_norm": 0.3818339705467224,
+      "learning_rate": 1.5735656125733766e-05,
+      "loss": 0.1092,
+      "step": 11260
+    },
+    {
+      "epoch": 1.067032758947169,
+      "grad_norm": 0.321243554353714,
+      "learning_rate": 1.5731868964211326e-05,
+      "loss": 0.1006,
+      "step": 11270
+    },
+    {
+      "epoch": 1.0679795493277788,
+      "grad_norm": 0.42666223645210266,
+      "learning_rate": 1.5728081802688886e-05,
+      "loss": 0.1087,
+      "step": 11280
+    },
+    {
+      "epoch": 1.0689263397083886,
+      "grad_norm": 0.42080941796302795,
+      "learning_rate": 1.5724294641166446e-05,
+      "loss": 0.1068,
+      "step": 11290
+    },
+    {
+      "epoch": 1.0698731300889983,
+      "grad_norm": 0.3022630512714386,
+      "learning_rate": 1.5720507479644006e-05,
+      "loss": 0.104,
+      "step": 11300
+    },
+    {
+      "epoch": 1.070819920469608,
+      "grad_norm": 0.45780861377716064,
+      "learning_rate": 1.571672031812157e-05,
+      "loss": 0.0998,
+      "step": 11310
+    },
+    {
+      "epoch": 1.0717667108502178,
+      "grad_norm": 0.33300885558128357,
+      "learning_rate": 1.571293315659913e-05,
+      "loss": 0.1107,
+      "step": 11320
+    },
+    {
+      "epoch": 1.0727135012308275,
+      "grad_norm": 0.3255704641342163,
+      "learning_rate": 1.570914599507669e-05,
+      "loss": 0.1127,
+      "step": 11330
+    },
+    {
+      "epoch": 1.0736602916114373,
+      "grad_norm": 0.3382815420627594,
+      "learning_rate": 1.5705358833554253e-05,
+      "loss": 0.1043,
+      "step": 11340
+    },
+    {
+      "epoch": 1.074607081992047,
+      "grad_norm": 0.27806100249290466,
+      "learning_rate": 1.5701571672031813e-05,
+      "loss": 0.1143,
+      "step": 11350
+    },
+    {
+      "epoch": 1.0755538723726568,
+      "grad_norm": 0.27662235498428345,
+      "learning_rate": 1.5697784510509377e-05,
+      "loss": 0.0919,
+      "step": 11360
+    },
+    {
+      "epoch": 1.0765006627532665,
+      "grad_norm": 0.32525408267974854,
+      "learning_rate": 1.5693997348986937e-05,
+      "loss": 0.1128,
+      "step": 11370
+    },
+    {
+      "epoch": 1.0774474531338762,
+      "grad_norm": 0.2928694188594818,
+      "learning_rate": 1.5690210187464497e-05,
+      "loss": 0.1114,
+      "step": 11380
+    },
+    {
+      "epoch": 1.078394243514486,
+      "grad_norm": 0.32543128728866577,
+      "learning_rate": 1.5686423025942057e-05,
+      "loss": 0.1113,
+      "step": 11390
+    },
+    {
+      "epoch": 1.0793410338950957,
+      "grad_norm": 0.3159891366958618,
+      "learning_rate": 1.568263586441962e-05,
+      "loss": 0.1111,
+      "step": 11400
+    },
+    {
+      "epoch": 1.0802878242757052,
+      "grad_norm": 0.6057949066162109,
+      "learning_rate": 1.567884870289718e-05,
+      "loss": 0.1102,
+      "step": 11410
+    },
+    {
+      "epoch": 1.081234614656315,
+      "grad_norm": 0.43823304772377014,
+      "learning_rate": 1.567506154137474e-05,
+      "loss": 0.1154,
+      "step": 11420
+    },
+    {
+      "epoch": 1.0821814050369247,
+      "grad_norm": 0.37226566672325134,
+      "learning_rate": 1.56712743798523e-05,
+      "loss": 0.1087,
+      "step": 11430
+    },
+    {
+      "epoch": 1.0831281954175345,
+      "grad_norm": 0.41247767210006714,
+      "learning_rate": 1.566748721832986e-05,
+      "loss": 0.1091,
+      "step": 11440
+    },
+    {
+      "epoch": 1.0840749857981442,
+      "grad_norm": 0.3546197712421417,
+      "learning_rate": 1.5663700056807424e-05,
+      "loss": 0.0982,
+      "step": 11450
+    },
+    {
+      "epoch": 1.085021776178754,
+      "grad_norm": 0.32991188764572144,
+      "learning_rate": 1.5659912895284984e-05,
+      "loss": 0.1032,
+      "step": 11460
+    },
+    {
+      "epoch": 1.0859685665593637,
+      "grad_norm": 0.3210223317146301,
+      "learning_rate": 1.5656125733762547e-05,
+      "loss": 0.1192,
+      "step": 11470
+    },
+    {
+      "epoch": 1.0869153569399734,
+      "grad_norm": 0.2751235067844391,
+      "learning_rate": 1.5652338572240107e-05,
+      "loss": 0.1184,
+      "step": 11480
+    },
+    {
+      "epoch": 1.0878621473205832,
+      "grad_norm": 0.45019084215164185,
+      "learning_rate": 1.564855141071767e-05,
+      "loss": 0.114,
+      "step": 11490
+    },
+    {
+      "epoch": 1.088808937701193,
+      "grad_norm": 0.3328302502632141,
+      "learning_rate": 1.564476424919523e-05,
+      "loss": 0.111,
+      "step": 11500
+    },
+    {
+      "epoch": 1.0897557280818027,
+      "grad_norm": 0.3211038410663605,
+      "learning_rate": 1.564097708767279e-05,
+      "loss": 0.1133,
+      "step": 11510
+    },
+    {
+      "epoch": 1.0907025184624124,
+      "grad_norm": 0.3597552180290222,
+      "learning_rate": 1.563718992615035e-05,
+      "loss": 0.1108,
+      "step": 11520
+    },
+    {
+      "epoch": 1.0916493088430221,
+      "grad_norm": 0.3381154239177704,
+      "learning_rate": 1.563340276462791e-05,
+      "loss": 0.1077,
+      "step": 11530
+    },
+    {
+      "epoch": 1.0925960992236319,
+      "grad_norm": 0.3959258198738098,
+      "learning_rate": 1.5629615603105474e-05,
+      "loss": 0.1208,
+      "step": 11540
+    },
+    {
+      "epoch": 1.0935428896042416,
+      "grad_norm": 0.3280211091041565,
+      "learning_rate": 1.5625828441583034e-05,
+      "loss": 0.1039,
+      "step": 11550
+    },
+    {
+      "epoch": 1.0944896799848514,
+      "grad_norm": 0.27923819422721863,
+      "learning_rate": 1.5622041280060595e-05,
+      "loss": 0.1066,
+      "step": 11560
+    },
+    {
+      "epoch": 1.095436470365461,
+      "grad_norm": 0.29647937417030334,
+      "learning_rate": 1.5618254118538158e-05,
+      "loss": 0.1001,
+      "step": 11570
+    },
+    {
+      "epoch": 1.0963832607460708,
+      "grad_norm": 0.41365155577659607,
+      "learning_rate": 1.5614466957015718e-05,
+      "loss": 0.1123,
+      "step": 11580
+    },
+    {
+      "epoch": 1.0973300511266806,
+      "grad_norm": 0.29978007078170776,
+      "learning_rate": 1.561067979549328e-05,
+      "loss": 0.1168,
+      "step": 11590
+    },
+    {
+      "epoch": 1.0982768415072903,
+      "grad_norm": 0.3278815746307373,
+      "learning_rate": 1.560689263397084e-05,
+      "loss": 0.1126,
+      "step": 11600
+    },
+    {
+      "epoch": 1.0992236318879,
+      "grad_norm": 0.34750813245773315,
+      "learning_rate": 1.56031054724484e-05,
+      "loss": 0.1062,
+      "step": 11610
+    },
+    {
+      "epoch": 1.1001704222685098,
+      "grad_norm": 0.5691556930541992,
+      "learning_rate": 1.559931831092596e-05,
+      "loss": 0.0999,
+      "step": 11620
+    },
+    {
+      "epoch": 1.1011172126491195,
+      "grad_norm": 0.43018022179603577,
+      "learning_rate": 1.5595531149403525e-05,
+      "loss": 0.1154,
+      "step": 11630
+    },
+    {
+      "epoch": 1.1020640030297293,
+      "grad_norm": 0.3963812589645386,
+      "learning_rate": 1.5591743987881085e-05,
+      "loss": 0.1001,
+      "step": 11640
+    },
+    {
+      "epoch": 1.103010793410339,
+      "grad_norm": 0.2910940945148468,
+      "learning_rate": 1.5587956826358645e-05,
+      "loss": 0.1131,
+      "step": 11650
+    },
+    {
+      "epoch": 1.1039575837909488,
+      "grad_norm": 0.42120805382728577,
+      "learning_rate": 1.5584169664836205e-05,
+      "loss": 0.1139,
+      "step": 11660
+    },
+    {
+      "epoch": 1.1049043741715585,
+      "grad_norm": 0.33243581652641296,
+      "learning_rate": 1.5580382503313765e-05,
+      "loss": 0.1058,
+      "step": 11670
+    },
+    {
+      "epoch": 1.1058511645521683,
+      "grad_norm": 0.360920250415802,
+      "learning_rate": 1.557659534179133e-05,
+      "loss": 0.1033,
+      "step": 11680
+    },
+    {
+      "epoch": 1.106797954932778,
+      "grad_norm": 0.3652898073196411,
+      "learning_rate": 1.557280818026889e-05,
+      "loss": 0.1107,
+      "step": 11690
+    },
+    {
+      "epoch": 1.1077447453133875,
+      "grad_norm": 0.3319729268550873,
+      "learning_rate": 1.5569021018746452e-05,
+      "loss": 0.1037,
+      "step": 11700
+    },
+    {
+      "epoch": 1.1086915356939973,
+      "grad_norm": 0.4759555757045746,
+      "learning_rate": 1.5565233857224012e-05,
+      "loss": 0.1157,
+      "step": 11710
+    },
+    {
+      "epoch": 1.109638326074607,
+      "grad_norm": 0.30491140484809875,
+      "learning_rate": 1.5561446695701572e-05,
+      "loss": 0.0995,
+      "step": 11720
+    },
+    {
+      "epoch": 1.1105851164552167,
+      "grad_norm": 0.3831186592578888,
+      "learning_rate": 1.5557659534179136e-05,
+      "loss": 0.1128,
+      "step": 11730
+    },
+    {
+      "epoch": 1.1115319068358265,
+      "grad_norm": 0.3806062936782837,
+      "learning_rate": 1.5553872372656696e-05,
+      "loss": 0.0988,
+      "step": 11740
+    },
+    {
+      "epoch": 1.1124786972164362,
+      "grad_norm": 0.3348376452922821,
+      "learning_rate": 1.5550085211134256e-05,
+      "loss": 0.1121,
+      "step": 11750
+    },
+    {
+      "epoch": 1.113425487597046,
+      "grad_norm": 0.3526676297187805,
+      "learning_rate": 1.5546298049611816e-05,
+      "loss": 0.1134,
+      "step": 11760
+    },
+    {
+      "epoch": 1.1143722779776557,
+      "grad_norm": 0.4545170068740845,
+      "learning_rate": 1.554251088808938e-05,
+      "loss": 0.1104,
+      "step": 11770
+    },
+    {
+      "epoch": 1.1153190683582654,
+      "grad_norm": 0.2890951335430145,
+      "learning_rate": 1.553872372656694e-05,
+      "loss": 0.1009,
+      "step": 11780
+    },
+    {
+      "epoch": 1.1162658587388752,
+      "grad_norm": 0.3283509612083435,
+      "learning_rate": 1.55349365650445e-05,
+      "loss": 0.097,
+      "step": 11790
+    },
+    {
+      "epoch": 1.117212649119485,
+      "grad_norm": 0.39825356006622314,
+      "learning_rate": 1.553114940352206e-05,
+      "loss": 0.108,
+      "step": 11800
+    },
+    {
+      "epoch": 1.1181594395000947,
+      "grad_norm": 0.3061431348323822,
+      "learning_rate": 1.5527362241999623e-05,
+      "loss": 0.1115,
+      "step": 11810
+    },
+    {
+      "epoch": 1.1191062298807044,
+      "grad_norm": 0.3241097629070282,
+      "learning_rate": 1.5523575080477183e-05,
+      "loss": 0.104,
+      "step": 11820
+    },
+    {
+      "epoch": 1.1200530202613141,
+      "grad_norm": 0.4179798662662506,
+      "learning_rate": 1.5519787918954746e-05,
+      "loss": 0.1111,
+      "step": 11830
+    },
+    {
+      "epoch": 1.1209998106419239,
+      "grad_norm": 0.52249675989151,
+      "learning_rate": 1.5516000757432306e-05,
+      "loss": 0.1128,
+      "step": 11840
+    },
+    {
+      "epoch": 1.1219466010225336,
+      "grad_norm": 0.42467227578163147,
+      "learning_rate": 1.5512213595909866e-05,
+      "loss": 0.1127,
+      "step": 11850
+    },
+    {
+      "epoch": 1.1228933914031434,
+      "grad_norm": 0.344828337430954,
+      "learning_rate": 1.5508426434387427e-05,
+      "loss": 0.1084,
+      "step": 11860
+    },
+    {
+      "epoch": 1.1238401817837531,
+      "grad_norm": 0.30831944942474365,
+      "learning_rate": 1.550463927286499e-05,
+      "loss": 0.1149,
+      "step": 11870
+    },
+    {
+      "epoch": 1.1247869721643629,
+      "grad_norm": 0.4202735424041748,
+      "learning_rate": 1.550085211134255e-05,
+      "loss": 0.1233,
+      "step": 11880
+    },
+    {
+      "epoch": 1.1257337625449726,
+      "grad_norm": 0.46457767486572266,
+      "learning_rate": 1.549706494982011e-05,
+      "loss": 0.1151,
+      "step": 11890
+    },
+    {
+      "epoch": 1.1266805529255823,
+      "grad_norm": 0.41014325618743896,
+      "learning_rate": 1.549327778829767e-05,
+      "loss": 0.1046,
+      "step": 11900
+    },
+    {
+      "epoch": 1.127627343306192,
+      "grad_norm": 0.3300694227218628,
+      "learning_rate": 1.5489490626775234e-05,
+      "loss": 0.1044,
+      "step": 11910
+    },
+    {
+      "epoch": 1.1285741336868018,
+      "grad_norm": 0.3922729194164276,
+      "learning_rate": 1.5485703465252794e-05,
+      "loss": 0.0988,
+      "step": 11920
+    },
+    {
+      "epoch": 1.1295209240674116,
+      "grad_norm": 0.36446335911750793,
+      "learning_rate": 1.5481916303730357e-05,
+      "loss": 0.1082,
+      "step": 11930
+    },
+    {
+      "epoch": 1.1304677144480213,
+      "grad_norm": 0.3100954294204712,
+      "learning_rate": 1.5478129142207917e-05,
+      "loss": 0.1062,
+      "step": 11940
+    },
+    {
+      "epoch": 1.1314145048286308,
+      "grad_norm": 0.3029150366783142,
+      "learning_rate": 1.5474341980685477e-05,
+      "loss": 0.1166,
+      "step": 11950
+    },
+    {
+      "epoch": 1.1323612952092406,
+      "grad_norm": 0.287994384765625,
+      "learning_rate": 1.547055481916304e-05,
+      "loss": 0.1009,
+      "step": 11960
+    },
+    {
+      "epoch": 1.1333080855898503,
+      "grad_norm": 0.3634592294692993,
+      "learning_rate": 1.54667676576406e-05,
+      "loss": 0.1107,
+      "step": 11970
+    },
+    {
+      "epoch": 1.13425487597046,
+      "grad_norm": 0.39057114720344543,
+      "learning_rate": 1.546298049611816e-05,
+      "loss": 0.1117,
+      "step": 11980
+    },
+    {
+      "epoch": 1.1352016663510698,
+      "grad_norm": 0.3468126356601715,
+      "learning_rate": 1.545919333459572e-05,
+      "loss": 0.1161,
+      "step": 11990
+    },
+    {
+      "epoch": 1.1361484567316795,
+      "grad_norm": 0.43798407912254333,
+      "learning_rate": 1.545540617307328e-05,
+      "loss": 0.1152,
+      "step": 12000
+    },
+    {
+      "epoch": 1.1370952471122893,
+      "grad_norm": 0.2758103907108307,
+      "learning_rate": 1.5451619011550844e-05,
+      "loss": 0.1125,
+      "step": 12010
+    },
+    {
+      "epoch": 1.138042037492899,
+      "grad_norm": 0.3223762512207031,
+      "learning_rate": 1.5447831850028404e-05,
+      "loss": 0.1206,
+      "step": 12020
+    },
+    {
+      "epoch": 1.1389888278735087,
+      "grad_norm": 0.2902495563030243,
+      "learning_rate": 1.5444044688505964e-05,
+      "loss": 0.1007,
+      "step": 12030
+    },
+    {
+      "epoch": 1.1399356182541185,
+      "grad_norm": 0.253961980342865,
+      "learning_rate": 1.5440257526983528e-05,
+      "loss": 0.0983,
+      "step": 12040
+    },
+    {
+      "epoch": 1.1408824086347282,
+      "grad_norm": 0.41474276781082153,
+      "learning_rate": 1.5436470365461088e-05,
+      "loss": 0.1021,
+      "step": 12050
+    },
+    {
+      "epoch": 1.141829199015338,
+      "grad_norm": 0.3824501037597656,
+      "learning_rate": 1.543268320393865e-05,
+      "loss": 0.1171,
+      "step": 12060
+    },
+    {
+      "epoch": 1.1427759893959477,
+      "grad_norm": 0.39116084575653076,
+      "learning_rate": 1.542889604241621e-05,
+      "loss": 0.1052,
+      "step": 12070
+    },
+    {
+      "epoch": 1.1437227797765575,
+      "grad_norm": 0.7037903070449829,
+      "learning_rate": 1.542510888089377e-05,
+      "loss": 0.0976,
+      "step": 12080
+    },
+    {
+      "epoch": 1.1446695701571672,
+      "grad_norm": 0.4843405485153198,
+      "learning_rate": 1.542132171937133e-05,
+      "loss": 0.1128,
+      "step": 12090
+    },
+    {
+      "epoch": 1.145616360537777,
+      "grad_norm": 0.36582592129707336,
+      "learning_rate": 1.5417534557848895e-05,
+      "loss": 0.1073,
+      "step": 12100
+    },
+    {
+      "epoch": 1.1465631509183867,
+      "grad_norm": 0.36037132143974304,
+      "learning_rate": 1.5413747396326455e-05,
+      "loss": 0.1109,
+      "step": 12110
+    },
+    {
+      "epoch": 1.1475099412989964,
+      "grad_norm": 0.30198612809181213,
+      "learning_rate": 1.5409960234804015e-05,
+      "loss": 0.109,
+      "step": 12120
+    },
+    {
+      "epoch": 1.1484567316796062,
+      "grad_norm": 0.32701000571250916,
+      "learning_rate": 1.5406173073281575e-05,
+      "loss": 0.1,
+      "step": 12130
+    },
+    {
+      "epoch": 1.149403522060216,
+      "grad_norm": 0.35844531655311584,
+      "learning_rate": 1.540238591175914e-05,
+      "loss": 0.1204,
+      "step": 12140
+    },
+    {
+      "epoch": 1.1503503124408256,
+      "grad_norm": 0.32474035024642944,
+      "learning_rate": 1.53985987502367e-05,
+      "loss": 0.1059,
+      "step": 12150
+    },
+    {
+      "epoch": 1.1512971028214354,
+      "grad_norm": 0.3637523651123047,
+      "learning_rate": 1.539481158871426e-05,
+      "loss": 0.1058,
+      "step": 12160
+    },
+    {
+      "epoch": 1.1522438932020451,
+      "grad_norm": 0.33581167459487915,
+      "learning_rate": 1.5391024427191822e-05,
+      "loss": 0.1123,
+      "step": 12170
+    },
+    {
+      "epoch": 1.1531906835826549,
+      "grad_norm": 0.2559072971343994,
+      "learning_rate": 1.5387237265669382e-05,
+      "loss": 0.1103,
+      "step": 12180
+    },
+    {
+      "epoch": 1.1541374739632646,
+      "grad_norm": 0.40483200550079346,
+      "learning_rate": 1.5383450104146945e-05,
+      "loss": 0.0937,
+      "step": 12190
+    },
+    {
+      "epoch": 1.1550842643438743,
+      "grad_norm": 0.29024216532707214,
+      "learning_rate": 1.5379662942624505e-05,
+      "loss": 0.1039,
+      "step": 12200
+    },
+    {
+      "epoch": 1.156031054724484,
+      "grad_norm": 0.3387871980667114,
+      "learning_rate": 1.5375875781102066e-05,
+      "loss": 0.1055,
+      "step": 12210
+    },
+    {
+      "epoch": 1.1569778451050938,
+      "grad_norm": 0.33456581830978394,
+      "learning_rate": 1.5372088619579626e-05,
+      "loss": 0.1102,
+      "step": 12220
+    },
+    {
+      "epoch": 1.1579246354857036,
+      "grad_norm": 0.25301578640937805,
+      "learning_rate": 1.5368301458057186e-05,
+      "loss": 0.1114,
+      "step": 12230
+    },
+    {
+      "epoch": 1.1588714258663133,
+      "grad_norm": 0.33042630553245544,
+      "learning_rate": 1.536451429653475e-05,
+      "loss": 0.11,
+      "step": 12240
+    },
+    {
+      "epoch": 1.159818216246923,
+      "grad_norm": 0.2601511478424072,
+      "learning_rate": 1.536072713501231e-05,
+      "loss": 0.1023,
+      "step": 12250
+    },
+    {
+      "epoch": 1.1607650066275326,
+      "grad_norm": 0.3771694004535675,
+      "learning_rate": 1.535693997348987e-05,
+      "loss": 0.1173,
+      "step": 12260
+    },
+    {
+      "epoch": 1.1617117970081423,
+      "grad_norm": 0.3950464725494385,
+      "learning_rate": 1.5353152811967433e-05,
+      "loss": 0.1156,
+      "step": 12270
+    },
+    {
+      "epoch": 1.162658587388752,
+      "grad_norm": 0.36990073323249817,
+      "learning_rate": 1.5349365650444993e-05,
+      "loss": 0.1082,
+      "step": 12280
+    },
+    {
+      "epoch": 1.1636053777693618,
+      "grad_norm": 0.4734830856323242,
+      "learning_rate": 1.5345578488922556e-05,
+      "loss": 0.1025,
+      "step": 12290
+    },
+    {
+      "epoch": 1.1645521681499715,
+      "grad_norm": 0.40759214758872986,
+      "learning_rate": 1.5341791327400116e-05,
+      "loss": 0.1204,
+      "step": 12300
+    },
+    {
+      "epoch": 1.1654989585305813,
+      "grad_norm": 0.30610671639442444,
+      "learning_rate": 1.5338004165877676e-05,
+      "loss": 0.1032,
+      "step": 12310
+    },
+    {
+      "epoch": 1.166445748911191,
+      "grad_norm": 0.3321300446987152,
+      "learning_rate": 1.5334217004355236e-05,
+      "loss": 0.114,
+      "step": 12320
+    },
+    {
+      "epoch": 1.1673925392918008,
+      "grad_norm": 0.31097903847694397,
+      "learning_rate": 1.53304298428328e-05,
+      "loss": 0.1151,
+      "step": 12330
+    },
+    {
+      "epoch": 1.1683393296724105,
+      "grad_norm": 0.32177960872650146,
+      "learning_rate": 1.532664268131036e-05,
+      "loss": 0.1137,
+      "step": 12340
+    },
+    {
+      "epoch": 1.1692861200530202,
+      "grad_norm": 0.3149729371070862,
+      "learning_rate": 1.532285551978792e-05,
+      "loss": 0.1197,
+      "step": 12350
+    },
+    {
+      "epoch": 1.17023291043363,
+      "grad_norm": 0.3405247628688812,
+      "learning_rate": 1.531906835826548e-05,
+      "loss": 0.1027,
+      "step": 12360
+    },
+    {
+      "epoch": 1.1711797008142397,
+      "grad_norm": 0.34930819272994995,
+      "learning_rate": 1.531528119674304e-05,
+      "loss": 0.1085,
+      "step": 12370
+    },
+    {
+      "epoch": 1.1721264911948495,
+      "grad_norm": 0.4254821240901947,
+      "learning_rate": 1.5311494035220603e-05,
+      "loss": 0.1117,
+      "step": 12380
+    },
+    {
+      "epoch": 1.1730732815754592,
+      "grad_norm": 0.380534827709198,
+      "learning_rate": 1.5307706873698163e-05,
+      "loss": 0.1112,
+      "step": 12390
+    },
+    {
+      "epoch": 1.174020071956069,
+      "grad_norm": 0.28946927189826965,
+      "learning_rate": 1.5303919712175727e-05,
+      "loss": 0.1039,
+      "step": 12400
+    },
+    {
+      "epoch": 1.1749668623366787,
+      "grad_norm": 0.2585335671901703,
+      "learning_rate": 1.5300132550653287e-05,
+      "loss": 0.0979,
+      "step": 12410
+    },
+    {
+      "epoch": 1.1759136527172884,
+      "grad_norm": 0.4145877957344055,
+      "learning_rate": 1.5296345389130847e-05,
+      "loss": 0.1072,
+      "step": 12420
+    },
+    {
+      "epoch": 1.1768604430978982,
+      "grad_norm": 0.3534521162509918,
+      "learning_rate": 1.529255822760841e-05,
+      "loss": 0.1125,
+      "step": 12430
+    },
+    {
+      "epoch": 1.177807233478508,
+      "grad_norm": 0.312716007232666,
+      "learning_rate": 1.528877106608597e-05,
+      "loss": 0.1124,
+      "step": 12440
+    },
+    {
+      "epoch": 1.1787540238591177,
+      "grad_norm": 0.3371047377586365,
+      "learning_rate": 1.528498390456353e-05,
+      "loss": 0.1013,
+      "step": 12450
+    },
+    {
+      "epoch": 1.1797008142397274,
+      "grad_norm": 0.3125591278076172,
+      "learning_rate": 1.528119674304109e-05,
+      "loss": 0.1095,
+      "step": 12460
+    },
+    {
+      "epoch": 1.1806476046203371,
+      "grad_norm": 0.3226798474788666,
+      "learning_rate": 1.5277409581518654e-05,
+      "loss": 0.111,
+      "step": 12470
+    },
+    {
+      "epoch": 1.1815943950009469,
+      "grad_norm": 0.3581395745277405,
+      "learning_rate": 1.5273622419996214e-05,
+      "loss": 0.1096,
+      "step": 12480
+    },
+    {
+      "epoch": 1.1825411853815566,
+      "grad_norm": 0.3344843089580536,
+      "learning_rate": 1.5269835258473774e-05,
+      "loss": 0.1111,
+      "step": 12490
+    },
+    {
+      "epoch": 1.1834879757621661,
+      "grad_norm": 0.40416479110717773,
+      "learning_rate": 1.5266048096951337e-05,
+      "loss": 0.1065,
+      "step": 12500
+    },
+    {
+      "epoch": 1.1844347661427759,
+      "grad_norm": 0.3474705219268799,
+      "learning_rate": 1.5262260935428897e-05,
+      "loss": 0.1017,
+      "step": 12510
+    },
+    {
+      "epoch": 1.1853815565233856,
+      "grad_norm": 0.374860554933548,
+      "learning_rate": 1.525847377390646e-05,
+      "loss": 0.1007,
+      "step": 12520
+    },
+    {
+      "epoch": 1.1863283469039954,
+      "grad_norm": 0.33770567178726196,
+      "learning_rate": 1.525468661238402e-05,
+      "loss": 0.1054,
+      "step": 12530
+    },
+    {
+      "epoch": 1.187275137284605,
+      "grad_norm": 0.3405020833015442,
+      "learning_rate": 1.5250899450861581e-05,
+      "loss": 0.1152,
+      "step": 12540
+    },
+    {
+      "epoch": 1.1882219276652148,
+      "grad_norm": 0.3371250033378601,
+      "learning_rate": 1.5247112289339141e-05,
+      "loss": 0.1026,
+      "step": 12550
+    },
+    {
+      "epoch": 1.1891687180458246,
+      "grad_norm": 0.4953118860721588,
+      "learning_rate": 1.5243325127816701e-05,
+      "loss": 0.1165,
+      "step": 12560
+    },
+    {
+      "epoch": 1.1901155084264343,
+      "grad_norm": 0.33394649624824524,
+      "learning_rate": 1.5239537966294265e-05,
+      "loss": 0.1181,
+      "step": 12570
+    },
+    {
+      "epoch": 1.191062298807044,
+      "grad_norm": 0.3339691758155823,
+      "learning_rate": 1.5235750804771825e-05,
+      "loss": 0.106,
+      "step": 12580
+    },
+    {
+      "epoch": 1.1920090891876538,
+      "grad_norm": 0.25260743498802185,
+      "learning_rate": 1.5231963643249386e-05,
+      "loss": 0.1196,
+      "step": 12590
+    },
+    {
+      "epoch": 1.1929558795682635,
+      "grad_norm": 0.307058185338974,
+      "learning_rate": 1.5228176481726946e-05,
+      "loss": 0.1159,
+      "step": 12600
+    },
+    {
+      "epoch": 1.1939026699488733,
+      "grad_norm": 0.3390623927116394,
+      "learning_rate": 1.5224389320204508e-05,
+      "loss": 0.1045,
+      "step": 12610
+    },
+    {
+      "epoch": 1.194849460329483,
+      "grad_norm": 0.357327401638031,
+      "learning_rate": 1.522060215868207e-05,
+      "loss": 0.0975,
+      "step": 12620
+    },
+    {
+      "epoch": 1.1957962507100928,
+      "grad_norm": 0.29861733317375183,
+      "learning_rate": 1.521681499715963e-05,
+      "loss": 0.1007,
+      "step": 12630
+    },
+    {
+      "epoch": 1.1967430410907025,
+      "grad_norm": 0.3431881368160248,
+      "learning_rate": 1.521302783563719e-05,
+      "loss": 0.1111,
+      "step": 12640
+    },
+    {
+      "epoch": 1.1976898314713122,
+      "grad_norm": 0.29756271839141846,
+      "learning_rate": 1.5209240674114752e-05,
+      "loss": 0.1014,
+      "step": 12650
+    },
+    {
+      "epoch": 1.198636621851922,
+      "grad_norm": 0.34353452920913696,
+      "learning_rate": 1.5205453512592313e-05,
+      "loss": 0.1107,
+      "step": 12660
+    },
+    {
+      "epoch": 1.1995834122325317,
+      "grad_norm": 0.31465011835098267,
+      "learning_rate": 1.5201666351069875e-05,
+      "loss": 0.1031,
+      "step": 12670
+    },
+    {
+      "epoch": 1.2005302026131415,
+      "grad_norm": 0.333942174911499,
+      "learning_rate": 1.5197879189547435e-05,
+      "loss": 0.1042,
+      "step": 12680
+    },
+    {
+      "epoch": 1.2014769929937512,
+      "grad_norm": 0.3030639588832855,
+      "learning_rate": 1.5194092028024995e-05,
+      "loss": 0.1015,
+      "step": 12690
+    },
+    {
+      "epoch": 1.202423783374361,
+      "grad_norm": 0.3866049349308014,
+      "learning_rate": 1.5190304866502557e-05,
+      "loss": 0.1069,
+      "step": 12700
+    },
+    {
+      "epoch": 1.2033705737549707,
+      "grad_norm": 0.3382660448551178,
+      "learning_rate": 1.5186517704980119e-05,
+      "loss": 0.1053,
+      "step": 12710
+    },
+    {
+      "epoch": 1.2043173641355804,
+      "grad_norm": 0.4327520430088043,
+      "learning_rate": 1.518273054345768e-05,
+      "loss": 0.1057,
+      "step": 12720
+    },
+    {
+      "epoch": 1.2052641545161902,
+      "grad_norm": 0.30221450328826904,
+      "learning_rate": 1.517894338193524e-05,
+      "loss": 0.1132,
+      "step": 12730
+    },
+    {
+      "epoch": 1.2062109448968,
+      "grad_norm": 0.39057356119155884,
+      "learning_rate": 1.51751562204128e-05,
+      "loss": 0.1081,
+      "step": 12740
+    },
+    {
+      "epoch": 1.2071577352774097,
+      "grad_norm": 0.31764695048332214,
+      "learning_rate": 1.5171369058890364e-05,
+      "loss": 0.1174,
+      "step": 12750
+    },
+    {
+      "epoch": 1.2081045256580194,
+      "grad_norm": 0.4214201271533966,
+      "learning_rate": 1.5167581897367924e-05,
+      "loss": 0.1076,
+      "step": 12760
+    },
+    {
+      "epoch": 1.2090513160386291,
+      "grad_norm": 0.4078432321548462,
+      "learning_rate": 1.5163794735845486e-05,
+      "loss": 0.1035,
+      "step": 12770
+    },
+    {
+      "epoch": 1.2099981064192389,
+      "grad_norm": 0.2991807460784912,
+      "learning_rate": 1.5160007574323046e-05,
+      "loss": 0.1172,
+      "step": 12780
+    },
+    {
+      "epoch": 1.2109448967998486,
+      "grad_norm": 0.3427339494228363,
+      "learning_rate": 1.5156220412800606e-05,
+      "loss": 0.1145,
+      "step": 12790
+    },
+    {
+      "epoch": 1.2118916871804584,
+      "grad_norm": 0.3660449981689453,
+      "learning_rate": 1.515243325127817e-05,
+      "loss": 0.1126,
+      "step": 12800
+    },
+    {
+      "epoch": 1.2128384775610679,
+      "grad_norm": 0.2685498595237732,
+      "learning_rate": 1.514864608975573e-05,
+      "loss": 0.1074,
+      "step": 12810
+    },
+    {
+      "epoch": 1.2137852679416776,
+      "grad_norm": 0.4036467969417572,
+      "learning_rate": 1.514485892823329e-05,
+      "loss": 0.1129,
+      "step": 12820
+    },
+    {
+      "epoch": 1.2147320583222874,
+      "grad_norm": 0.28507718443870544,
+      "learning_rate": 1.5141071766710851e-05,
+      "loss": 0.1054,
+      "step": 12830
+    },
+    {
+      "epoch": 1.215678848702897,
+      "grad_norm": 0.32858598232269287,
+      "learning_rate": 1.5137284605188411e-05,
+      "loss": 0.1081,
+      "step": 12840
+    },
+    {
+      "epoch": 1.2166256390835068,
+      "grad_norm": 0.3813497722148895,
+      "learning_rate": 1.5133497443665975e-05,
+      "loss": 0.1152,
+      "step": 12850
+    },
+    {
+      "epoch": 1.2175724294641166,
+      "grad_norm": 0.31586164236068726,
+      "learning_rate": 1.5129710282143535e-05,
+      "loss": 0.1076,
+      "step": 12860
+    },
+    {
+      "epoch": 1.2185192198447263,
+      "grad_norm": 0.44221732020378113,
+      "learning_rate": 1.5125923120621095e-05,
+      "loss": 0.1092,
+      "step": 12870
+    },
+    {
+      "epoch": 1.219466010225336,
+      "grad_norm": 0.4099266231060028,
+      "learning_rate": 1.5122135959098657e-05,
+      "loss": 0.1171,
+      "step": 12880
+    },
+    {
+      "epoch": 1.2204128006059458,
+      "grad_norm": 0.2790457606315613,
+      "learning_rate": 1.5118348797576218e-05,
+      "loss": 0.1051,
+      "step": 12890
+    },
+    {
+      "epoch": 1.2213595909865556,
+      "grad_norm": 0.40873250365257263,
+      "learning_rate": 1.511456163605378e-05,
+      "loss": 0.1068,
+      "step": 12900
+    },
+    {
+      "epoch": 1.2223063813671653,
+      "grad_norm": 0.34618380665779114,
+      "learning_rate": 1.511077447453134e-05,
+      "loss": 0.1111,
+      "step": 12910
+    },
+    {
+      "epoch": 1.223253171747775,
+      "grad_norm": 0.3440490961074829,
+      "learning_rate": 1.51069873130089e-05,
+      "loss": 0.1128,
+      "step": 12920
+    },
+    {
+      "epoch": 1.2241999621283848,
+      "grad_norm": 0.34496238827705383,
+      "learning_rate": 1.5103200151486462e-05,
+      "loss": 0.0993,
+      "step": 12930
+    },
+    {
+      "epoch": 1.2251467525089945,
+      "grad_norm": 0.3635864853858948,
+      "learning_rate": 1.5099412989964024e-05,
+      "loss": 0.0994,
+      "step": 12940
+    },
+    {
+      "epoch": 1.2260935428896043,
+      "grad_norm": 0.3953652083873749,
+      "learning_rate": 1.5095625828441585e-05,
+      "loss": 0.1148,
+      "step": 12950
+    },
+    {
+      "epoch": 1.227040333270214,
+      "grad_norm": 0.31408146023750305,
+      "learning_rate": 1.5091838666919145e-05,
+      "loss": 0.0987,
+      "step": 12960
+    },
+    {
+      "epoch": 1.2279871236508237,
+      "grad_norm": 0.47466138005256653,
+      "learning_rate": 1.5088051505396706e-05,
+      "loss": 0.1049,
+      "step": 12970
+    },
+    {
+      "epoch": 1.2289339140314335,
+      "grad_norm": 0.36139219999313354,
+      "learning_rate": 1.5084264343874267e-05,
+      "loss": 0.1182,
+      "step": 12980
+    },
+    {
+      "epoch": 1.2298807044120432,
+      "grad_norm": 0.31297993659973145,
+      "learning_rate": 1.5080477182351829e-05,
+      "loss": 0.1084,
+      "step": 12990
+    },
+    {
+      "epoch": 1.230827494792653,
+      "grad_norm": 0.35729625821113586,
+      "learning_rate": 1.5076690020829389e-05,
+      "loss": 0.1043,
+      "step": 13000
+    },
+    {
+      "epoch": 1.2317742851732627,
+      "grad_norm": 0.4169113039970398,
+      "learning_rate": 1.507290285930695e-05,
+      "loss": 0.1033,
+      "step": 13010
+    },
+    {
+      "epoch": 1.2327210755538724,
+      "grad_norm": 0.39845699071884155,
+      "learning_rate": 1.506911569778451e-05,
+      "loss": 0.1143,
+      "step": 13020
+    },
+    {
+      "epoch": 1.2336678659344822,
+      "grad_norm": 0.270967960357666,
+      "learning_rate": 1.5065328536262074e-05,
+      "loss": 0.1063,
+      "step": 13030
+    },
+    {
+      "epoch": 1.234614656315092,
+      "grad_norm": 0.3269328474998474,
+      "learning_rate": 1.5061541374739634e-05,
+      "loss": 0.0983,
+      "step": 13040
+    },
+    {
+      "epoch": 1.2355614466957014,
+      "grad_norm": 0.3136744499206543,
+      "learning_rate": 1.5057754213217194e-05,
+      "loss": 0.0919,
+      "step": 13050
+    },
+    {
+      "epoch": 1.2365082370763112,
+      "grad_norm": 0.3122495710849762,
+      "learning_rate": 1.5053967051694756e-05,
+      "loss": 0.1076,
+      "step": 13060
+    },
+    {
+      "epoch": 1.237455027456921,
+      "grad_norm": 0.4010438024997711,
+      "learning_rate": 1.5050179890172316e-05,
+      "loss": 0.1116,
+      "step": 13070
+    },
+    {
+      "epoch": 1.2384018178375307,
+      "grad_norm": 0.4302702248096466,
+      "learning_rate": 1.504639272864988e-05,
+      "loss": 0.1219,
+      "step": 13080
+    },
+    {
+      "epoch": 1.2393486082181404,
+      "grad_norm": 0.35059264302253723,
+      "learning_rate": 1.504260556712744e-05,
+      "loss": 0.12,
+      "step": 13090
+    },
+    {
+      "epoch": 1.2402953985987502,
+      "grad_norm": 0.30450910329818726,
+      "learning_rate": 1.5038818405605e-05,
+      "loss": 0.1018,
+      "step": 13100
+    },
+    {
+      "epoch": 1.24124218897936,
+      "grad_norm": 0.3359741270542145,
+      "learning_rate": 1.5035031244082561e-05,
+      "loss": 0.1174,
+      "step": 13110
+    },
+    {
+      "epoch": 1.2421889793599696,
+      "grad_norm": 0.2931978702545166,
+      "learning_rate": 1.5031244082560122e-05,
+      "loss": 0.1101,
+      "step": 13120
+    },
+    {
+      "epoch": 1.2431357697405794,
+      "grad_norm": 0.27383294701576233,
+      "learning_rate": 1.5027456921037685e-05,
+      "loss": 0.1108,
+      "step": 13130
+    },
+    {
+      "epoch": 1.2440825601211891,
+      "grad_norm": 0.33028000593185425,
+      "learning_rate": 1.5023669759515245e-05,
+      "loss": 0.1055,
+      "step": 13140
+    },
+    {
+      "epoch": 1.2450293505017989,
+      "grad_norm": 0.3667839765548706,
+      "learning_rate": 1.5019882597992805e-05,
+      "loss": 0.1099,
+      "step": 13150
+    },
+    {
+      "epoch": 1.2459761408824086,
+      "grad_norm": 0.3085506558418274,
+      "learning_rate": 1.5016095436470367e-05,
+      "loss": 0.1124,
+      "step": 13160
+    },
+    {
+      "epoch": 1.2469229312630183,
+      "grad_norm": 0.3162235617637634,
+      "learning_rate": 1.5012308274947929e-05,
+      "loss": 0.1133,
+      "step": 13170
+    },
+    {
+      "epoch": 1.247869721643628,
+      "grad_norm": 0.3025159239768982,
+      "learning_rate": 1.5008521113425489e-05,
+      "loss": 0.1079,
+      "step": 13180
+    },
+    {
+      "epoch": 1.2488165120242378,
+      "grad_norm": 0.3852561414241791,
+      "learning_rate": 1.500473395190305e-05,
+      "loss": 0.1041,
+      "step": 13190
+    },
+    {
+      "epoch": 1.2497633024048476,
+      "grad_norm": 0.3863688111305237,
+      "learning_rate": 1.500094679038061e-05,
+      "loss": 0.1081,
+      "step": 13200
+    },
+    {
+      "epoch": 1.2507100927854573,
+      "grad_norm": 0.42680519819259644,
+      "learning_rate": 1.499715962885817e-05,
+      "loss": 0.1011,
+      "step": 13210
+    },
+    {
+      "epoch": 1.251656883166067,
+      "grad_norm": 0.3259578049182892,
+      "learning_rate": 1.4993372467335734e-05,
+      "loss": 0.1116,
+      "step": 13220
+    },
+    {
+      "epoch": 1.2526036735466768,
+      "grad_norm": 0.3074738383293152,
+      "learning_rate": 1.4989585305813294e-05,
+      "loss": 0.1133,
+      "step": 13230
+    },
+    {
+      "epoch": 1.2535504639272865,
+      "grad_norm": 0.4082379639148712,
+      "learning_rate": 1.4985798144290856e-05,
+      "loss": 0.1107,
+      "step": 13240
+    },
+    {
+      "epoch": 1.2544972543078963,
+      "grad_norm": 0.35676753520965576,
+      "learning_rate": 1.4982010982768416e-05,
+      "loss": 0.1068,
+      "step": 13250
+    },
+    {
+      "epoch": 1.255444044688506,
+      "grad_norm": 0.34967681765556335,
+      "learning_rate": 1.4978223821245976e-05,
+      "loss": 0.1081,
+      "step": 13260
+    },
+    {
+      "epoch": 1.2563908350691158,
+      "grad_norm": 0.37591052055358887,
+      "learning_rate": 1.497443665972354e-05,
+      "loss": 0.1185,
+      "step": 13270
+    },
+    {
+      "epoch": 1.2573376254497255,
+      "grad_norm": 0.3394329845905304,
+      "learning_rate": 1.49706494982011e-05,
+      "loss": 0.1147,
+      "step": 13280
+    },
+    {
+      "epoch": 1.2582844158303352,
+      "grad_norm": 0.26928722858428955,
+      "learning_rate": 1.4966862336678661e-05,
+      "loss": 0.1077,
+      "step": 13290
+    },
+    {
+      "epoch": 1.259231206210945,
+      "grad_norm": 0.35786399245262146,
+      "learning_rate": 1.4963075175156221e-05,
+      "loss": 0.1024,
+      "step": 13300
+    },
+    {
+      "epoch": 1.2601779965915547,
+      "grad_norm": 0.2981221377849579,
+      "learning_rate": 1.4959288013633784e-05,
+      "loss": 0.1101,
+      "step": 13310
+    },
+    {
+      "epoch": 1.2611247869721645,
+      "grad_norm": 0.33681800961494446,
+      "learning_rate": 1.4955500852111345e-05,
+      "loss": 0.1158,
+      "step": 13320
+    },
+    {
+      "epoch": 1.2620715773527742,
+      "grad_norm": 0.3591706156730652,
+      "learning_rate": 1.4951713690588905e-05,
+      "loss": 0.114,
+      "step": 13330
+    },
+    {
+      "epoch": 1.263018367733384,
+      "grad_norm": 0.35171523690223694,
+      "learning_rate": 1.4947926529066466e-05,
+      "loss": 0.1057,
+      "step": 13340
+    },
+    {
+      "epoch": 1.2639651581139937,
+      "grad_norm": 0.36935341358184814,
+      "learning_rate": 1.4944139367544026e-05,
+      "loss": 0.1111,
+      "step": 13350
+    },
+    {
+      "epoch": 1.2649119484946034,
+      "grad_norm": 0.3233521580696106,
+      "learning_rate": 1.4940352206021588e-05,
+      "loss": 0.1015,
+      "step": 13360
+    },
+    {
+      "epoch": 1.2658587388752132,
+      "grad_norm": 0.408893346786499,
+      "learning_rate": 1.493656504449915e-05,
+      "loss": 0.1197,
+      "step": 13370
+    },
+    {
+      "epoch": 1.2668055292558227,
+      "grad_norm": 0.38509801030158997,
+      "learning_rate": 1.493277788297671e-05,
+      "loss": 0.1036,
+      "step": 13380
+    },
+    {
+      "epoch": 1.2677523196364324,
+      "grad_norm": 0.3889254331588745,
+      "learning_rate": 1.492899072145427e-05,
+      "loss": 0.1089,
+      "step": 13390
+    },
+    {
+      "epoch": 1.2686991100170422,
+      "grad_norm": 0.3505185544490814,
+      "learning_rate": 1.4925203559931832e-05,
+      "loss": 0.1083,
+      "step": 13400
+    },
+    {
+      "epoch": 1.269645900397652,
+      "grad_norm": 0.3532257676124573,
+      "learning_rate": 1.4921416398409393e-05,
+      "loss": 0.1123,
+      "step": 13410
+    },
+    {
+      "epoch": 1.2705926907782616,
+      "grad_norm": 0.4355742633342743,
+      "learning_rate": 1.4917629236886955e-05,
+      "loss": 0.1194,
+      "step": 13420
+    },
+    {
+      "epoch": 1.2715394811588714,
+      "grad_norm": 0.3455640971660614,
+      "learning_rate": 1.4913842075364515e-05,
+      "loss": 0.1062,
+      "step": 13430
+    },
+    {
+      "epoch": 1.2724862715394811,
+      "grad_norm": 0.420749306678772,
+      "learning_rate": 1.4910054913842075e-05,
+      "loss": 0.1005,
+      "step": 13440
+    },
+    {
+      "epoch": 1.2734330619200909,
+      "grad_norm": 0.4022374451160431,
+      "learning_rate": 1.4906267752319639e-05,
+      "loss": 0.1089,
+      "step": 13450
+    },
+    {
+      "epoch": 1.2743798523007006,
+      "grad_norm": 0.31215524673461914,
+      "learning_rate": 1.4902480590797199e-05,
+      "loss": 0.1126,
+      "step": 13460
+    },
+    {
+      "epoch": 1.2753266426813104,
+      "grad_norm": 0.37119176983833313,
+      "learning_rate": 1.489869342927476e-05,
+      "loss": 0.1088,
+      "step": 13470
+    },
+    {
+      "epoch": 1.27627343306192,
+      "grad_norm": 0.31066417694091797,
+      "learning_rate": 1.489490626775232e-05,
+      "loss": 0.1089,
+      "step": 13480
+    },
+    {
+      "epoch": 1.2772202234425298,
+      "grad_norm": 0.3264316916465759,
+      "learning_rate": 1.489111910622988e-05,
+      "loss": 0.1151,
+      "step": 13490
+    },
+    {
+      "epoch": 1.2781670138231396,
+      "grad_norm": 0.2913450002670288,
+      "learning_rate": 1.4887331944707444e-05,
+      "loss": 0.1135,
+      "step": 13500
+    },
+    {
+      "epoch": 1.2791138042037493,
+      "grad_norm": 0.3532226085662842,
+      "learning_rate": 1.4883544783185004e-05,
+      "loss": 0.1064,
+      "step": 13510
+    },
+    {
+      "epoch": 1.280060594584359,
+      "grad_norm": 0.3865976929664612,
+      "learning_rate": 1.4879757621662566e-05,
+      "loss": 0.1015,
+      "step": 13520
+    },
+    {
+      "epoch": 1.2810073849649688,
+      "grad_norm": 0.37290051579475403,
+      "learning_rate": 1.4875970460140126e-05,
+      "loss": 0.0982,
+      "step": 13530
+    },
+    {
+      "epoch": 1.2819541753455785,
+      "grad_norm": 0.4275021553039551,
+      "learning_rate": 1.4872183298617686e-05,
+      "loss": 0.1177,
+      "step": 13540
+    },
+    {
+      "epoch": 1.2829009657261883,
+      "grad_norm": 0.30412721633911133,
+      "learning_rate": 1.486839613709525e-05,
+      "loss": 0.1116,
+      "step": 13550
+    },
+    {
+      "epoch": 1.283847756106798,
+      "grad_norm": 0.31443697214126587,
+      "learning_rate": 1.486460897557281e-05,
+      "loss": 0.1128,
+      "step": 13560
+    },
+    {
+      "epoch": 1.2847945464874078,
+      "grad_norm": 0.23646575212478638,
+      "learning_rate": 1.486082181405037e-05,
+      "loss": 0.107,
+      "step": 13570
+    },
+    {
+      "epoch": 1.2857413368680175,
+      "grad_norm": 0.4218645691871643,
+      "learning_rate": 1.4857034652527931e-05,
+      "loss": 0.1021,
+      "step": 13580
+    },
+    {
+      "epoch": 1.286688127248627,
+      "grad_norm": 0.3243103623390198,
+      "learning_rate": 1.4853247491005493e-05,
+      "loss": 0.1037,
+      "step": 13590
+    },
+    {
+      "epoch": 1.2876349176292368,
+      "grad_norm": 0.3228418529033661,
+      "learning_rate": 1.4849460329483055e-05,
+      "loss": 0.1002,
+      "step": 13600
+    },
+    {
+      "epoch": 1.2885817080098465,
+      "grad_norm": 0.3226083517074585,
+      "learning_rate": 1.4845673167960615e-05,
+      "loss": 0.1123,
+      "step": 13610
+    },
+    {
+      "epoch": 1.2895284983904562,
+      "grad_norm": 0.39514216780662537,
+      "learning_rate": 1.4841886006438175e-05,
+      "loss": 0.1105,
+      "step": 13620
+    },
+    {
+      "epoch": 1.290475288771066,
+      "grad_norm": 0.31612706184387207,
+      "learning_rate": 1.4838098844915737e-05,
+      "loss": 0.1117,
+      "step": 13630
+    },
+    {
+      "epoch": 1.2914220791516757,
+      "grad_norm": 0.3052833378314972,
+      "learning_rate": 1.4834311683393298e-05,
+      "loss": 0.108,
+      "step": 13640
+    },
+    {
+      "epoch": 1.2923688695322855,
+      "grad_norm": 0.32270005345344543,
+      "learning_rate": 1.483052452187086e-05,
+      "loss": 0.107,
+      "step": 13650
+    },
+    {
+      "epoch": 1.2933156599128952,
+      "grad_norm": 0.3900124430656433,
+      "learning_rate": 1.482673736034842e-05,
+      "loss": 0.1161,
+      "step": 13660
+    },
+    {
+      "epoch": 1.294262450293505,
+      "grad_norm": 0.3898795247077942,
+      "learning_rate": 1.482295019882598e-05,
+      "loss": 0.109,
+      "step": 13670
+    },
+    {
+      "epoch": 1.2952092406741147,
+      "grad_norm": 0.3327631652355194,
+      "learning_rate": 1.4819163037303542e-05,
+      "loss": 0.1065,
+      "step": 13680
+    },
+    {
+      "epoch": 1.2961560310547244,
+      "grad_norm": 0.2671869695186615,
+      "learning_rate": 1.4815375875781104e-05,
+      "loss": 0.1151,
+      "step": 13690
+    },
+    {
+      "epoch": 1.2971028214353342,
+      "grad_norm": 0.36693522334098816,
+      "learning_rate": 1.4811588714258665e-05,
+      "loss": 0.1253,
+      "step": 13700
+    },
+    {
+      "epoch": 1.298049611815944,
+      "grad_norm": 0.37485846877098083,
+      "learning_rate": 1.4807801552736225e-05,
+      "loss": 0.1075,
+      "step": 13710
+    },
+    {
+      "epoch": 1.2989964021965537,
+      "grad_norm": 0.3453369736671448,
+      "learning_rate": 1.4804014391213785e-05,
+      "loss": 0.1098,
+      "step": 13720
+    },
+    {
+      "epoch": 1.2999431925771634,
+      "grad_norm": 0.29103124141693115,
+      "learning_rate": 1.4800227229691349e-05,
+      "loss": 0.1063,
+      "step": 13730
+    },
+    {
+      "epoch": 1.3008899829577731,
+      "grad_norm": 0.5688313245773315,
+      "learning_rate": 1.4796440068168909e-05,
+      "loss": 0.1125,
+      "step": 13740
+    },
+    {
+      "epoch": 1.3018367733383829,
+      "grad_norm": 0.4438546597957611,
+      "learning_rate": 1.4792652906646469e-05,
+      "loss": 0.1085,
+      "step": 13750
+    },
+    {
+      "epoch": 1.3027835637189926,
+      "grad_norm": 0.39846259355545044,
+      "learning_rate": 1.478886574512403e-05,
+      "loss": 0.1084,
+      "step": 13760
+    },
+    {
+      "epoch": 1.3037303540996024,
+      "grad_norm": 0.362238347530365,
+      "learning_rate": 1.478507858360159e-05,
+      "loss": 0.1084,
+      "step": 13770
+    },
+    {
+      "epoch": 1.304677144480212,
+      "grad_norm": 0.4283710718154907,
+      "learning_rate": 1.4781291422079154e-05,
+      "loss": 0.1117,
+      "step": 13780
+    },
+    {
+      "epoch": 1.3056239348608218,
+      "grad_norm": 0.2781156301498413,
+      "learning_rate": 1.4777504260556714e-05,
+      "loss": 0.1013,
+      "step": 13790
+    },
+    {
+      "epoch": 1.3065707252414316,
+      "grad_norm": 0.31439098715782166,
+      "learning_rate": 1.4773717099034274e-05,
+      "loss": 0.1084,
+      "step": 13800
+    },
+    {
+      "epoch": 1.3075175156220413,
+      "grad_norm": 0.4003913998603821,
+      "learning_rate": 1.4769929937511836e-05,
+      "loss": 0.1096,
+      "step": 13810
+    },
+    {
+      "epoch": 1.308464306002651,
+      "grad_norm": 0.39215022325515747,
+      "learning_rate": 1.4766142775989396e-05,
+      "loss": 0.1009,
+      "step": 13820
+    },
+    {
+      "epoch": 1.3094110963832608,
+      "grad_norm": 0.5251069068908691,
+      "learning_rate": 1.476235561446696e-05,
+      "loss": 0.11,
+      "step": 13830
+    },
+    {
+      "epoch": 1.3103578867638705,
+      "grad_norm": 0.4530577063560486,
+      "learning_rate": 1.475856845294452e-05,
+      "loss": 0.104,
+      "step": 13840
+    },
+    {
+      "epoch": 1.3113046771444803,
+      "grad_norm": 0.3963939845561981,
+      "learning_rate": 1.475478129142208e-05,
+      "loss": 0.1095,
+      "step": 13850
+    },
+    {
+      "epoch": 1.31225146752509,
+      "grad_norm": 0.2831088602542877,
+      "learning_rate": 1.4750994129899641e-05,
+      "loss": 0.0973,
+      "step": 13860
+    },
+    {
+      "epoch": 1.3131982579056998,
+      "grad_norm": 0.3418594002723694,
+      "learning_rate": 1.4747206968377203e-05,
+      "loss": 0.099,
+      "step": 13870
+    },
+    {
+      "epoch": 1.3141450482863095,
+      "grad_norm": 0.35267600417137146,
+      "learning_rate": 1.4743419806854765e-05,
+      "loss": 0.106,
+      "step": 13880
+    },
+    {
+      "epoch": 1.3150918386669193,
+      "grad_norm": 0.35564425587654114,
+      "learning_rate": 1.4739632645332325e-05,
+      "loss": 0.121,
+      "step": 13890
+    },
+    {
+      "epoch": 1.316038629047529,
+      "grad_norm": 0.37613263726234436,
+      "learning_rate": 1.4735845483809885e-05,
+      "loss": 0.1255,
+      "step": 13900
+    },
+    {
+      "epoch": 1.3169854194281387,
+      "grad_norm": 0.3647848665714264,
+      "learning_rate": 1.4732058322287445e-05,
+      "loss": 0.1113,
+      "step": 13910
+    },
+    {
+      "epoch": 1.3179322098087485,
+      "grad_norm": 0.4477173686027527,
+      "learning_rate": 1.4728271160765008e-05,
+      "loss": 0.1099,
+      "step": 13920
+    },
+    {
+      "epoch": 1.318879000189358,
+      "grad_norm": 0.3369513154029846,
+      "learning_rate": 1.4724483999242569e-05,
+      "loss": 0.108,
+      "step": 13930
+    },
+    {
+      "epoch": 1.3198257905699677,
+      "grad_norm": 0.32130277156829834,
+      "learning_rate": 1.472069683772013e-05,
+      "loss": 0.1066,
+      "step": 13940
+    },
+    {
+      "epoch": 1.3207725809505775,
+      "grad_norm": 0.3958997428417206,
+      "learning_rate": 1.471690967619769e-05,
+      "loss": 0.1084,
+      "step": 13950
+    },
+    {
+      "epoch": 1.3217193713311872,
+      "grad_norm": 0.27947115898132324,
+      "learning_rate": 1.4713122514675254e-05,
+      "loss": 0.1051,
+      "step": 13960
+    },
+    {
+      "epoch": 1.322666161711797,
+      "grad_norm": 0.2896656095981598,
+      "learning_rate": 1.4709335353152814e-05,
+      "loss": 0.0985,
+      "step": 13970
+    },
+    {
+      "epoch": 1.3236129520924067,
+      "grad_norm": 0.3368990123271942,
+      "learning_rate": 1.4705548191630374e-05,
+      "loss": 0.1037,
+      "step": 13980
+    },
+    {
+      "epoch": 1.3245597424730164,
+      "grad_norm": 0.4216310977935791,
+      "learning_rate": 1.4701761030107936e-05,
+      "loss": 0.1084,
+      "step": 13990
+    },
+    {
+      "epoch": 1.3255065328536262,
+      "grad_norm": 0.29863274097442627,
+      "learning_rate": 1.4697973868585496e-05,
+      "loss": 0.1108,
+      "step": 14000
+    },
+    {
+      "epoch": 1.326453323234236,
+      "grad_norm": 0.3625233769416809,
+      "learning_rate": 1.4694186707063059e-05,
+      "loss": 0.1151,
+      "step": 14010
+    },
+    {
+      "epoch": 1.3274001136148457,
+      "grad_norm": 0.3410845696926117,
+      "learning_rate": 1.4690399545540619e-05,
+      "loss": 0.1129,
+      "step": 14020
+    },
+    {
+      "epoch": 1.3283469039954554,
+      "grad_norm": 0.4437841773033142,
+      "learning_rate": 1.468661238401818e-05,
+      "loss": 0.112,
+      "step": 14030
+    },
+    {
+      "epoch": 1.3292936943760651,
+      "grad_norm": 0.3411904573440552,
+      "learning_rate": 1.4682825222495741e-05,
+      "loss": 0.1099,
+      "step": 14040
+    },
+    {
+      "epoch": 1.330240484756675,
+      "grad_norm": 0.36379116773605347,
+      "learning_rate": 1.4679038060973301e-05,
+      "loss": 0.1121,
+      "step": 14050
+    },
+    {
+      "epoch": 1.3311872751372846,
+      "grad_norm": 0.36595743894577026,
+      "learning_rate": 1.4675250899450864e-05,
+      "loss": 0.1061,
+      "step": 14060
+    },
+    {
+      "epoch": 1.3321340655178944,
+      "grad_norm": 0.3377189040184021,
+      "learning_rate": 1.4671463737928424e-05,
+      "loss": 0.1058,
+      "step": 14070
+    },
+    {
+      "epoch": 1.3330808558985041,
+      "grad_norm": 0.321304589509964,
+      "learning_rate": 1.4667676576405985e-05,
+      "loss": 0.116,
+      "step": 14080
+    },
+    {
+      "epoch": 1.3340276462791139,
+      "grad_norm": 0.3352893590927124,
+      "learning_rate": 1.4663889414883545e-05,
+      "loss": 0.1121,
+      "step": 14090
+    },
+    {
+      "epoch": 1.3349744366597236,
+      "grad_norm": 0.3742593228816986,
+      "learning_rate": 1.4660102253361108e-05,
+      "loss": 0.1076,
+      "step": 14100
+    },
+    {
+      "epoch": 1.3359212270403333,
+      "grad_norm": 0.2847396731376648,
+      "learning_rate": 1.4656315091838668e-05,
+      "loss": 0.1029,
+      "step": 14110
+    },
+    {
+      "epoch": 1.336868017420943,
+      "grad_norm": 0.36076590418815613,
+      "learning_rate": 1.465252793031623e-05,
+      "loss": 0.0999,
+      "step": 14120
+    },
+    {
+      "epoch": 1.3378148078015528,
+      "grad_norm": 0.3037276566028595,
+      "learning_rate": 1.464874076879379e-05,
+      "loss": 0.107,
+      "step": 14130
+    },
+    {
+      "epoch": 1.3387615981821623,
+      "grad_norm": 0.41113731265068054,
+      "learning_rate": 1.464495360727135e-05,
+      "loss": 0.1035,
+      "step": 14140
+    },
+    {
+      "epoch": 1.339708388562772,
+      "grad_norm": 0.3692213296890259,
+      "learning_rate": 1.4641166445748913e-05,
+      "loss": 0.1125,
+      "step": 14150
+    },
+    {
+      "epoch": 1.3406551789433818,
+      "grad_norm": 0.31594541668891907,
+      "learning_rate": 1.4637379284226473e-05,
+      "loss": 0.1016,
+      "step": 14160
+    },
+    {
+      "epoch": 1.3416019693239916,
+      "grad_norm": 0.3881237506866455,
+      "learning_rate": 1.4633592122704035e-05,
+      "loss": 0.1063,
+      "step": 14170
+    },
+    {
+      "epoch": 1.3425487597046013,
+      "grad_norm": 0.34778666496276855,
+      "learning_rate": 1.4629804961181595e-05,
+      "loss": 0.1044,
+      "step": 14180
+    },
+    {
+      "epoch": 1.343495550085211,
+      "grad_norm": 0.38068491220474243,
+      "learning_rate": 1.4626017799659155e-05,
+      "loss": 0.1088,
+      "step": 14190
+    },
+    {
+      "epoch": 1.3444423404658208,
+      "grad_norm": 0.379708856344223,
+      "learning_rate": 1.4622230638136719e-05,
+      "loss": 0.1065,
+      "step": 14200
+    },
+    {
+      "epoch": 1.3453891308464305,
+      "grad_norm": 0.47989198565483093,
+      "learning_rate": 1.4618443476614279e-05,
+      "loss": 0.1083,
+      "step": 14210
+    },
+    {
+      "epoch": 1.3463359212270403,
+      "grad_norm": 0.3261682689189911,
+      "learning_rate": 1.461465631509184e-05,
+      "loss": 0.1131,
+      "step": 14220
+    },
+    {
+      "epoch": 1.34728271160765,
+      "grad_norm": 0.27690833806991577,
+      "learning_rate": 1.46108691535694e-05,
+      "loss": 0.1107,
+      "step": 14230
+    },
+    {
+      "epoch": 1.3482295019882597,
+      "grad_norm": 0.3644765317440033,
+      "learning_rate": 1.4607081992046962e-05,
+      "loss": 0.1078,
+      "step": 14240
+    },
+    {
+      "epoch": 1.3491762923688695,
+      "grad_norm": 0.40070703625679016,
+      "learning_rate": 1.4603294830524524e-05,
+      "loss": 0.101,
+      "step": 14250
+    },
+    {
+      "epoch": 1.3501230827494792,
+      "grad_norm": 0.4271068871021271,
+      "learning_rate": 1.4599507669002084e-05,
+      "loss": 0.0979,
+      "step": 14260
+    },
+    {
+      "epoch": 1.351069873130089,
+      "grad_norm": 0.3761310875415802,
+      "learning_rate": 1.4595720507479644e-05,
+      "loss": 0.1151,
+      "step": 14270
+    },
+    {
+      "epoch": 1.3520166635106987,
+      "grad_norm": 0.4187332093715668,
+      "learning_rate": 1.4591933345957206e-05,
+      "loss": 0.1112,
+      "step": 14280
+    },
+    {
+      "epoch": 1.3529634538913085,
+      "grad_norm": 0.2853829264640808,
+      "learning_rate": 1.4588146184434768e-05,
+      "loss": 0.1051,
+      "step": 14290
+    },
+    {
+      "epoch": 1.3539102442719182,
+      "grad_norm": 0.4218035638332367,
+      "learning_rate": 1.458435902291233e-05,
+      "loss": 0.1153,
+      "step": 14300
+    },
+    {
+      "epoch": 1.354857034652528,
+      "grad_norm": 0.46550899744033813,
+      "learning_rate": 1.458057186138989e-05,
+      "loss": 0.1111,
+      "step": 14310
+    },
+    {
+      "epoch": 1.3558038250331377,
+      "grad_norm": 0.3463427722454071,
+      "learning_rate": 1.457678469986745e-05,
+      "loss": 0.1056,
+      "step": 14320
+    },
+    {
+      "epoch": 1.3567506154137474,
+      "grad_norm": 0.35000738501548767,
+      "learning_rate": 1.4572997538345011e-05,
+      "loss": 0.1043,
+      "step": 14330
+    },
+    {
+      "epoch": 1.3576974057943572,
+      "grad_norm": 0.3538738191127777,
+      "learning_rate": 1.4569210376822573e-05,
+      "loss": 0.1177,
+      "step": 14340
+    },
+    {
+      "epoch": 1.358644196174967,
+      "grad_norm": 0.33624202013015747,
+      "learning_rate": 1.4565423215300135e-05,
+      "loss": 0.1143,
+      "step": 14350
+    },
+    {
+      "epoch": 1.3595909865555766,
+      "grad_norm": 0.3360025882720947,
+      "learning_rate": 1.4561636053777695e-05,
+      "loss": 0.1113,
+      "step": 14360
+    },
+    {
+      "epoch": 1.3605377769361864,
+      "grad_norm": 0.400475412607193,
+      "learning_rate": 1.4557848892255255e-05,
+      "loss": 0.106,
+      "step": 14370
+    },
+    {
+      "epoch": 1.3614845673167961,
+      "grad_norm": 0.318945050239563,
+      "learning_rate": 1.4554061730732818e-05,
+      "loss": 0.123,
+      "step": 14380
+    },
+    {
+      "epoch": 1.3624313576974059,
+      "grad_norm": 0.3328794836997986,
+      "learning_rate": 1.4550274569210378e-05,
+      "loss": 0.1189,
+      "step": 14390
+    },
+    {
+      "epoch": 1.3633781480780156,
+      "grad_norm": 0.3011545240879059,
+      "learning_rate": 1.454648740768794e-05,
+      "loss": 0.1053,
+      "step": 14400
+    },
+    {
+      "epoch": 1.3643249384586253,
+      "grad_norm": 0.3530985116958618,
+      "learning_rate": 1.45427002461655e-05,
+      "loss": 0.1049,
+      "step": 14410
+    },
+    {
+      "epoch": 1.365271728839235,
+      "grad_norm": 0.31439208984375,
+      "learning_rate": 1.453891308464306e-05,
+      "loss": 0.1172,
+      "step": 14420
+    },
+    {
+      "epoch": 1.3662185192198448,
+      "grad_norm": 0.2875416874885559,
+      "learning_rate": 1.4535125923120624e-05,
+      "loss": 0.1038,
+      "step": 14430
+    },
+    {
+      "epoch": 1.3671653096004546,
+      "grad_norm": 0.36960655450820923,
+      "learning_rate": 1.4531338761598184e-05,
+      "loss": 0.1199,
+      "step": 14440
+    },
+    {
+      "epoch": 1.3681120999810643,
+      "grad_norm": 0.34971728920936584,
+      "learning_rate": 1.4527551600075744e-05,
+      "loss": 0.1161,
+      "step": 14450
+    },
+    {
+      "epoch": 1.369058890361674,
+      "grad_norm": 0.31043538451194763,
+      "learning_rate": 1.4523764438553305e-05,
+      "loss": 0.1091,
+      "step": 14460
+    },
+    {
+      "epoch": 1.3700056807422838,
+      "grad_norm": 0.3760201632976532,
+      "learning_rate": 1.4519977277030865e-05,
+      "loss": 0.1136,
+      "step": 14470
+    },
+    {
+      "epoch": 1.3709524711228933,
+      "grad_norm": 0.3248160779476166,
+      "learning_rate": 1.4516190115508429e-05,
+      "loss": 0.1093,
+      "step": 14480
+    },
+    {
+      "epoch": 1.371899261503503,
+      "grad_norm": 0.4631254971027374,
+      "learning_rate": 1.4512402953985989e-05,
+      "loss": 0.1025,
+      "step": 14490
+    },
+    {
+      "epoch": 1.3728460518841128,
+      "grad_norm": 0.31244543194770813,
+      "learning_rate": 1.4508615792463549e-05,
+      "loss": 0.1068,
+      "step": 14500
+    },
+    {
+      "epoch": 1.3737928422647225,
+      "grad_norm": 0.35992249846458435,
+      "learning_rate": 1.450482863094111e-05,
+      "loss": 0.1116,
+      "step": 14510
+    },
+    {
+      "epoch": 1.3747396326453323,
+      "grad_norm": 0.4582676291465759,
+      "learning_rate": 1.4501041469418672e-05,
+      "loss": 0.1007,
+      "step": 14520
+    },
+    {
+      "epoch": 1.375686423025942,
+      "grad_norm": 0.3570781648159027,
+      "learning_rate": 1.4497254307896234e-05,
+      "loss": 0.1125,
+      "step": 14530
+    },
+    {
+      "epoch": 1.3766332134065518,
+      "grad_norm": 0.29250195622444153,
+      "learning_rate": 1.4493467146373794e-05,
+      "loss": 0.1095,
+      "step": 14540
+    },
+    {
+      "epoch": 1.3775800037871615,
+      "grad_norm": 0.4136636257171631,
+      "learning_rate": 1.4489679984851354e-05,
+      "loss": 0.1174,
+      "step": 14550
+    },
+    {
+      "epoch": 1.3785267941677712,
+      "grad_norm": 0.3163720369338989,
+      "learning_rate": 1.4485892823328916e-05,
+      "loss": 0.1028,
+      "step": 14560
+    },
+    {
+      "epoch": 1.379473584548381,
+      "grad_norm": 0.43609631061553955,
+      "learning_rate": 1.4482105661806478e-05,
+      "loss": 0.1098,
+      "step": 14570
+    },
+    {
+      "epoch": 1.3804203749289907,
+      "grad_norm": 0.364156037569046,
+      "learning_rate": 1.447831850028404e-05,
+      "loss": 0.111,
+      "step": 14580
+    },
+    {
+      "epoch": 1.3813671653096005,
+      "grad_norm": 0.3868848383426666,
+      "learning_rate": 1.44745313387616e-05,
+      "loss": 0.1288,
+      "step": 14590
+    },
+    {
+      "epoch": 1.3823139556902102,
+      "grad_norm": 0.30255648493766785,
+      "learning_rate": 1.447074417723916e-05,
+      "loss": 0.1122,
+      "step": 14600
+    },
+    {
+      "epoch": 1.38326074607082,
+      "grad_norm": 0.4294314682483673,
+      "learning_rate": 1.4466957015716721e-05,
+      "loss": 0.1088,
+      "step": 14610
+    },
+    {
+      "epoch": 1.3842075364514297,
+      "grad_norm": 0.3444744050502777,
+      "learning_rate": 1.4463169854194283e-05,
+      "loss": 0.106,
+      "step": 14620
+    },
+    {
+      "epoch": 1.3851543268320394,
+      "grad_norm": 0.3731893002986908,
+      "learning_rate": 1.4459382692671843e-05,
+      "loss": 0.1074,
+      "step": 14630
+    },
+    {
+      "epoch": 1.3861011172126492,
+      "grad_norm": 0.30317267775535583,
+      "learning_rate": 1.4455595531149405e-05,
+      "loss": 0.1052,
+      "step": 14640
+    },
+    {
+      "epoch": 1.387047907593259,
+      "grad_norm": 0.3629275858402252,
+      "learning_rate": 1.4451808369626965e-05,
+      "loss": 0.1024,
+      "step": 14650
+    },
+    {
+      "epoch": 1.3879946979738687,
+      "grad_norm": 0.3758595883846283,
+      "learning_rate": 1.4448021208104528e-05,
+      "loss": 0.1152,
+      "step": 14660
+    },
+    {
+      "epoch": 1.3889414883544784,
+      "grad_norm": 0.3961845636367798,
+      "learning_rate": 1.4444234046582088e-05,
+      "loss": 0.1112,
+      "step": 14670
+    },
+    {
+      "epoch": 1.3898882787350881,
+      "grad_norm": 0.3658260405063629,
+      "learning_rate": 1.4440446885059648e-05,
+      "loss": 0.1094,
+      "step": 14680
+    },
+    {
+      "epoch": 1.3908350691156977,
+      "grad_norm": 0.3775550425052643,
+      "learning_rate": 1.443665972353721e-05,
+      "loss": 0.1136,
+      "step": 14690
+    },
+    {
+      "epoch": 1.3917818594963074,
+      "grad_norm": 0.30337873101234436,
+      "learning_rate": 1.443287256201477e-05,
+      "loss": 0.1073,
+      "step": 14700
+    },
+    {
+      "epoch": 1.3927286498769171,
+      "grad_norm": 0.36682024598121643,
+      "learning_rate": 1.4429085400492334e-05,
+      "loss": 0.1066,
+      "step": 14710
+    },
+    {
+      "epoch": 1.3936754402575269,
+      "grad_norm": 0.5544394254684448,
+      "learning_rate": 1.4425298238969894e-05,
+      "loss": 0.1134,
+      "step": 14720
+    },
+    {
+      "epoch": 1.3946222306381366,
+      "grad_norm": 0.3843153715133667,
+      "learning_rate": 1.4421511077447454e-05,
+      "loss": 0.1048,
+      "step": 14730
+    },
+    {
+      "epoch": 1.3955690210187464,
+      "grad_norm": 0.40148961544036865,
+      "learning_rate": 1.4417723915925016e-05,
+      "loss": 0.1068,
+      "step": 14740
+    },
+    {
+      "epoch": 1.396515811399356,
+      "grad_norm": 0.30784979462623596,
+      "learning_rate": 1.4413936754402576e-05,
+      "loss": 0.1017,
+      "step": 14750
+    },
+    {
+      "epoch": 1.3974626017799658,
+      "grad_norm": 0.30099502205848694,
+      "learning_rate": 1.4410149592880139e-05,
+      "loss": 0.1049,
+      "step": 14760
+    },
+    {
+      "epoch": 1.3984093921605756,
+      "grad_norm": 0.30565211176872253,
+      "learning_rate": 1.4406362431357699e-05,
+      "loss": 0.1109,
+      "step": 14770
+    },
+    {
+      "epoch": 1.3993561825411853,
+      "grad_norm": 0.33714359998703003,
+      "learning_rate": 1.4402575269835259e-05,
+      "loss": 0.1105,
+      "step": 14780
+    },
+    {
+      "epoch": 1.400302972921795,
+      "grad_norm": 0.420604407787323,
+      "learning_rate": 1.4398788108312821e-05,
+      "loss": 0.0969,
+      "step": 14790
+    },
+    {
+      "epoch": 1.4012497633024048,
+      "grad_norm": 0.32643163204193115,
+      "learning_rate": 1.4395000946790383e-05,
+      "loss": 0.0992,
+      "step": 14800
+    },
+    {
+      "epoch": 1.4021965536830145,
+      "grad_norm": 0.4230499267578125,
+      "learning_rate": 1.4391213785267943e-05,
+      "loss": 0.1153,
+      "step": 14810
+    },
+    {
+      "epoch": 1.4031433440636243,
+      "grad_norm": 0.4465436637401581,
+      "learning_rate": 1.4387426623745504e-05,
+      "loss": 0.1096,
+      "step": 14820
+    },
+    {
+      "epoch": 1.404090134444234,
+      "grad_norm": 0.37599942088127136,
+      "learning_rate": 1.4383639462223064e-05,
+      "loss": 0.1142,
+      "step": 14830
+    },
+    {
+      "epoch": 1.4050369248248438,
+      "grad_norm": 0.340032160282135,
+      "learning_rate": 1.4379852300700625e-05,
+      "loss": 0.106,
+      "step": 14840
+    },
+    {
+      "epoch": 1.4059837152054535,
+      "grad_norm": 0.35222116112709045,
+      "learning_rate": 1.4376065139178188e-05,
+      "loss": 0.1116,
+      "step": 14850
+    },
+    {
+      "epoch": 1.4069305055860633,
+      "grad_norm": 0.3487146496772766,
+      "learning_rate": 1.4372277977655748e-05,
+      "loss": 0.1183,
+      "step": 14860
+    },
+    {
+      "epoch": 1.407877295966673,
+      "grad_norm": 0.3430015742778778,
+      "learning_rate": 1.436849081613331e-05,
+      "loss": 0.1118,
+      "step": 14870
+    },
+    {
+      "epoch": 1.4088240863472827,
+      "grad_norm": 0.2878393828868866,
+      "learning_rate": 1.436470365461087e-05,
+      "loss": 0.106,
+      "step": 14880
+    },
+    {
+      "epoch": 1.4097708767278925,
+      "grad_norm": 0.31863144040107727,
+      "learning_rate": 1.436091649308843e-05,
+      "loss": 0.1075,
+      "step": 14890
+    },
+    {
+      "epoch": 1.4107176671085022,
+      "grad_norm": 0.3831043243408203,
+      "learning_rate": 1.4357129331565993e-05,
+      "loss": 0.1032,
+      "step": 14900
+    },
+    {
+      "epoch": 1.411664457489112,
+      "grad_norm": 0.39178845286369324,
+      "learning_rate": 1.4353342170043553e-05,
+      "loss": 0.1145,
+      "step": 14910
+    },
+    {
+      "epoch": 1.4126112478697217,
+      "grad_norm": 0.34716176986694336,
+      "learning_rate": 1.4349555008521115e-05,
+      "loss": 0.1067,
+      "step": 14920
+    },
+    {
+      "epoch": 1.4135580382503314,
+      "grad_norm": 0.30537331104278564,
+      "learning_rate": 1.4345767846998675e-05,
+      "loss": 0.1123,
+      "step": 14930
+    },
+    {
+      "epoch": 1.4145048286309412,
+      "grad_norm": 0.4346376657485962,
+      "learning_rate": 1.4341980685476239e-05,
+      "loss": 0.1033,
+      "step": 14940
+    },
+    {
+      "epoch": 1.415451619011551,
+      "grad_norm": 0.33927392959594727,
+      "learning_rate": 1.4338193523953799e-05,
+      "loss": 0.1059,
+      "step": 14950
+    },
+    {
+      "epoch": 1.4163984093921607,
+      "grad_norm": 0.3123628795146942,
+      "learning_rate": 1.4334406362431359e-05,
+      "loss": 0.1012,
+      "step": 14960
+    },
+    {
+      "epoch": 1.4173451997727704,
+      "grad_norm": 0.36858904361724854,
+      "learning_rate": 1.433061920090892e-05,
+      "loss": 0.1102,
+      "step": 14970
+    },
+    {
+      "epoch": 1.4182919901533801,
+      "grad_norm": 0.29052573442459106,
+      "learning_rate": 1.432683203938648e-05,
+      "loss": 0.1007,
+      "step": 14980
+    },
+    {
+      "epoch": 1.4192387805339899,
+      "grad_norm": 0.41456088423728943,
+      "learning_rate": 1.4323044877864042e-05,
+      "loss": 0.1134,
+      "step": 14990
+    },
+    {
+      "epoch": 1.4201855709145996,
+      "grad_norm": 0.333921879529953,
+      "learning_rate": 1.4319257716341604e-05,
+      "loss": 0.1013,
+      "step": 15000
+    },
+    {
+      "epoch": 1.4211323612952094,
+      "grad_norm": 0.36147958040237427,
+      "learning_rate": 1.4315470554819164e-05,
+      "loss": 0.1115,
+      "step": 15010
+    },
+    {
+      "epoch": 1.422079151675819,
+      "grad_norm": 0.37708374857902527,
+      "learning_rate": 1.4311683393296724e-05,
+      "loss": 0.1092,
+      "step": 15020
+    },
+    {
+      "epoch": 1.4230259420564286,
+      "grad_norm": 0.3549210727214813,
+      "learning_rate": 1.4307896231774286e-05,
+      "loss": 0.1065,
+      "step": 15030
+    },
+    {
+      "epoch": 1.4239727324370384,
+      "grad_norm": 0.35664400458335876,
+      "learning_rate": 1.4304109070251848e-05,
+      "loss": 0.1124,
+      "step": 15040
+    },
+    {
+      "epoch": 1.424919522817648,
+      "grad_norm": 0.33038046956062317,
+      "learning_rate": 1.430032190872941e-05,
+      "loss": 0.0909,
+      "step": 15050
+    },
+    {
+      "epoch": 1.4258663131982579,
+      "grad_norm": 0.46176019310951233,
+      "learning_rate": 1.429653474720697e-05,
+      "loss": 0.1019,
+      "step": 15060
+    },
+    {
+      "epoch": 1.4268131035788676,
+      "grad_norm": 0.4170028269290924,
+      "learning_rate": 1.429274758568453e-05,
+      "loss": 0.109,
+      "step": 15070
+    },
+    {
+      "epoch": 1.4277598939594773,
+      "grad_norm": 0.42809316515922546,
+      "learning_rate": 1.4288960424162093e-05,
+      "loss": 0.1107,
+      "step": 15080
+    },
+    {
+      "epoch": 1.428706684340087,
+      "grad_norm": 0.2779739499092102,
+      "learning_rate": 1.4285173262639653e-05,
+      "loss": 0.1163,
+      "step": 15090
+    },
+    {
+      "epoch": 1.4296534747206968,
+      "grad_norm": 0.37938955426216125,
+      "learning_rate": 1.4281386101117215e-05,
+      "loss": 0.1082,
+      "step": 15100
+    },
+    {
+      "epoch": 1.4306002651013066,
+      "grad_norm": 0.3991941511631012,
+      "learning_rate": 1.4277598939594775e-05,
+      "loss": 0.1017,
+      "step": 15110
+    },
+    {
+      "epoch": 1.4315470554819163,
+      "grad_norm": 0.31340453028678894,
+      "learning_rate": 1.4273811778072335e-05,
+      "loss": 0.1174,
+      "step": 15120
+    },
+    {
+      "epoch": 1.432493845862526,
+      "grad_norm": 0.5108481645584106,
+      "learning_rate": 1.4270024616549898e-05,
+      "loss": 0.1105,
+      "step": 15130
+    },
+    {
+      "epoch": 1.4334406362431358,
+      "grad_norm": 0.4092690944671631,
+      "learning_rate": 1.4266237455027458e-05,
+      "loss": 0.1144,
+      "step": 15140
+    },
+    {
+      "epoch": 1.4343874266237455,
+      "grad_norm": 0.32388442754745483,
+      "learning_rate": 1.426245029350502e-05,
+      "loss": 0.1156,
+      "step": 15150
+    },
+    {
+      "epoch": 1.4353342170043553,
+      "grad_norm": 0.3596822917461395,
+      "learning_rate": 1.425866313198258e-05,
+      "loss": 0.1048,
+      "step": 15160
+    },
+    {
+      "epoch": 1.436281007384965,
+      "grad_norm": 0.3799607455730438,
+      "learning_rate": 1.425487597046014e-05,
+      "loss": 0.1118,
+      "step": 15170
+    },
+    {
+      "epoch": 1.4372277977655747,
+      "grad_norm": 0.3617545962333679,
+      "learning_rate": 1.4251088808937703e-05,
+      "loss": 0.1067,
+      "step": 15180
+    },
+    {
+      "epoch": 1.4381745881461845,
+      "grad_norm": 0.37102869153022766,
+      "learning_rate": 1.4247301647415264e-05,
+      "loss": 0.1164,
+      "step": 15190
+    },
+    {
+      "epoch": 1.4391213785267942,
+      "grad_norm": 0.347303181886673,
+      "learning_rate": 1.4243514485892824e-05,
+      "loss": 0.1087,
+      "step": 15200
+    },
+    {
+      "epoch": 1.440068168907404,
+      "grad_norm": 0.3193877041339874,
+      "learning_rate": 1.4239727324370385e-05,
+      "loss": 0.1154,
+      "step": 15210
+    },
+    {
+      "epoch": 1.4410149592880137,
+      "grad_norm": 0.502229630947113,
+      "learning_rate": 1.4235940162847947e-05,
+      "loss": 0.1136,
+      "step": 15220
+    },
+    {
+      "epoch": 1.4419617496686234,
+      "grad_norm": 0.4352882504463196,
+      "learning_rate": 1.4232153001325509e-05,
+      "loss": 0.1054,
+      "step": 15230
+    },
+    {
+      "epoch": 1.442908540049233,
+      "grad_norm": 0.3593296706676483,
+      "learning_rate": 1.4228365839803069e-05,
+      "loss": 0.1083,
+      "step": 15240
+    },
+    {
+      "epoch": 1.4438553304298427,
+      "grad_norm": 0.3581804931163788,
+      "learning_rate": 1.4224578678280629e-05,
+      "loss": 0.1136,
+      "step": 15250
+    },
+    {
+      "epoch": 1.4448021208104525,
+      "grad_norm": 0.272034615278244,
+      "learning_rate": 1.422079151675819e-05,
+      "loss": 0.0999,
+      "step": 15260
+    },
+    {
+      "epoch": 1.4457489111910622,
+      "grad_norm": 0.35017022490501404,
+      "learning_rate": 1.4217004355235752e-05,
+      "loss": 0.1154,
+      "step": 15270
+    },
+    {
+      "epoch": 1.446695701571672,
+      "grad_norm": 0.3309600055217743,
+      "learning_rate": 1.4213217193713314e-05,
+      "loss": 0.1033,
+      "step": 15280
+    },
+    {
+      "epoch": 1.4476424919522817,
+      "grad_norm": 0.3215767443180084,
+      "learning_rate": 1.4209430032190874e-05,
+      "loss": 0.1088,
+      "step": 15290
+    },
+    {
+      "epoch": 1.4485892823328914,
+      "grad_norm": 0.3672831356525421,
+      "learning_rate": 1.4205642870668434e-05,
+      "loss": 0.1059,
+      "step": 15300
+    },
+    {
+      "epoch": 1.4495360727135012,
+      "grad_norm": 0.3399512767791748,
+      "learning_rate": 1.4201855709145996e-05,
+      "loss": 0.1054,
+      "step": 15310
+    },
+    {
+      "epoch": 1.450482863094111,
+      "grad_norm": 0.2895975708961487,
+      "learning_rate": 1.4198068547623558e-05,
+      "loss": 0.11,
+      "step": 15320
+    },
+    {
+      "epoch": 1.4514296534747206,
+      "grad_norm": 0.38185954093933105,
+      "learning_rate": 1.419428138610112e-05,
+      "loss": 0.1186,
+      "step": 15330
+    },
+    {
+      "epoch": 1.4523764438553304,
+      "grad_norm": 0.3593434691429138,
+      "learning_rate": 1.419049422457868e-05,
+      "loss": 0.0985,
+      "step": 15340
+    },
+    {
+      "epoch": 1.4533232342359401,
+      "grad_norm": 0.38047701120376587,
+      "learning_rate": 1.418670706305624e-05,
+      "loss": 0.1052,
+      "step": 15350
+    },
+    {
+      "epoch": 1.4542700246165499,
+      "grad_norm": 0.34423530101776123,
+      "learning_rate": 1.4182919901533803e-05,
+      "loss": 0.1071,
+      "step": 15360
+    },
+    {
+      "epoch": 1.4552168149971596,
+      "grad_norm": 0.27197402715682983,
+      "learning_rate": 1.4179132740011363e-05,
+      "loss": 0.1089,
+      "step": 15370
+    },
+    {
+      "epoch": 1.4561636053777693,
+      "grad_norm": 0.3939764201641083,
+      "learning_rate": 1.4175345578488923e-05,
+      "loss": 0.1069,
+      "step": 15380
+    },
+    {
+      "epoch": 1.457110395758379,
+      "grad_norm": 0.30846384167671204,
+      "learning_rate": 1.4171558416966485e-05,
+      "loss": 0.1062,
+      "step": 15390
+    },
+    {
+      "epoch": 1.4580571861389888,
+      "grad_norm": 0.30915161967277527,
+      "learning_rate": 1.4167771255444045e-05,
+      "loss": 0.0926,
+      "step": 15400
+    },
+    {
+      "epoch": 1.4590039765195986,
+      "grad_norm": 0.33449786901474,
+      "learning_rate": 1.4163984093921608e-05,
+      "loss": 0.1106,
+      "step": 15410
+    },
+    {
+      "epoch": 1.4599507669002083,
+      "grad_norm": 0.27424463629722595,
+      "learning_rate": 1.4160196932399168e-05,
+      "loss": 0.1045,
+      "step": 15420
+    },
+    {
+      "epoch": 1.460897557280818,
+      "grad_norm": 0.2804360091686249,
+      "learning_rate": 1.4156409770876728e-05,
+      "loss": 0.1177,
+      "step": 15430
+    },
+    {
+      "epoch": 1.4618443476614278,
+      "grad_norm": 0.35507193207740784,
+      "learning_rate": 1.415262260935429e-05,
+      "loss": 0.1058,
+      "step": 15440
+    },
+    {
+      "epoch": 1.4627911380420375,
+      "grad_norm": 0.3734015226364136,
+      "learning_rate": 1.414883544783185e-05,
+      "loss": 0.1046,
+      "step": 15450
+    },
+    {
+      "epoch": 1.4637379284226473,
+      "grad_norm": 0.29109764099121094,
+      "learning_rate": 1.4145048286309414e-05,
+      "loss": 0.1055,
+      "step": 15460
+    },
+    {
+      "epoch": 1.464684718803257,
+      "grad_norm": 0.3349165916442871,
+      "learning_rate": 1.4141261124786974e-05,
+      "loss": 0.1029,
+      "step": 15470
+    },
+    {
+      "epoch": 1.4656315091838668,
+      "grad_norm": 0.3439183235168457,
+      "learning_rate": 1.4137473963264534e-05,
+      "loss": 0.1088,
+      "step": 15480
+    },
+    {
+      "epoch": 1.4665782995644765,
+      "grad_norm": 0.31033629179000854,
+      "learning_rate": 1.4133686801742095e-05,
+      "loss": 0.1001,
+      "step": 15490
+    },
+    {
+      "epoch": 1.4675250899450862,
+      "grad_norm": 0.3343507945537567,
+      "learning_rate": 1.4129899640219657e-05,
+      "loss": 0.1085,
+      "step": 15500
+    },
+    {
+      "epoch": 1.468471880325696,
+      "grad_norm": 0.3156386911869049,
+      "learning_rate": 1.4126112478697219e-05,
+      "loss": 0.1126,
+      "step": 15510
+    },
+    {
+      "epoch": 1.4694186707063057,
+      "grad_norm": 0.3920190632343292,
+      "learning_rate": 1.4122325317174779e-05,
+      "loss": 0.1108,
+      "step": 15520
+    },
+    {
+      "epoch": 1.4703654610869155,
+      "grad_norm": 0.2956010699272156,
+      "learning_rate": 1.4118538155652339e-05,
+      "loss": 0.0993,
+      "step": 15530
+    },
+    {
+      "epoch": 1.4713122514675252,
+      "grad_norm": 0.363127201795578,
+      "learning_rate": 1.4114750994129899e-05,
+      "loss": 0.0994,
+      "step": 15540
+    },
+    {
+      "epoch": 1.472259041848135,
+      "grad_norm": 0.3261021673679352,
+      "learning_rate": 1.4110963832607463e-05,
+      "loss": 0.1056,
+      "step": 15550
+    },
+    {
+      "epoch": 1.4732058322287447,
+      "grad_norm": 0.37980979681015015,
+      "learning_rate": 1.4107176671085023e-05,
+      "loss": 0.1151,
+      "step": 15560
+    },
+    {
+      "epoch": 1.4741526226093544,
+      "grad_norm": 0.35632893443107605,
+      "learning_rate": 1.4103389509562584e-05,
+      "loss": 0.1067,
+      "step": 15570
+    },
+    {
+      "epoch": 1.475099412989964,
+      "grad_norm": 0.3723986744880676,
+      "learning_rate": 1.4099602348040144e-05,
+      "loss": 0.1052,
+      "step": 15580
+    },
+    {
+      "epoch": 1.4760462033705737,
+      "grad_norm": 0.3155215382575989,
+      "learning_rate": 1.4095815186517704e-05,
+      "loss": 0.1047,
+      "step": 15590
+    },
+    {
+      "epoch": 1.4769929937511834,
+      "grad_norm": 0.36644500494003296,
+      "learning_rate": 1.4092028024995268e-05,
+      "loss": 0.1038,
+      "step": 15600
+    },
+    {
+      "epoch": 1.4779397841317932,
+      "grad_norm": 0.3605925738811493,
+      "learning_rate": 1.4088240863472828e-05,
+      "loss": 0.111,
+      "step": 15610
+    },
+    {
+      "epoch": 1.478886574512403,
+      "grad_norm": 0.39198005199432373,
+      "learning_rate": 1.408445370195039e-05,
+      "loss": 0.1148,
+      "step": 15620
+    },
+    {
+      "epoch": 1.4798333648930126,
+      "grad_norm": 0.36386531591415405,
+      "learning_rate": 1.408066654042795e-05,
+      "loss": 0.1043,
+      "step": 15630
+    },
+    {
+      "epoch": 1.4807801552736224,
+      "grad_norm": 0.44443851709365845,
+      "learning_rate": 1.4076879378905513e-05,
+      "loss": 0.1029,
+      "step": 15640
+    },
+    {
+      "epoch": 1.4817269456542321,
+      "grad_norm": 0.4252612590789795,
+      "learning_rate": 1.4073092217383073e-05,
+      "loss": 0.1065,
+      "step": 15650
+    },
+    {
+      "epoch": 1.4826737360348419,
+      "grad_norm": 0.3278532028198242,
+      "learning_rate": 1.4069305055860633e-05,
+      "loss": 0.1099,
+      "step": 15660
+    },
+    {
+      "epoch": 1.4836205264154516,
+      "grad_norm": 0.30991101264953613,
+      "learning_rate": 1.4065517894338195e-05,
+      "loss": 0.1065,
+      "step": 15670
+    },
+    {
+      "epoch": 1.4845673167960614,
+      "grad_norm": 0.39886489510536194,
+      "learning_rate": 1.4061730732815755e-05,
+      "loss": 0.11,
+      "step": 15680
+    },
+    {
+      "epoch": 1.485514107176671,
+      "grad_norm": 0.3394196331501007,
+      "learning_rate": 1.4057943571293318e-05,
+      "loss": 0.1159,
+      "step": 15690
+    },
+    {
+      "epoch": 1.4864608975572808,
+      "grad_norm": 0.32270580530166626,
+      "learning_rate": 1.4054156409770879e-05,
+      "loss": 0.118,
+      "step": 15700
+    },
+    {
+      "epoch": 1.4874076879378906,
+      "grad_norm": 0.3170892000198364,
+      "learning_rate": 1.4050369248248439e-05,
+      "loss": 0.1064,
+      "step": 15710
+    },
+    {
+      "epoch": 1.4883544783185003,
+      "grad_norm": 0.4652809202671051,
+      "learning_rate": 1.4046582086725999e-05,
+      "loss": 0.1251,
+      "step": 15720
+    },
+    {
+      "epoch": 1.48930126869911,
+      "grad_norm": 0.34130120277404785,
+      "learning_rate": 1.404279492520356e-05,
+      "loss": 0.1066,
+      "step": 15730
+    },
+    {
+      "epoch": 1.4902480590797198,
+      "grad_norm": 0.34349262714385986,
+      "learning_rate": 1.4039007763681122e-05,
+      "loss": 0.1042,
+      "step": 15740
+    },
+    {
+      "epoch": 1.4911948494603295,
+      "grad_norm": 0.38822877407073975,
+      "learning_rate": 1.4035220602158684e-05,
+      "loss": 0.1116,
+      "step": 15750
+    },
+    {
+      "epoch": 1.4921416398409393,
+      "grad_norm": 0.31036290526390076,
+      "learning_rate": 1.4031433440636244e-05,
+      "loss": 0.0986,
+      "step": 15760
+    },
+    {
+      "epoch": 1.493088430221549,
+      "grad_norm": 0.33042672276496887,
+      "learning_rate": 1.4027646279113804e-05,
+      "loss": 0.1002,
+      "step": 15770
+    },
+    {
+      "epoch": 1.4940352206021588,
+      "grad_norm": 0.3388994038105011,
+      "learning_rate": 1.4023859117591367e-05,
+      "loss": 0.1034,
+      "step": 15780
+    },
+    {
+      "epoch": 1.4949820109827683,
+      "grad_norm": 0.44724178314208984,
+      "learning_rate": 1.4020071956068927e-05,
+      "loss": 0.1108,
+      "step": 15790
+    },
+    {
+      "epoch": 1.495928801363378,
+      "grad_norm": 0.358551949262619,
+      "learning_rate": 1.401628479454649e-05,
+      "loss": 0.1043,
+      "step": 15800
+    },
+    {
+      "epoch": 1.4968755917439878,
+      "grad_norm": 0.32363811135292053,
+      "learning_rate": 1.401249763302405e-05,
+      "loss": 0.098,
+      "step": 15810
+    },
+    {
+      "epoch": 1.4978223821245975,
+      "grad_norm": 0.3617938160896301,
+      "learning_rate": 1.400871047150161e-05,
+      "loss": 0.1137,
+      "step": 15820
+    },
+    {
+      "epoch": 1.4987691725052072,
+      "grad_norm": 0.33651456236839294,
+      "learning_rate": 1.4004923309979173e-05,
+      "loss": 0.1052,
+      "step": 15830
+    },
+    {
+      "epoch": 1.499715962885817,
+      "grad_norm": 0.3093239963054657,
+      "learning_rate": 1.4001136148456733e-05,
+      "loss": 0.106,
+      "step": 15840
+    },
+    {
+      "epoch": 1.5006627532664267,
+      "grad_norm": 0.36069008708000183,
+      "learning_rate": 1.3997348986934295e-05,
+      "loss": 0.1059,
+      "step": 15850
+    },
+    {
+      "epoch": 1.5016095436470365,
+      "grad_norm": 0.35160404443740845,
+      "learning_rate": 1.3993561825411855e-05,
+      "loss": 0.1072,
+      "step": 15860
+    },
+    {
+      "epoch": 1.5025563340276462,
+      "grad_norm": 0.3883945345878601,
+      "learning_rate": 1.3989774663889415e-05,
+      "loss": 0.106,
+      "step": 15870
+    },
+    {
+      "epoch": 1.503503124408256,
+      "grad_norm": 0.4188264012336731,
+      "learning_rate": 1.3985987502366978e-05,
+      "loss": 0.1135,
+      "step": 15880
+    },
+    {
+      "epoch": 1.5044499147888657,
+      "grad_norm": 0.3728765547275543,
+      "learning_rate": 1.3982200340844538e-05,
+      "loss": 0.1204,
+      "step": 15890
+    },
+    {
+      "epoch": 1.5053967051694754,
+      "grad_norm": 0.30158689618110657,
+      "learning_rate": 1.3978413179322098e-05,
+      "loss": 0.1085,
+      "step": 15900
+    },
+    {
+      "epoch": 1.5063434955500852,
+      "grad_norm": 0.30330297350883484,
+      "learning_rate": 1.397462601779966e-05,
+      "loss": 0.1121,
+      "step": 15910
+    },
+    {
+      "epoch": 1.507290285930695,
+      "grad_norm": 0.36662396788597107,
+      "learning_rate": 1.3970838856277222e-05,
+      "loss": 0.1103,
+      "step": 15920
+    },
+    {
+      "epoch": 1.5082370763113047,
+      "grad_norm": 0.4485093653202057,
+      "learning_rate": 1.3967051694754783e-05,
+      "loss": 0.1229,
+      "step": 15930
+    },
+    {
+      "epoch": 1.5091838666919144,
+      "grad_norm": 0.34482085704803467,
+      "learning_rate": 1.3963264533232343e-05,
+      "loss": 0.1104,
+      "step": 15940
+    },
+    {
+      "epoch": 1.5101306570725241,
+      "grad_norm": 0.5053472518920898,
+      "learning_rate": 1.3959477371709904e-05,
+      "loss": 0.1076,
+      "step": 15950
+    },
+    {
+      "epoch": 1.5110774474531339,
+      "grad_norm": 0.34632086753845215,
+      "learning_rate": 1.3955690210187465e-05,
+      "loss": 0.0972,
+      "step": 15960
+    },
+    {
+      "epoch": 1.5120242378337436,
+      "grad_norm": 0.3810611367225647,
+      "learning_rate": 1.3951903048665027e-05,
+      "loss": 0.1073,
+      "step": 15970
+    },
+    {
+      "epoch": 1.5129710282143534,
+      "grad_norm": 0.364725798368454,
+      "learning_rate": 1.3948115887142589e-05,
+      "loss": 0.1093,
+      "step": 15980
+    },
+    {
+      "epoch": 1.513917818594963,
+      "grad_norm": 0.32327204942703247,
+      "learning_rate": 1.3944328725620149e-05,
+      "loss": 0.1197,
+      "step": 15990
+    },
+    {
+      "epoch": 1.5148646089755728,
+      "grad_norm": 0.4384889602661133,
+      "learning_rate": 1.3940541564097709e-05,
+      "loss": 0.1057,
+      "step": 16000
+    },
+    {
+      "epoch": 1.5158113993561826,
+      "grad_norm": 0.29192763566970825,
+      "learning_rate": 1.393675440257527e-05,
+      "loss": 0.1043,
+      "step": 16010
+    },
+    {
+      "epoch": 1.5167581897367923,
+      "grad_norm": 0.37460416555404663,
+      "learning_rate": 1.3932967241052832e-05,
+      "loss": 0.1147,
+      "step": 16020
+    },
+    {
+      "epoch": 1.517704980117402,
+      "grad_norm": 0.29063111543655396,
+      "learning_rate": 1.3929180079530394e-05,
+      "loss": 0.0987,
+      "step": 16030
+    },
+    {
+      "epoch": 1.5186517704980118,
+      "grad_norm": 0.3550906777381897,
+      "learning_rate": 1.3925392918007954e-05,
+      "loss": 0.1092,
+      "step": 16040
+    },
+    {
+      "epoch": 1.5195985608786216,
+      "grad_norm": 0.3875524699687958,
+      "learning_rate": 1.3921605756485514e-05,
+      "loss": 0.1118,
+      "step": 16050
+    },
+    {
+      "epoch": 1.5205453512592313,
+      "grad_norm": 0.29305535554885864,
+      "learning_rate": 1.3917818594963078e-05,
+      "loss": 0.1153,
+      "step": 16060
+    },
+    {
+      "epoch": 1.521492141639841,
+      "grad_norm": 0.28514429926872253,
+      "learning_rate": 1.3914031433440638e-05,
+      "loss": 0.11,
+      "step": 16070
+    },
+    {
+      "epoch": 1.5224389320204508,
+      "grad_norm": 0.426382452249527,
+      "learning_rate": 1.3910244271918198e-05,
+      "loss": 0.1287,
+      "step": 16080
+    },
+    {
+      "epoch": 1.5233857224010605,
+      "grad_norm": 0.381534218788147,
+      "learning_rate": 1.390645711039576e-05,
+      "loss": 0.095,
+      "step": 16090
+    },
+    {
+      "epoch": 1.5243325127816703,
+      "grad_norm": 0.3806873559951782,
+      "learning_rate": 1.390266994887332e-05,
+      "loss": 0.1051,
+      "step": 16100
+    },
+    {
+      "epoch": 1.52527930316228,
+      "grad_norm": 0.29320812225341797,
+      "learning_rate": 1.3898882787350883e-05,
+      "loss": 0.109,
+      "step": 16110
+    },
+    {
+      "epoch": 1.5262260935428897,
+      "grad_norm": 0.32169806957244873,
+      "learning_rate": 1.3895095625828443e-05,
+      "loss": 0.0986,
+      "step": 16120
+    },
+    {
+      "epoch": 1.5271728839234995,
+      "grad_norm": 0.40018782019615173,
+      "learning_rate": 1.3891308464306003e-05,
+      "loss": 0.1125,
+      "step": 16130
+    },
+    {
+      "epoch": 1.5281196743041092,
+      "grad_norm": 0.3513498604297638,
+      "learning_rate": 1.3887521302783565e-05,
+      "loss": 0.1134,
+      "step": 16140
+    },
+    {
+      "epoch": 1.529066464684719,
+      "grad_norm": 0.3241359293460846,
+      "learning_rate": 1.3883734141261125e-05,
+      "loss": 0.1014,
+      "step": 16150
+    },
+    {
+      "epoch": 1.5300132550653285,
+      "grad_norm": 0.3017043173313141,
+      "learning_rate": 1.3879946979738688e-05,
+      "loss": 0.1074,
+      "step": 16160
+    },
+    {
+      "epoch": 1.5309600454459382,
+      "grad_norm": 0.3104637563228607,
+      "learning_rate": 1.3876159818216248e-05,
+      "loss": 0.105,
+      "step": 16170
+    },
+    {
+      "epoch": 1.531906835826548,
+      "grad_norm": 0.36349010467529297,
+      "learning_rate": 1.3872372656693808e-05,
+      "loss": 0.1064,
+      "step": 16180
+    },
+    {
+      "epoch": 1.5328536262071577,
+      "grad_norm": 0.35329166054725647,
+      "learning_rate": 1.386858549517137e-05,
+      "loss": 0.1091,
+      "step": 16190
+    },
+    {
+      "epoch": 1.5338004165877674,
+      "grad_norm": 0.36299046874046326,
+      "learning_rate": 1.3864798333648932e-05,
+      "loss": 0.1067,
+      "step": 16200
+    },
+    {
+      "epoch": 1.5347472069683772,
+      "grad_norm": 0.46190503239631653,
+      "learning_rate": 1.3861011172126494e-05,
+      "loss": 0.1032,
+      "step": 16210
+    },
+    {
+      "epoch": 1.535693997348987,
+      "grad_norm": 0.3034611642360687,
+      "learning_rate": 1.3857224010604054e-05,
+      "loss": 0.1112,
+      "step": 16220
+    },
+    {
+      "epoch": 1.5366407877295967,
+      "grad_norm": 0.3905613422393799,
+      "learning_rate": 1.3853436849081614e-05,
+      "loss": 0.1063,
+      "step": 16230
+    },
+    {
+      "epoch": 1.5375875781102064,
+      "grad_norm": 0.34141162037849426,
+      "learning_rate": 1.3849649687559175e-05,
+      "loss": 0.1279,
+      "step": 16240
+    },
+    {
+      "epoch": 1.5385343684908162,
+      "grad_norm": 0.3614260256290436,
+      "learning_rate": 1.3845862526036737e-05,
+      "loss": 0.1103,
+      "step": 16250
+    },
+    {
+      "epoch": 1.539481158871426,
+      "grad_norm": 0.43407753109931946,
+      "learning_rate": 1.3842075364514297e-05,
+      "loss": 0.1082,
+      "step": 16260
+    },
+    {
+      "epoch": 1.5404279492520356,
+      "grad_norm": 0.39427345991134644,
+      "learning_rate": 1.3838288202991859e-05,
+      "loss": 0.1121,
+      "step": 16270
+    },
+    {
+      "epoch": 1.5413747396326454,
+      "grad_norm": 0.34145137667655945,
+      "learning_rate": 1.3834501041469419e-05,
+      "loss": 0.0989,
+      "step": 16280
+    },
+    {
+      "epoch": 1.5423215300132551,
+      "grad_norm": 0.5783584713935852,
+      "learning_rate": 1.3830713879946979e-05,
+      "loss": 0.094,
+      "step": 16290
+    },
+    {
+      "epoch": 1.5432683203938646,
+      "grad_norm": 0.31390732526779175,
+      "learning_rate": 1.3826926718424543e-05,
+      "loss": 0.1118,
+      "step": 16300
+    },
+    {
+      "epoch": 1.5442151107744744,
+      "grad_norm": 0.3206731379032135,
+      "learning_rate": 1.3823139556902103e-05,
+      "loss": 0.1121,
+      "step": 16310
+    },
+    {
+      "epoch": 1.5451619011550841,
+      "grad_norm": 0.47795799374580383,
+      "learning_rate": 1.3819352395379664e-05,
+      "loss": 0.1062,
+      "step": 16320
+    },
+    {
+      "epoch": 1.5461086915356939,
+      "grad_norm": 0.36711663007736206,
+      "learning_rate": 1.3815565233857224e-05,
+      "loss": 0.1116,
+      "step": 16330
+    },
+    {
+      "epoch": 1.5470554819163036,
+      "grad_norm": 0.37463775277137756,
+      "learning_rate": 1.3811778072334788e-05,
+      "loss": 0.104,
+      "step": 16340
+    },
+    {
+      "epoch": 1.5480022722969133,
+      "grad_norm": 0.3962617814540863,
+      "learning_rate": 1.3807990910812348e-05,
+      "loss": 0.1018,
+      "step": 16350
+    },
+    {
+      "epoch": 1.548949062677523,
+      "grad_norm": 0.31192144751548767,
+      "learning_rate": 1.3804203749289908e-05,
+      "loss": 0.1057,
+      "step": 16360
+    },
+    {
+      "epoch": 1.5498958530581328,
+      "grad_norm": 0.3603304326534271,
+      "learning_rate": 1.380041658776747e-05,
+      "loss": 0.1028,
+      "step": 16370
+    },
+    {
+      "epoch": 1.5508426434387426,
+      "grad_norm": 0.35720306634902954,
+      "learning_rate": 1.379662942624503e-05,
+      "loss": 0.1126,
+      "step": 16380
+    },
+    {
+      "epoch": 1.5517894338193523,
+      "grad_norm": 0.3946895897388458,
+      "learning_rate": 1.3792842264722593e-05,
+      "loss": 0.1091,
+      "step": 16390
+    },
+    {
+      "epoch": 1.552736224199962,
+      "grad_norm": 0.3394911587238312,
+      "learning_rate": 1.3789055103200153e-05,
+      "loss": 0.1128,
+      "step": 16400
+    },
+    {
+      "epoch": 1.5536830145805718,
+      "grad_norm": 0.36512428522109985,
+      "learning_rate": 1.3785267941677713e-05,
+      "loss": 0.1059,
+      "step": 16410
+    },
+    {
+      "epoch": 1.5546298049611815,
+      "grad_norm": 0.417980819940567,
+      "learning_rate": 1.3781480780155275e-05,
+      "loss": 0.116,
+      "step": 16420
+    },
+    {
+      "epoch": 1.5555765953417913,
+      "grad_norm": 0.36091163754463196,
+      "learning_rate": 1.3777693618632835e-05,
+      "loss": 0.1078,
+      "step": 16430
+    },
+    {
+      "epoch": 1.556523385722401,
+      "grad_norm": 0.29771095514297485,
+      "learning_rate": 1.3773906457110397e-05,
+      "loss": 0.1051,
+      "step": 16440
+    },
+    {
+      "epoch": 1.5574701761030108,
+      "grad_norm": 0.39906641840934753,
+      "learning_rate": 1.3770119295587958e-05,
+      "loss": 0.1164,
+      "step": 16450
+    },
+    {
+      "epoch": 1.5584169664836205,
+      "grad_norm": 0.35542377829551697,
+      "learning_rate": 1.3766332134065519e-05,
+      "loss": 0.0973,
+      "step": 16460
+    },
+    {
+      "epoch": 1.5593637568642302,
+      "grad_norm": 0.3593170642852783,
+      "learning_rate": 1.3762544972543079e-05,
+      "loss": 0.1127,
+      "step": 16470
+    },
+    {
+      "epoch": 1.56031054724484,
+      "grad_norm": 0.34388262033462524,
+      "learning_rate": 1.3758757811020642e-05,
+      "loss": 0.1107,
+      "step": 16480
+    },
+    {
+      "epoch": 1.5612573376254497,
+      "grad_norm": 0.38229548931121826,
+      "learning_rate": 1.3754970649498202e-05,
+      "loss": 0.1057,
+      "step": 16490
+    },
+    {
+      "epoch": 1.5622041280060595,
+      "grad_norm": 0.38837409019470215,
+      "learning_rate": 1.3751183487975764e-05,
+      "loss": 0.1151,
+      "step": 16500
+    },
+    {
+      "epoch": 1.5631509183866692,
+      "grad_norm": 0.3817042112350464,
+      "learning_rate": 1.3747396326453324e-05,
+      "loss": 0.1076,
+      "step": 16510
+    },
+    {
+      "epoch": 1.564097708767279,
+      "grad_norm": 0.5364170074462891,
+      "learning_rate": 1.3743609164930884e-05,
+      "loss": 0.11,
+      "step": 16520
+    },
+    {
+      "epoch": 1.5650444991478887,
+      "grad_norm": 0.30082473158836365,
+      "learning_rate": 1.3739822003408447e-05,
+      "loss": 0.125,
+      "step": 16530
+    },
+    {
+      "epoch": 1.5659912895284984,
+      "grad_norm": 0.28244081139564514,
+      "learning_rate": 1.3736034841886007e-05,
+      "loss": 0.1131,
+      "step": 16540
+    },
+    {
+      "epoch": 1.5669380799091082,
+      "grad_norm": 0.34583744406700134,
+      "learning_rate": 1.3732247680363569e-05,
+      "loss": 0.0999,
+      "step": 16550
+    },
+    {
+      "epoch": 1.567884870289718,
+      "grad_norm": 0.30678942799568176,
+      "learning_rate": 1.372846051884113e-05,
+      "loss": 0.1166,
+      "step": 16560
+    },
+    {
+      "epoch": 1.5688316606703276,
+      "grad_norm": 0.3636748194694519,
+      "learning_rate": 1.372467335731869e-05,
+      "loss": 0.1055,
+      "step": 16570
+    },
+    {
+      "epoch": 1.5697784510509374,
+      "grad_norm": 0.4010995030403137,
+      "learning_rate": 1.3720886195796253e-05,
+      "loss": 0.1131,
+      "step": 16580
+    },
+    {
+      "epoch": 1.5707252414315471,
+      "grad_norm": 0.43221795558929443,
+      "learning_rate": 1.3717099034273813e-05,
+      "loss": 0.1149,
+      "step": 16590
+    },
+    {
+      "epoch": 1.5716720318121569,
+      "grad_norm": 0.3476791977882385,
+      "learning_rate": 1.3713311872751374e-05,
+      "loss": 0.1044,
+      "step": 16600
+    },
+    {
+      "epoch": 1.5726188221927666,
+      "grad_norm": 0.36815622448921204,
+      "learning_rate": 1.3709524711228935e-05,
+      "loss": 0.1139,
+      "step": 16610
+    },
+    {
+      "epoch": 1.5735656125733763,
+      "grad_norm": 0.42477601766586304,
+      "learning_rate": 1.3705737549706496e-05,
+      "loss": 0.1132,
+      "step": 16620
+    },
+    {
+      "epoch": 1.574512402953986,
+      "grad_norm": 0.35573610663414,
+      "learning_rate": 1.3701950388184058e-05,
+      "loss": 0.1137,
+      "step": 16630
+    },
+    {
+      "epoch": 1.5754591933345958,
+      "grad_norm": 0.38760095834732056,
+      "learning_rate": 1.3698163226661618e-05,
+      "loss": 0.114,
+      "step": 16640
+    },
+    {
+      "epoch": 1.5764059837152056,
+      "grad_norm": 0.34179815649986267,
+      "learning_rate": 1.3694376065139178e-05,
+      "loss": 0.1087,
+      "step": 16650
+    },
+    {
+      "epoch": 1.5773527740958153,
+      "grad_norm": 0.3777365982532501,
+      "learning_rate": 1.369058890361674e-05,
+      "loss": 0.118,
+      "step": 16660
+    },
+    {
+      "epoch": 1.578299564476425,
+      "grad_norm": 0.3674735724925995,
+      "learning_rate": 1.3686801742094302e-05,
+      "loss": 0.1049,
+      "step": 16670
+    },
+    {
+      "epoch": 1.5792463548570348,
+      "grad_norm": 0.3519558608531952,
+      "learning_rate": 1.3683014580571863e-05,
+      "loss": 0.1135,
+      "step": 16680
+    },
+    {
+      "epoch": 1.5801931452376445,
+      "grad_norm": 0.3152370750904083,
+      "learning_rate": 1.3679227419049423e-05,
+      "loss": 0.1034,
+      "step": 16690
+    },
+    {
+      "epoch": 1.5811399356182543,
+      "grad_norm": 0.40113574266433716,
+      "learning_rate": 1.3675440257526983e-05,
+      "loss": 0.111,
+      "step": 16700
+    },
+    {
+      "epoch": 1.5820867259988638,
+      "grad_norm": 0.3402666747570038,
+      "learning_rate": 1.3671653096004547e-05,
+      "loss": 0.1102,
+      "step": 16710
+    },
+    {
+      "epoch": 1.5830335163794735,
+      "grad_norm": 0.33529624342918396,
+      "learning_rate": 1.3667865934482107e-05,
+      "loss": 0.1112,
+      "step": 16720
+    },
+    {
+      "epoch": 1.5839803067600833,
+      "grad_norm": 0.30931976437568665,
+      "learning_rate": 1.3664078772959669e-05,
+      "loss": 0.1048,
+      "step": 16730
+    },
+    {
+      "epoch": 1.584927097140693,
+      "grad_norm": 0.3059307932853699,
+      "learning_rate": 1.3660291611437229e-05,
+      "loss": 0.1111,
+      "step": 16740
+    },
+    {
+      "epoch": 1.5858738875213028,
+      "grad_norm": 0.41115766763687134,
+      "learning_rate": 1.3656504449914789e-05,
+      "loss": 0.11,
+      "step": 16750
+    },
+    {
+      "epoch": 1.5868206779019125,
+      "grad_norm": 0.3679191768169403,
+      "learning_rate": 1.3652717288392352e-05,
+      "loss": 0.1011,
+      "step": 16760
+    },
+    {
+      "epoch": 1.5877674682825222,
+      "grad_norm": 0.32862117886543274,
+      "learning_rate": 1.3648930126869912e-05,
+      "loss": 0.1108,
+      "step": 16770
+    },
+    {
+      "epoch": 1.588714258663132,
+      "grad_norm": 0.25885462760925293,
+      "learning_rate": 1.3645142965347474e-05,
+      "loss": 0.1102,
+      "step": 16780
+    },
+    {
+      "epoch": 1.5896610490437417,
+      "grad_norm": 0.401241272687912,
+      "learning_rate": 1.3641355803825034e-05,
+      "loss": 0.1099,
+      "step": 16790
+    },
+    {
+      "epoch": 1.5906078394243515,
+      "grad_norm": 0.3936287760734558,
+      "learning_rate": 1.3637568642302594e-05,
+      "loss": 0.1048,
+      "step": 16800
+    },
+    {
+      "epoch": 1.5915546298049612,
+      "grad_norm": 0.41480880975723267,
+      "learning_rate": 1.3633781480780158e-05,
+      "loss": 0.1053,
+      "step": 16810
+    },
+    {
+      "epoch": 1.592501420185571,
+      "grad_norm": 0.42180556058883667,
+      "learning_rate": 1.3629994319257718e-05,
+      "loss": 0.1123,
+      "step": 16820
+    },
+    {
+      "epoch": 1.5934482105661807,
+      "grad_norm": 0.3735913634300232,
+      "learning_rate": 1.3626207157735278e-05,
+      "loss": 0.1139,
+      "step": 16830
+    },
+    {
+      "epoch": 1.5943950009467904,
+      "grad_norm": 0.38873815536499023,
+      "learning_rate": 1.362241999621284e-05,
+      "loss": 0.1036,
+      "step": 16840
+    },
+    {
+      "epoch": 1.5953417913274,
+      "grad_norm": 0.3312840461730957,
+      "learning_rate": 1.3618632834690401e-05,
+      "loss": 0.1036,
+      "step": 16850
+    },
+    {
+      "epoch": 1.5962885817080097,
+      "grad_norm": 0.344212144613266,
+      "learning_rate": 1.3614845673167963e-05,
+      "loss": 0.1139,
+      "step": 16860
+    },
+    {
+      "epoch": 1.5972353720886194,
+      "grad_norm": 0.3515395224094391,
+      "learning_rate": 1.3611058511645523e-05,
+      "loss": 0.1082,
+      "step": 16870
+    },
+    {
+      "epoch": 1.5981821624692292,
+      "grad_norm": 0.30493828654289246,
+      "learning_rate": 1.3607271350123083e-05,
+      "loss": 0.1117,
+      "step": 16880
+    },
+    {
+      "epoch": 1.599128952849839,
+      "grad_norm": 0.29671570658683777,
+      "learning_rate": 1.3603484188600645e-05,
+      "loss": 0.104,
+      "step": 16890
+    },
+    {
+      "epoch": 1.6000757432304487,
+      "grad_norm": 0.3543199598789215,
+      "learning_rate": 1.3599697027078206e-05,
+      "loss": 0.1084,
+      "step": 16900
+    },
+    {
+      "epoch": 1.6010225336110584,
+      "grad_norm": 0.4090178310871124,
+      "learning_rate": 1.3595909865555768e-05,
+      "loss": 0.1056,
+      "step": 16910
+    },
+    {
+      "epoch": 1.6019693239916681,
+      "grad_norm": 0.35988327860832214,
+      "learning_rate": 1.3592122704033328e-05,
+      "loss": 0.1055,
+      "step": 16920
+    },
+    {
+      "epoch": 1.6029161143722779,
+      "grad_norm": 0.4046424925327301,
+      "learning_rate": 1.3588335542510888e-05,
+      "loss": 0.1097,
+      "step": 16930
+    },
+    {
+      "epoch": 1.6038629047528876,
+      "grad_norm": 0.330355167388916,
+      "learning_rate": 1.358454838098845e-05,
+      "loss": 0.1104,
+      "step": 16940
+    },
+    {
+      "epoch": 1.6048096951334974,
+      "grad_norm": 0.3554498553276062,
+      "learning_rate": 1.3580761219466012e-05,
+      "loss": 0.0995,
+      "step": 16950
+    },
+    {
+      "epoch": 1.605756485514107,
+      "grad_norm": 0.3867628872394562,
+      "learning_rate": 1.3576974057943574e-05,
+      "loss": 0.1146,
+      "step": 16960
+    },
+    {
+      "epoch": 1.6067032758947168,
+      "grad_norm": 0.3537595272064209,
+      "learning_rate": 1.3573186896421134e-05,
+      "loss": 0.1039,
+      "step": 16970
+    },
+    {
+      "epoch": 1.6076500662753266,
+      "grad_norm": 0.33180224895477295,
+      "learning_rate": 1.3569399734898694e-05,
+      "loss": 0.1067,
+      "step": 16980
+    },
+    {
+      "epoch": 1.6085968566559363,
+      "grad_norm": 0.36756113171577454,
+      "learning_rate": 1.3565612573376257e-05,
+      "loss": 0.107,
+      "step": 16990
+    },
+    {
+      "epoch": 1.609543647036546,
+      "grad_norm": 0.33853188157081604,
+      "learning_rate": 1.3561825411853817e-05,
+      "loss": 0.1221,
+      "step": 17000
+    },
+    {
+      "epoch": 1.6104904374171558,
+      "grad_norm": 0.2996186316013336,
+      "learning_rate": 1.3558038250331377e-05,
+      "loss": 0.1043,
+      "step": 17010
+    },
+    {
+      "epoch": 1.6114372277977655,
+      "grad_norm": 0.3746570944786072,
+      "learning_rate": 1.3554251088808939e-05,
+      "loss": 0.1121,
+      "step": 17020
+    },
+    {
+      "epoch": 1.6123840181783753,
+      "grad_norm": 0.3464304804801941,
+      "learning_rate": 1.3550463927286499e-05,
+      "loss": 0.1071,
+      "step": 17030
+    },
+    {
+      "epoch": 1.613330808558985,
+      "grad_norm": 0.3381328284740448,
+      "learning_rate": 1.3546676765764062e-05,
+      "loss": 0.1097,
+      "step": 17040
+    },
+    {
+      "epoch": 1.6142775989395948,
+      "grad_norm": 0.3596116304397583,
+      "learning_rate": 1.3542889604241622e-05,
+      "loss": 0.106,
+      "step": 17050
+    },
+    {
+      "epoch": 1.6152243893202045,
+      "grad_norm": 0.3500080406665802,
+      "learning_rate": 1.3539102442719182e-05,
+      "loss": 0.101,
+      "step": 17060
+    },
+    {
+      "epoch": 1.6161711797008143,
+      "grad_norm": 0.30177420377731323,
+      "learning_rate": 1.3535315281196744e-05,
+      "loss": 0.1022,
+      "step": 17070
+    },
+    {
+      "epoch": 1.617117970081424,
+      "grad_norm": 0.3444776237010956,
+      "learning_rate": 1.3531528119674304e-05,
+      "loss": 0.1164,
+      "step": 17080
+    },
+    {
+      "epoch": 1.6180647604620337,
+      "grad_norm": 0.3821253180503845,
+      "learning_rate": 1.3527740958151868e-05,
+      "loss": 0.1183,
+      "step": 17090
+    },
+    {
+      "epoch": 1.6190115508426435,
+      "grad_norm": 0.39384153485298157,
+      "learning_rate": 1.3523953796629428e-05,
+      "loss": 0.0943,
+      "step": 17100
+    },
+    {
+      "epoch": 1.6199583412232532,
+      "grad_norm": 0.4101506769657135,
+      "learning_rate": 1.3520166635106988e-05,
+      "loss": 0.1035,
+      "step": 17110
+    },
+    {
+      "epoch": 1.620905131603863,
+      "grad_norm": 0.4028509557247162,
+      "learning_rate": 1.351637947358455e-05,
+      "loss": 0.1063,
+      "step": 17120
+    },
+    {
+      "epoch": 1.6218519219844727,
+      "grad_norm": 0.33712711930274963,
+      "learning_rate": 1.3512592312062111e-05,
+      "loss": 0.1069,
+      "step": 17130
+    },
+    {
+      "epoch": 1.6227987123650824,
+      "grad_norm": 0.3235998749732971,
+      "learning_rate": 1.3508805150539673e-05,
+      "loss": 0.1101,
+      "step": 17140
+    },
+    {
+      "epoch": 1.6237455027456922,
+      "grad_norm": 0.33888810873031616,
+      "learning_rate": 1.3505017989017233e-05,
+      "loss": 0.1136,
+      "step": 17150
+    },
+    {
+      "epoch": 1.624692293126302,
+      "grad_norm": 0.351136714220047,
+      "learning_rate": 1.3501230827494793e-05,
+      "loss": 0.109,
+      "step": 17160
+    },
+    {
+      "epoch": 1.6256390835069117,
+      "grad_norm": 0.3584769070148468,
+      "learning_rate": 1.3497443665972353e-05,
+      "loss": 0.108,
+      "step": 17170
+    },
+    {
+      "epoch": 1.6265858738875214,
+      "grad_norm": 0.39292648434638977,
+      "learning_rate": 1.3493656504449917e-05,
+      "loss": 0.1024,
+      "step": 17180
+    },
+    {
+      "epoch": 1.6275326642681311,
+      "grad_norm": 0.33220037817955017,
+      "learning_rate": 1.3489869342927477e-05,
+      "loss": 0.1042,
+      "step": 17190
+    },
+    {
+      "epoch": 1.6284794546487409,
+      "grad_norm": 0.3405236303806305,
+      "learning_rate": 1.3486082181405038e-05,
+      "loss": 0.1156,
+      "step": 17200
+    },
+    {
+      "epoch": 1.6294262450293506,
+      "grad_norm": 0.31284892559051514,
+      "learning_rate": 1.3482295019882598e-05,
+      "loss": 0.1054,
+      "step": 17210
+    },
+    {
+      "epoch": 1.6303730354099604,
+      "grad_norm": 0.34259989857673645,
+      "learning_rate": 1.3478507858360159e-05,
+      "loss": 0.1017,
+      "step": 17220
+    },
+    {
+      "epoch": 1.63131982579057,
+      "grad_norm": 0.37608641386032104,
+      "learning_rate": 1.3474720696837722e-05,
+      "loss": 0.1173,
+      "step": 17230
+    },
+    {
+      "epoch": 1.6322666161711799,
+      "grad_norm": 0.38453209400177,
+      "learning_rate": 1.3470933535315282e-05,
+      "loss": 0.1049,
+      "step": 17240
+    },
+    {
+      "epoch": 1.6332134065517896,
+      "grad_norm": 0.27649539709091187,
+      "learning_rate": 1.3467146373792844e-05,
+      "loss": 0.1031,
+      "step": 17250
+    },
+    {
+      "epoch": 1.6341601969323991,
+      "grad_norm": 0.30744364857673645,
+      "learning_rate": 1.3463359212270404e-05,
+      "loss": 0.1038,
+      "step": 17260
+    },
+    {
+      "epoch": 1.6351069873130089,
+      "grad_norm": 0.30624982714653015,
+      "learning_rate": 1.3459572050747967e-05,
+      "loss": 0.1157,
+      "step": 17270
+    },
+    {
+      "epoch": 1.6360537776936186,
+      "grad_norm": 0.40259572863578796,
+      "learning_rate": 1.3455784889225527e-05,
+      "loss": 0.1124,
+      "step": 17280
+    },
+    {
+      "epoch": 1.6370005680742283,
+      "grad_norm": 0.3984117805957794,
+      "learning_rate": 1.3451997727703087e-05,
+      "loss": 0.1047,
+      "step": 17290
+    },
+    {
+      "epoch": 1.637947358454838,
+      "grad_norm": 0.34300553798675537,
+      "learning_rate": 1.3448210566180649e-05,
+      "loss": 0.1167,
+      "step": 17300
+    },
+    {
+      "epoch": 1.6388941488354478,
+      "grad_norm": 0.3874812424182892,
+      "learning_rate": 1.3444423404658209e-05,
+      "loss": 0.0996,
+      "step": 17310
+    },
+    {
+      "epoch": 1.6398409392160576,
+      "grad_norm": 0.31231194734573364,
+      "learning_rate": 1.3440636243135773e-05,
+      "loss": 0.1049,
+      "step": 17320
+    },
+    {
+      "epoch": 1.6407877295966673,
+      "grad_norm": 0.35242578387260437,
+      "learning_rate": 1.3436849081613333e-05,
+      "loss": 0.1014,
+      "step": 17330
+    },
+    {
+      "epoch": 1.641734519977277,
+      "grad_norm": 0.3678164482116699,
+      "learning_rate": 1.3433061920090893e-05,
+      "loss": 0.1019,
+      "step": 17340
+    },
+    {
+      "epoch": 1.6426813103578868,
+      "grad_norm": 0.500481903553009,
+      "learning_rate": 1.3429274758568453e-05,
+      "loss": 0.111,
+      "step": 17350
+    },
+    {
+      "epoch": 1.6436281007384965,
+      "grad_norm": 0.35619476437568665,
+      "learning_rate": 1.3425487597046014e-05,
+      "loss": 0.1058,
+      "step": 17360
+    },
+    {
+      "epoch": 1.6445748911191063,
+      "grad_norm": 0.28489652276039124,
+      "learning_rate": 1.3421700435523576e-05,
+      "loss": 0.107,
+      "step": 17370
+    },
+    {
+      "epoch": 1.645521681499716,
+      "grad_norm": 0.3987047076225281,
+      "learning_rate": 1.3417913274001138e-05,
+      "loss": 0.1102,
+      "step": 17380
+    },
+    {
+      "epoch": 1.6464684718803257,
+      "grad_norm": 0.29614952206611633,
+      "learning_rate": 1.3414126112478698e-05,
+      "loss": 0.1016,
+      "step": 17390
+    },
+    {
+      "epoch": 1.6474152622609353,
+      "grad_norm": 0.4422844648361206,
+      "learning_rate": 1.3410338950956258e-05,
+      "loss": 0.1109,
+      "step": 17400
+    },
+    {
+      "epoch": 1.648362052641545,
+      "grad_norm": 0.485542356967926,
+      "learning_rate": 1.3406551789433821e-05,
+      "loss": 0.125,
+      "step": 17410
+    },
+    {
+      "epoch": 1.6493088430221547,
+      "grad_norm": 0.3143191337585449,
+      "learning_rate": 1.3402764627911382e-05,
+      "loss": 0.0982,
+      "step": 17420
+    },
+    {
+      "epoch": 1.6502556334027645,
+      "grad_norm": 0.34800443053245544,
+      "learning_rate": 1.3398977466388943e-05,
+      "loss": 0.1222,
+      "step": 17430
+    },
+    {
+      "epoch": 1.6512024237833742,
+      "grad_norm": 0.4274339973926544,
+      "learning_rate": 1.3395190304866503e-05,
+      "loss": 0.1128,
+      "step": 17440
+    },
+    {
+      "epoch": 1.652149214163984,
+      "grad_norm": 0.35616034269332886,
+      "learning_rate": 1.3391403143344063e-05,
+      "loss": 0.1101,
+      "step": 17450
+    },
+    {
+      "epoch": 1.6530960045445937,
+      "grad_norm": 0.34067031741142273,
+      "learning_rate": 1.3387615981821627e-05,
+      "loss": 0.1077,
+      "step": 17460
+    },
+    {
+      "epoch": 1.6540427949252035,
+      "grad_norm": 0.367699533700943,
+      "learning_rate": 1.3383828820299187e-05,
+      "loss": 0.1086,
+      "step": 17470
+    },
+    {
+      "epoch": 1.6549895853058132,
+      "grad_norm": 0.34683769941329956,
+      "learning_rate": 1.3380041658776749e-05,
+      "loss": 0.1103,
+      "step": 17480
+    },
+    {
+      "epoch": 1.655936375686423,
+      "grad_norm": 0.31630000472068787,
+      "learning_rate": 1.3376254497254309e-05,
+      "loss": 0.1099,
+      "step": 17490
+    },
+    {
+      "epoch": 1.6568831660670327,
+      "grad_norm": 0.30700427293777466,
+      "learning_rate": 1.3372467335731869e-05,
+      "loss": 0.1102,
+      "step": 17500
+    },
+    {
+      "epoch": 1.6578299564476424,
+      "grad_norm": 0.3312118351459503,
+      "learning_rate": 1.3368680174209432e-05,
+      "loss": 0.109,
+      "step": 17510
+    },
+    {
+      "epoch": 1.6587767468282522,
+      "grad_norm": 0.3572567403316498,
+      "learning_rate": 1.3364893012686992e-05,
+      "loss": 0.1098,
+      "step": 17520
+    },
+    {
+      "epoch": 1.659723537208862,
+      "grad_norm": 0.35258644819259644,
+      "learning_rate": 1.3361105851164552e-05,
+      "loss": 0.1026,
+      "step": 17530
+    },
+    {
+      "epoch": 1.6606703275894716,
+      "grad_norm": 0.4267875552177429,
+      "learning_rate": 1.3357318689642114e-05,
+      "loss": 0.1126,
+      "step": 17540
+    },
+    {
+      "epoch": 1.6616171179700814,
+      "grad_norm": 0.4066093862056732,
+      "learning_rate": 1.3353531528119676e-05,
+      "loss": 0.1037,
+      "step": 17550
+    },
+    {
+      "epoch": 1.6625639083506911,
+      "grad_norm": 0.39801302552223206,
+      "learning_rate": 1.3349744366597237e-05,
+      "loss": 0.1166,
+      "step": 17560
+    },
+    {
+      "epoch": 1.6635106987313009,
+      "grad_norm": 0.3564015030860901,
+      "learning_rate": 1.3345957205074798e-05,
+      "loss": 0.1115,
+      "step": 17570
+    },
+    {
+      "epoch": 1.6644574891119106,
+      "grad_norm": 0.3493454158306122,
+      "learning_rate": 1.3342170043552358e-05,
+      "loss": 0.1121,
+      "step": 17580
+    },
+    {
+      "epoch": 1.6654042794925203,
+      "grad_norm": 0.33144834637641907,
+      "learning_rate": 1.333838288202992e-05,
+      "loss": 0.1018,
+      "step": 17590
+    },
+    {
+      "epoch": 1.66635106987313,
+      "grad_norm": 0.295213907957077,
+      "learning_rate": 1.3334595720507481e-05,
+      "loss": 0.1048,
+      "step": 17600
+    },
+    {
+      "epoch": 1.6672978602537398,
+      "grad_norm": 0.35501471161842346,
+      "learning_rate": 1.3330808558985043e-05,
+      "loss": 0.1038,
+      "step": 17610
+    },
+    {
+      "epoch": 1.6682446506343496,
+      "grad_norm": 0.373725026845932,
+      "learning_rate": 1.3327021397462603e-05,
+      "loss": 0.1049,
+      "step": 17620
+    },
+    {
+      "epoch": 1.6691914410149593,
+      "grad_norm": 0.35085785388946533,
+      "learning_rate": 1.3323234235940163e-05,
+      "loss": 0.1079,
+      "step": 17630
+    },
+    {
+      "epoch": 1.670138231395569,
+      "grad_norm": 0.28753459453582764,
+      "learning_rate": 1.3319447074417725e-05,
+      "loss": 0.0992,
+      "step": 17640
+    },
+    {
+      "epoch": 1.6710850217761788,
+      "grad_norm": 0.3040468990802765,
+      "learning_rate": 1.3315659912895286e-05,
+      "loss": 0.1077,
+      "step": 17650
+    },
+    {
+      "epoch": 1.6720318121567885,
+      "grad_norm": 0.2903306782245636,
+      "learning_rate": 1.3311872751372848e-05,
+      "loss": 0.1025,
+      "step": 17660
+    },
+    {
+      "epoch": 1.6729786025373983,
+      "grad_norm": 0.377331018447876,
+      "learning_rate": 1.3308085589850408e-05,
+      "loss": 0.1116,
+      "step": 17670
+    },
+    {
+      "epoch": 1.673925392918008,
+      "grad_norm": 0.3333212435245514,
+      "learning_rate": 1.3304298428327968e-05,
+      "loss": 0.1141,
+      "step": 17680
+    },
+    {
+      "epoch": 1.6748721832986178,
+      "grad_norm": 0.342886358499527,
+      "learning_rate": 1.3300511266805532e-05,
+      "loss": 0.1043,
+      "step": 17690
+    },
+    {
+      "epoch": 1.6758189736792275,
+      "grad_norm": 0.3984515368938446,
+      "learning_rate": 1.3296724105283092e-05,
+      "loss": 0.1001,
+      "step": 17700
+    },
+    {
+      "epoch": 1.6767657640598372,
+      "grad_norm": 0.32483118772506714,
+      "learning_rate": 1.3292936943760652e-05,
+      "loss": 0.1174,
+      "step": 17710
+    },
+    {
+      "epoch": 1.677712554440447,
+      "grad_norm": 0.4028964638710022,
+      "learning_rate": 1.3289149782238214e-05,
+      "loss": 0.1056,
+      "step": 17720
+    },
+    {
+      "epoch": 1.6786593448210567,
+      "grad_norm": 0.5397863984107971,
+      "learning_rate": 1.3285362620715774e-05,
+      "loss": 0.1081,
+      "step": 17730
+    },
+    {
+      "epoch": 1.6796061352016665,
+      "grad_norm": 0.34457170963287354,
+      "learning_rate": 1.3281575459193337e-05,
+      "loss": 0.1088,
+      "step": 17740
+    },
+    {
+      "epoch": 1.6805529255822762,
+      "grad_norm": 0.3310573101043701,
+      "learning_rate": 1.3277788297670897e-05,
+      "loss": 0.107,
+      "step": 17750
+    },
+    {
+      "epoch": 1.681499715962886,
+      "grad_norm": 0.475350558757782,
+      "learning_rate": 1.3274001136148457e-05,
+      "loss": 0.1054,
+      "step": 17760
+    },
+    {
+      "epoch": 1.6824465063434957,
+      "grad_norm": 0.32560041546821594,
+      "learning_rate": 1.3270213974626019e-05,
+      "loss": 0.0932,
+      "step": 17770
+    },
+    {
+      "epoch": 1.6833932967241054,
+      "grad_norm": 0.2892504930496216,
+      "learning_rate": 1.3266426813103579e-05,
+      "loss": 0.107,
+      "step": 17780
+    },
+    {
+      "epoch": 1.6843400871047152,
+      "grad_norm": 0.4393444359302521,
+      "learning_rate": 1.3262639651581142e-05,
+      "loss": 0.103,
+      "step": 17790
+    },
+    {
+      "epoch": 1.685286877485325,
+      "grad_norm": 0.3609474003314972,
+      "learning_rate": 1.3258852490058702e-05,
+      "loss": 0.1035,
+      "step": 17800
+    },
+    {
+      "epoch": 1.6862336678659344,
+      "grad_norm": 0.3656294643878937,
+      "learning_rate": 1.3255065328536262e-05,
+      "loss": 0.1054,
+      "step": 17810
+    },
+    {
+      "epoch": 1.6871804582465442,
+      "grad_norm": 0.32531723380088806,
+      "learning_rate": 1.3251278167013824e-05,
+      "loss": 0.1109,
+      "step": 17820
+    },
+    {
+      "epoch": 1.688127248627154,
+      "grad_norm": 0.42885199189186096,
+      "learning_rate": 1.3247491005491386e-05,
+      "loss": 0.1159,
+      "step": 17830
+    },
+    {
+      "epoch": 1.6890740390077637,
+      "grad_norm": 0.34545162320137024,
+      "learning_rate": 1.3243703843968948e-05,
+      "loss": 0.1137,
+      "step": 17840
+    },
+    {
+      "epoch": 1.6900208293883734,
+      "grad_norm": 0.45957455039024353,
+      "learning_rate": 1.3239916682446508e-05,
+      "loss": 0.1126,
+      "step": 17850
+    },
+    {
+      "epoch": 1.6909676197689831,
+      "grad_norm": 0.3322799503803253,
+      "learning_rate": 1.3236129520924068e-05,
+      "loss": 0.113,
+      "step": 17860
+    },
+    {
+      "epoch": 1.6919144101495929,
+      "grad_norm": 0.4249228239059448,
+      "learning_rate": 1.323234235940163e-05,
+      "loss": 0.1114,
+      "step": 17870
+    },
+    {
+      "epoch": 1.6928612005302026,
+      "grad_norm": 0.3494950830936432,
+      "learning_rate": 1.3228555197879191e-05,
+      "loss": 0.1083,
+      "step": 17880
+    },
+    {
+      "epoch": 1.6938079909108124,
+      "grad_norm": 0.34661203622817993,
+      "learning_rate": 1.3224768036356751e-05,
+      "loss": 0.108,
+      "step": 17890
+    },
+    {
+      "epoch": 1.694754781291422,
+      "grad_norm": 0.35037049651145935,
+      "learning_rate": 1.3220980874834313e-05,
+      "loss": 0.1079,
+      "step": 17900
+    },
+    {
+      "epoch": 1.6957015716720318,
+      "grad_norm": 0.3720286190509796,
+      "learning_rate": 1.3217193713311873e-05,
+      "loss": 0.1062,
+      "step": 17910
+    },
+    {
+      "epoch": 1.6966483620526416,
+      "grad_norm": 0.37974655628204346,
+      "learning_rate": 1.3213406551789433e-05,
+      "loss": 0.0993,
+      "step": 17920
+    },
+    {
+      "epoch": 1.6975951524332513,
+      "grad_norm": 0.28876063227653503,
+      "learning_rate": 1.3209619390266997e-05,
+      "loss": 0.0907,
+      "step": 17930
+    },
+    {
+      "epoch": 1.698541942813861,
+      "grad_norm": 0.33278247714042664,
+      "learning_rate": 1.3205832228744557e-05,
+      "loss": 0.1087,
+      "step": 17940
+    },
+    {
+      "epoch": 1.6994887331944706,
+      "grad_norm": 0.2906360924243927,
+      "learning_rate": 1.3202045067222118e-05,
+      "loss": 0.1123,
+      "step": 17950
+    },
+    {
+      "epoch": 1.7004355235750803,
+      "grad_norm": 0.32706132531166077,
+      "learning_rate": 1.3198257905699678e-05,
+      "loss": 0.1011,
+      "step": 17960
+    },
+    {
+      "epoch": 1.70138231395569,
+      "grad_norm": 0.3792358934879303,
+      "learning_rate": 1.3194470744177242e-05,
+      "loss": 0.104,
+      "step": 17970
+    },
+    {
+      "epoch": 1.7023291043362998,
+      "grad_norm": 0.5245835185050964,
+      "learning_rate": 1.3190683582654802e-05,
+      "loss": 0.1009,
+      "step": 17980
+    },
+    {
+      "epoch": 1.7032758947169095,
+      "grad_norm": 0.3694629669189453,
+      "learning_rate": 1.3186896421132362e-05,
+      "loss": 0.1104,
+      "step": 17990
+    },
+    {
+      "epoch": 1.7042226850975193,
+      "grad_norm": 0.27885186672210693,
+      "learning_rate": 1.3183109259609924e-05,
+      "loss": 0.1059,
+      "step": 18000
+    },
+    {
+      "epoch": 1.705169475478129,
+      "grad_norm": 0.5485479831695557,
+      "learning_rate": 1.3179322098087484e-05,
+      "loss": 0.115,
+      "step": 18010
+    },
+    {
+      "epoch": 1.7061162658587388,
+      "grad_norm": 0.4096299707889557,
+      "learning_rate": 1.3175534936565047e-05,
+      "loss": 0.1192,
+      "step": 18020
+    },
+    {
+      "epoch": 1.7070630562393485,
+      "grad_norm": 0.3202039301395416,
+      "learning_rate": 1.3171747775042607e-05,
+      "loss": 0.0996,
+      "step": 18030
+    },
+    {
+      "epoch": 1.7080098466199582,
+      "grad_norm": 0.6081573367118835,
+      "learning_rate": 1.3167960613520167e-05,
+      "loss": 0.1155,
+      "step": 18040
+    },
+    {
+      "epoch": 1.708956637000568,
+      "grad_norm": 0.37756624817848206,
+      "learning_rate": 1.3164173451997729e-05,
+      "loss": 0.1093,
+      "step": 18050
+    },
+    {
+      "epoch": 1.7099034273811777,
+      "grad_norm": 0.4071192443370819,
+      "learning_rate": 1.3160386290475289e-05,
+      "loss": 0.1026,
+      "step": 18060
+    },
+    {
+      "epoch": 1.7108502177617875,
+      "grad_norm": 0.3811383545398712,
+      "learning_rate": 1.315659912895285e-05,
+      "loss": 0.0979,
+      "step": 18070
+    },
+    {
+      "epoch": 1.7117970081423972,
+      "grad_norm": 0.32877886295318604,
+      "learning_rate": 1.3152811967430413e-05,
+      "loss": 0.1148,
+      "step": 18080
+    },
+    {
+      "epoch": 1.712743798523007,
+      "grad_norm": 0.42139747738838196,
+      "learning_rate": 1.3149024805907973e-05,
+      "loss": 0.1129,
+      "step": 18090
+    },
+    {
+      "epoch": 1.7136905889036167,
+      "grad_norm": 0.2595408260822296,
+      "learning_rate": 1.3145237644385533e-05,
+      "loss": 0.1049,
+      "step": 18100
+    },
+    {
+      "epoch": 1.7146373792842264,
+      "grad_norm": 0.38902461528778076,
+      "learning_rate": 1.3141450482863096e-05,
+      "loss": 0.1105,
+      "step": 18110
+    },
+    {
+      "epoch": 1.7155841696648362,
+      "grad_norm": 0.3774598240852356,
+      "learning_rate": 1.3137663321340656e-05,
+      "loss": 0.114,
+      "step": 18120
+    },
+    {
+      "epoch": 1.716530960045446,
+      "grad_norm": 0.29289478063583374,
+      "learning_rate": 1.3133876159818218e-05,
+      "loss": 0.0974,
+      "step": 18130
+    },
+    {
+      "epoch": 1.7174777504260557,
+      "grad_norm": 0.383760541677475,
+      "learning_rate": 1.3130088998295778e-05,
+      "loss": 0.1078,
+      "step": 18140
+    },
+    {
+      "epoch": 1.7184245408066654,
+      "grad_norm": 0.33194929361343384,
+      "learning_rate": 1.3126301836773338e-05,
+      "loss": 0.1192,
+      "step": 18150
+    },
+    {
+      "epoch": 1.7193713311872751,
+      "grad_norm": 0.41859087347984314,
+      "learning_rate": 1.3122514675250901e-05,
+      "loss": 0.1125,
+      "step": 18160
+    },
+    {
+      "epoch": 1.7203181215678849,
+      "grad_norm": 0.3526773750782013,
+      "learning_rate": 1.3118727513728461e-05,
+      "loss": 0.1102,
+      "step": 18170
+    },
+    {
+      "epoch": 1.7212649119484946,
+      "grad_norm": 0.3202621042728424,
+      "learning_rate": 1.3114940352206023e-05,
+      "loss": 0.1103,
+      "step": 18180
+    },
+    {
+      "epoch": 1.7222117023291044,
+      "grad_norm": 0.37917235493659973,
+      "learning_rate": 1.3111153190683583e-05,
+      "loss": 0.1153,
+      "step": 18190
+    },
+    {
+      "epoch": 1.723158492709714,
+      "grad_norm": 0.41578903794288635,
+      "learning_rate": 1.3107366029161143e-05,
+      "loss": 0.1048,
+      "step": 18200
+    },
+    {
+      "epoch": 1.7241052830903238,
+      "grad_norm": 0.33654308319091797,
+      "learning_rate": 1.3103578867638707e-05,
+      "loss": 0.1089,
+      "step": 18210
+    },
+    {
+      "epoch": 1.7250520734709336,
+      "grad_norm": 0.3018990457057953,
+      "learning_rate": 1.3099791706116267e-05,
+      "loss": 0.1097,
+      "step": 18220
+    },
+    {
+      "epoch": 1.7259988638515433,
+      "grad_norm": 0.36322179436683655,
+      "learning_rate": 1.3096004544593829e-05,
+      "loss": 0.0998,
+      "step": 18230
+    },
+    {
+      "epoch": 1.726945654232153,
+      "grad_norm": 0.3884753882884979,
+      "learning_rate": 1.3092217383071389e-05,
+      "loss": 0.118,
+      "step": 18240
+    },
+    {
+      "epoch": 1.7278924446127628,
+      "grad_norm": 0.31357282400131226,
+      "learning_rate": 1.308843022154895e-05,
+      "loss": 0.1103,
+      "step": 18250
+    },
+    {
+      "epoch": 1.7288392349933726,
+      "grad_norm": 0.3505796492099762,
+      "learning_rate": 1.3084643060026512e-05,
+      "loss": 0.1044,
+      "step": 18260
+    },
+    {
+      "epoch": 1.7297860253739823,
+      "grad_norm": 0.37499359250068665,
+      "learning_rate": 1.3080855898504072e-05,
+      "loss": 0.1157,
+      "step": 18270
+    },
+    {
+      "epoch": 1.730732815754592,
+      "grad_norm": 0.42914479970932007,
+      "learning_rate": 1.3077068736981632e-05,
+      "loss": 0.116,
+      "step": 18280
+    },
+    {
+      "epoch": 1.7316796061352018,
+      "grad_norm": 0.3526166081428528,
+      "learning_rate": 1.3073281575459194e-05,
+      "loss": 0.1052,
+      "step": 18290
+    },
+    {
+      "epoch": 1.7326263965158115,
+      "grad_norm": 0.2813660204410553,
+      "learning_rate": 1.3069494413936756e-05,
+      "loss": 0.1033,
+      "step": 18300
+    },
+    {
+      "epoch": 1.7335731868964213,
+      "grad_norm": 0.3836597800254822,
+      "learning_rate": 1.3065707252414317e-05,
+      "loss": 0.1045,
+      "step": 18310
+    },
+    {
+      "epoch": 1.734519977277031,
+      "grad_norm": 0.41530275344848633,
+      "learning_rate": 1.3061920090891877e-05,
+      "loss": 0.0978,
+      "step": 18320
+    },
+    {
+      "epoch": 1.7354667676576407,
+      "grad_norm": 0.3948558270931244,
+      "learning_rate": 1.3058132929369438e-05,
+      "loss": 0.1063,
+      "step": 18330
+    },
+    {
+      "epoch": 1.7364135580382505,
+      "grad_norm": 0.34573134779930115,
+      "learning_rate": 1.3054345767847e-05,
+      "loss": 0.1063,
+      "step": 18340
+    },
+    {
+      "epoch": 1.7373603484188602,
+      "grad_norm": 0.41142192482948303,
+      "learning_rate": 1.3050558606324561e-05,
+      "loss": 0.116,
+      "step": 18350
+    },
+    {
+      "epoch": 1.7383071387994697,
+      "grad_norm": 0.34815993905067444,
+      "learning_rate": 1.3046771444802123e-05,
+      "loss": 0.1106,
+      "step": 18360
+    },
+    {
+      "epoch": 1.7392539291800795,
+      "grad_norm": 0.3826438784599304,
+      "learning_rate": 1.3042984283279683e-05,
+      "loss": 0.1089,
+      "step": 18370
+    },
+    {
+      "epoch": 1.7402007195606892,
+      "grad_norm": 0.3685339093208313,
+      "learning_rate": 1.3039197121757243e-05,
+      "loss": 0.103,
+      "step": 18380
+    },
+    {
+      "epoch": 1.741147509941299,
+      "grad_norm": 0.33224424719810486,
+      "learning_rate": 1.3035409960234806e-05,
+      "loss": 0.1075,
+      "step": 18390
+    },
+    {
+      "epoch": 1.7420943003219087,
+      "grad_norm": 0.4313163757324219,
+      "learning_rate": 1.3031622798712366e-05,
+      "loss": 0.1098,
+      "step": 18400
+    },
+    {
+      "epoch": 1.7430410907025184,
+      "grad_norm": 0.3623204827308655,
+      "learning_rate": 1.3027835637189928e-05,
+      "loss": 0.1059,
+      "step": 18410
+    },
+    {
+      "epoch": 1.7439878810831282,
+      "grad_norm": 0.44435903429985046,
+      "learning_rate": 1.3024048475667488e-05,
+      "loss": 0.1094,
+      "step": 18420
+    },
+    {
+      "epoch": 1.744934671463738,
+      "grad_norm": 0.41246020793914795,
+      "learning_rate": 1.3020261314145048e-05,
+      "loss": 0.1143,
+      "step": 18430
+    },
+    {
+      "epoch": 1.7458814618443477,
+      "grad_norm": 0.3123202323913574,
+      "learning_rate": 1.3016474152622612e-05,
+      "loss": 0.1027,
+      "step": 18440
+    },
+    {
+      "epoch": 1.7468282522249574,
+      "grad_norm": 0.5212844014167786,
+      "learning_rate": 1.3012686991100172e-05,
+      "loss": 0.1157,
+      "step": 18450
+    },
+    {
+      "epoch": 1.7477750426055672,
+      "grad_norm": 0.43488380312919617,
+      "learning_rate": 1.3008899829577732e-05,
+      "loss": 0.1086,
+      "step": 18460
+    },
+    {
+      "epoch": 1.748721832986177,
+      "grad_norm": 0.3901176452636719,
+      "learning_rate": 1.3005112668055293e-05,
+      "loss": 0.1078,
+      "step": 18470
+    },
+    {
+      "epoch": 1.7496686233667866,
+      "grad_norm": 0.2928845286369324,
+      "learning_rate": 1.3001325506532854e-05,
+      "loss": 0.1166,
+      "step": 18480
+    },
+    {
+      "epoch": 1.7506154137473964,
+      "grad_norm": 0.340564101934433,
+      "learning_rate": 1.2997538345010417e-05,
+      "loss": 0.1087,
+      "step": 18490
+    },
+    {
+      "epoch": 1.751562204128006,
+      "grad_norm": 0.3431471288204193,
+      "learning_rate": 1.2993751183487977e-05,
+      "loss": 0.1094,
+      "step": 18500
+    },
+    {
+      "epoch": 1.7525089945086156,
+      "grad_norm": 0.35341206192970276,
+      "learning_rate": 1.2989964021965537e-05,
+      "loss": 0.0969,
+      "step": 18510
+    },
+    {
+      "epoch": 1.7534557848892254,
+      "grad_norm": 0.3784622550010681,
+      "learning_rate": 1.2986176860443099e-05,
+      "loss": 0.1087,
+      "step": 18520
+    },
+    {
+      "epoch": 1.7544025752698351,
+      "grad_norm": 0.3758784532546997,
+      "learning_rate": 1.298238969892066e-05,
+      "loss": 0.1087,
+      "step": 18530
+    },
+    {
+      "epoch": 1.7553493656504449,
+      "grad_norm": 0.31814923882484436,
+      "learning_rate": 1.2978602537398222e-05,
+      "loss": 0.1035,
+      "step": 18540
+    },
+    {
+      "epoch": 1.7562961560310546,
+      "grad_norm": 0.28400737047195435,
+      "learning_rate": 1.2974815375875782e-05,
+      "loss": 0.1037,
+      "step": 18550
+    },
+    {
+      "epoch": 1.7572429464116643,
+      "grad_norm": 0.3504233956336975,
+      "learning_rate": 1.2971028214353342e-05,
+      "loss": 0.1049,
+      "step": 18560
+    },
+    {
+      "epoch": 1.758189736792274,
+      "grad_norm": 0.3481403589248657,
+      "learning_rate": 1.2967241052830904e-05,
+      "loss": 0.1078,
+      "step": 18570
+    },
+    {
+      "epoch": 1.7591365271728838,
+      "grad_norm": 0.3445204794406891,
+      "learning_rate": 1.2963453891308466e-05,
+      "loss": 0.1093,
+      "step": 18580
+    },
+    {
+      "epoch": 1.7600833175534936,
+      "grad_norm": 0.3659699559211731,
+      "learning_rate": 1.2959666729786028e-05,
+      "loss": 0.1137,
+      "step": 18590
+    },
+    {
+      "epoch": 1.7610301079341033,
+      "grad_norm": 0.35296133160591125,
+      "learning_rate": 1.2955879568263588e-05,
+      "loss": 0.107,
+      "step": 18600
+    },
+    {
+      "epoch": 1.761976898314713,
+      "grad_norm": 0.30524009466171265,
+      "learning_rate": 1.2952092406741148e-05,
+      "loss": 0.1009,
+      "step": 18610
+    },
+    {
+      "epoch": 1.7629236886953228,
+      "grad_norm": 0.3883345127105713,
+      "learning_rate": 1.2948305245218708e-05,
+      "loss": 0.1173,
+      "step": 18620
+    },
+    {
+      "epoch": 1.7638704790759325,
+      "grad_norm": 0.2975183427333832,
+      "learning_rate": 1.2944518083696271e-05,
+      "loss": 0.117,
+      "step": 18630
+    },
+    {
+      "epoch": 1.7648172694565423,
+      "grad_norm": 0.35033711791038513,
+      "learning_rate": 1.2940730922173831e-05,
+      "loss": 0.1081,
+      "step": 18640
+    },
+    {
+      "epoch": 1.765764059837152,
+      "grad_norm": 0.3302345871925354,
+      "learning_rate": 1.2936943760651393e-05,
+      "loss": 0.1063,
+      "step": 18650
+    },
+    {
+      "epoch": 1.7667108502177618,
+      "grad_norm": 0.30319857597351074,
+      "learning_rate": 1.2933156599128953e-05,
+      "loss": 0.0999,
+      "step": 18660
+    },
+    {
+      "epoch": 1.7676576405983715,
+      "grad_norm": 0.39417383074760437,
+      "learning_rate": 1.2929369437606516e-05,
+      "loss": 0.1059,
+      "step": 18670
+    },
+    {
+      "epoch": 1.7686044309789812,
+      "grad_norm": 0.39419564604759216,
+      "learning_rate": 1.2925582276084077e-05,
+      "loss": 0.1081,
+      "step": 18680
+    },
+    {
+      "epoch": 1.769551221359591,
+      "grad_norm": 0.30187517404556274,
+      "learning_rate": 1.2921795114561637e-05,
+      "loss": 0.1064,
+      "step": 18690
+    },
+    {
+      "epoch": 1.7704980117402007,
+      "grad_norm": 0.3162728548049927,
+      "learning_rate": 1.2918007953039198e-05,
+      "loss": 0.1037,
+      "step": 18700
+    },
+    {
+      "epoch": 1.7714448021208105,
+      "grad_norm": 0.31417393684387207,
+      "learning_rate": 1.2914220791516758e-05,
+      "loss": 0.0992,
+      "step": 18710
+    },
+    {
+      "epoch": 1.7723915925014202,
+      "grad_norm": 0.3579910695552826,
+      "learning_rate": 1.2910433629994322e-05,
+      "loss": 0.1048,
+      "step": 18720
+    },
+    {
+      "epoch": 1.77333838288203,
+      "grad_norm": 0.4408959746360779,
+      "learning_rate": 1.2906646468471882e-05,
+      "loss": 0.1051,
+      "step": 18730
+    },
+    {
+      "epoch": 1.7742851732626397,
+      "grad_norm": 0.32164129614830017,
+      "learning_rate": 1.2902859306949442e-05,
+      "loss": 0.1097,
+      "step": 18740
+    },
+    {
+      "epoch": 1.7752319636432494,
+      "grad_norm": 0.28425174951553345,
+      "learning_rate": 1.2899072145427004e-05,
+      "loss": 0.11,
+      "step": 18750
+    },
+    {
+      "epoch": 1.7761787540238592,
+      "grad_norm": 0.3800603449344635,
+      "learning_rate": 1.2895284983904564e-05,
+      "loss": 0.1081,
+      "step": 18760
+    },
+    {
+      "epoch": 1.777125544404469,
+      "grad_norm": 0.3860698938369751,
+      "learning_rate": 1.2891497822382127e-05,
+      "loss": 0.1078,
+      "step": 18770
+    },
+    {
+      "epoch": 1.7780723347850786,
+      "grad_norm": 0.3523519039154053,
+      "learning_rate": 1.2887710660859687e-05,
+      "loss": 0.1115,
+      "step": 18780
+    },
+    {
+      "epoch": 1.7790191251656884,
+      "grad_norm": 0.5016382932662964,
+      "learning_rate": 1.2883923499337247e-05,
+      "loss": 0.1106,
+      "step": 18790
+    },
+    {
+      "epoch": 1.7799659155462981,
+      "grad_norm": 0.34837406873703003,
+      "learning_rate": 1.2880136337814807e-05,
+      "loss": 0.1046,
+      "step": 18800
+    },
+    {
+      "epoch": 1.7809127059269079,
+      "grad_norm": 0.34380877017974854,
+      "learning_rate": 1.287634917629237e-05,
+      "loss": 0.1028,
+      "step": 18810
+    },
+    {
+      "epoch": 1.7818594963075176,
+      "grad_norm": 0.32999187707901,
+      "learning_rate": 1.287256201476993e-05,
+      "loss": 0.0998,
+      "step": 18820
+    },
+    {
+      "epoch": 1.7828062866881274,
+      "grad_norm": 0.36616331338882446,
+      "learning_rate": 1.2868774853247493e-05,
+      "loss": 0.0974,
+      "step": 18830
+    },
+    {
+      "epoch": 1.783753077068737,
+      "grad_norm": 0.3709386885166168,
+      "learning_rate": 1.2864987691725053e-05,
+      "loss": 0.1101,
+      "step": 18840
+    },
+    {
+      "epoch": 1.7846998674493468,
+      "grad_norm": 0.3592256009578705,
+      "learning_rate": 1.2861200530202613e-05,
+      "loss": 0.1136,
+      "step": 18850
+    },
+    {
+      "epoch": 1.7856466578299566,
+      "grad_norm": 0.532692015171051,
+      "learning_rate": 1.2857413368680176e-05,
+      "loss": 0.1119,
+      "step": 18860
+    },
+    {
+      "epoch": 1.7865934482105663,
+      "grad_norm": 0.36976680159568787,
+      "learning_rate": 1.2853626207157736e-05,
+      "loss": 0.1098,
+      "step": 18870
+    },
+    {
+      "epoch": 1.787540238591176,
+      "grad_norm": 0.452336847782135,
+      "learning_rate": 1.2849839045635298e-05,
+      "loss": 0.1029,
+      "step": 18880
+    },
+    {
+      "epoch": 1.7884870289717858,
+      "grad_norm": 0.44521087408065796,
+      "learning_rate": 1.2846051884112858e-05,
+      "loss": 0.111,
+      "step": 18890
+    },
+    {
+      "epoch": 1.7894338193523955,
+      "grad_norm": 0.34995198249816895,
+      "learning_rate": 1.2842264722590418e-05,
+      "loss": 0.1091,
+      "step": 18900
+    },
+    {
+      "epoch": 1.790380609733005,
+      "grad_norm": 0.3400953710079193,
+      "learning_rate": 1.2838477561067981e-05,
+      "loss": 0.1009,
+      "step": 18910
+    },
+    {
+      "epoch": 1.7913274001136148,
+      "grad_norm": 0.37189778685569763,
+      "learning_rate": 1.2834690399545541e-05,
+      "loss": 0.1034,
+      "step": 18920
+    },
+    {
+      "epoch": 1.7922741904942245,
+      "grad_norm": 0.3677590787410736,
+      "learning_rate": 1.2830903238023103e-05,
+      "loss": 0.1039,
+      "step": 18930
+    },
+    {
+      "epoch": 1.7932209808748343,
+      "grad_norm": 0.32119473814964294,
+      "learning_rate": 1.2827116076500663e-05,
+      "loss": 0.1073,
+      "step": 18940
+    },
+    {
+      "epoch": 1.794167771255444,
+      "grad_norm": 0.3377484679222107,
+      "learning_rate": 1.2823328914978227e-05,
+      "loss": 0.1176,
+      "step": 18950
+    },
+    {
+      "epoch": 1.7951145616360538,
+      "grad_norm": 0.2578378915786743,
+      "learning_rate": 1.2819541753455787e-05,
+      "loss": 0.0994,
+      "step": 18960
+    },
+    {
+      "epoch": 1.7960613520166635,
+      "grad_norm": 0.34310463070869446,
+      "learning_rate": 1.2815754591933347e-05,
+      "loss": 0.1026,
+      "step": 18970
+    },
+    {
+      "epoch": 1.7970081423972732,
+      "grad_norm": 0.33588457107543945,
+      "learning_rate": 1.2811967430410907e-05,
+      "loss": 0.1039,
+      "step": 18980
+    },
+    {
+      "epoch": 1.797954932777883,
+      "grad_norm": 0.3982495367527008,
+      "learning_rate": 1.2808180268888469e-05,
+      "loss": 0.1154,
+      "step": 18990
+    },
+    {
+      "epoch": 1.7989017231584927,
+      "grad_norm": 0.4286862909793854,
+      "learning_rate": 1.280439310736603e-05,
+      "loss": 0.1143,
+      "step": 19000
+    },
+    {
+      "epoch": 1.7998485135391025,
+      "grad_norm": 0.37063848972320557,
+      "learning_rate": 1.2800605945843592e-05,
+      "loss": 0.1077,
+      "step": 19010
+    },
+    {
+      "epoch": 1.8007953039197122,
+      "grad_norm": 0.4058857262134552,
+      "learning_rate": 1.2796818784321152e-05,
+      "loss": 0.1165,
+      "step": 19020
+    },
+    {
+      "epoch": 1.801742094300322,
+      "grad_norm": 0.3987621068954468,
+      "learning_rate": 1.2793031622798712e-05,
+      "loss": 0.1118,
+      "step": 19030
+    },
+    {
+      "epoch": 1.8026888846809317,
+      "grad_norm": 0.41691696643829346,
+      "learning_rate": 1.2789244461276274e-05,
+      "loss": 0.1178,
+      "step": 19040
+    },
+    {
+      "epoch": 1.8036356750615412,
+      "grad_norm": 0.4202606678009033,
+      "learning_rate": 1.2785457299753836e-05,
+      "loss": 0.1091,
+      "step": 19050
+    },
+    {
+      "epoch": 1.804582465442151,
+      "grad_norm": 0.39799049496650696,
+      "learning_rate": 1.2781670138231397e-05,
+      "loss": 0.1096,
+      "step": 19060
+    },
+    {
+      "epoch": 1.8055292558227607,
+      "grad_norm": 0.3521597385406494,
+      "learning_rate": 1.2777882976708957e-05,
+      "loss": 0.1062,
+      "step": 19070
+    },
+    {
+      "epoch": 1.8064760462033704,
+      "grad_norm": 0.3163425326347351,
+      "learning_rate": 1.2774095815186517e-05,
+      "loss": 0.1077,
+      "step": 19080
+    },
+    {
+      "epoch": 1.8074228365839802,
+      "grad_norm": 0.3193988800048828,
+      "learning_rate": 1.2770308653664081e-05,
+      "loss": 0.1037,
+      "step": 19090
+    },
+    {
+      "epoch": 1.80836962696459,
+      "grad_norm": 0.34172874689102173,
+      "learning_rate": 1.2766521492141641e-05,
+      "loss": 0.1086,
+      "step": 19100
+    },
+    {
+      "epoch": 1.8093164173451997,
+      "grad_norm": 0.3774087429046631,
+      "learning_rate": 1.2762734330619203e-05,
+      "loss": 0.1042,
+      "step": 19110
+    },
+    {
+      "epoch": 1.8102632077258094,
+      "grad_norm": 0.4073845148086548,
+      "learning_rate": 1.2758947169096763e-05,
+      "loss": 0.1124,
+      "step": 19120
+    },
+    {
+      "epoch": 1.8112099981064191,
+      "grad_norm": 0.2993011474609375,
+      "learning_rate": 1.2755160007574323e-05,
+      "loss": 0.0977,
+      "step": 19130
+    },
+    {
+      "epoch": 1.8121567884870289,
+      "grad_norm": 0.3964138925075531,
+      "learning_rate": 1.2751372846051886e-05,
+      "loss": 0.1123,
+      "step": 19140
+    },
+    {
+      "epoch": 1.8131035788676386,
+      "grad_norm": 0.39932727813720703,
+      "learning_rate": 1.2747585684529446e-05,
+      "loss": 0.1022,
+      "step": 19150
+    },
+    {
+      "epoch": 1.8140503692482484,
+      "grad_norm": 0.24434198439121246,
+      "learning_rate": 1.2743798523007006e-05,
+      "loss": 0.0994,
+      "step": 19160
+    },
+    {
+      "epoch": 1.814997159628858,
+      "grad_norm": 0.431255042552948,
+      "learning_rate": 1.2740011361484568e-05,
+      "loss": 0.1157,
+      "step": 19170
+    },
+    {
+      "epoch": 1.8159439500094678,
+      "grad_norm": 0.35100871324539185,
+      "learning_rate": 1.2736224199962128e-05,
+      "loss": 0.1129,
+      "step": 19180
+    },
+    {
+      "epoch": 1.8168907403900776,
+      "grad_norm": 0.357479989528656,
+      "learning_rate": 1.2732437038439692e-05,
+      "loss": 0.1034,
+      "step": 19190
+    },
+    {
+      "epoch": 1.8178375307706873,
+      "grad_norm": 0.31612756848335266,
+      "learning_rate": 1.2728649876917252e-05,
+      "loss": 0.1147,
+      "step": 19200
+    },
+    {
+      "epoch": 1.818784321151297,
+      "grad_norm": 0.3879270851612091,
+      "learning_rate": 1.2724862715394812e-05,
+      "loss": 0.1071,
+      "step": 19210
+    },
+    {
+      "epoch": 1.8197311115319068,
+      "grad_norm": 0.3949774205684662,
+      "learning_rate": 1.2721075553872373e-05,
+      "loss": 0.1131,
+      "step": 19220
+    },
+    {
+      "epoch": 1.8206779019125165,
+      "grad_norm": 0.3450389802455902,
+      "learning_rate": 1.2717288392349935e-05,
+      "loss": 0.1172,
+      "step": 19230
+    },
+    {
+      "epoch": 1.8216246922931263,
+      "grad_norm": 0.30399641394615173,
+      "learning_rate": 1.2713501230827497e-05,
+      "loss": 0.1061,
+      "step": 19240
+    },
+    {
+      "epoch": 1.822571482673736,
+      "grad_norm": 0.3874061703681946,
+      "learning_rate": 1.2709714069305057e-05,
+      "loss": 0.1176,
+      "step": 19250
+    },
+    {
+      "epoch": 1.8235182730543458,
+      "grad_norm": 0.3001534640789032,
+      "learning_rate": 1.2705926907782617e-05,
+      "loss": 0.1109,
+      "step": 19260
+    },
+    {
+      "epoch": 1.8244650634349555,
+      "grad_norm": 0.3698557913303375,
+      "learning_rate": 1.2702139746260179e-05,
+      "loss": 0.1094,
+      "step": 19270
+    },
+    {
+      "epoch": 1.8254118538155653,
+      "grad_norm": 0.33681055903434753,
+      "learning_rate": 1.269835258473774e-05,
+      "loss": 0.1043,
+      "step": 19280
+    },
+    {
+      "epoch": 1.826358644196175,
+      "grad_norm": 0.33553245663642883,
+      "learning_rate": 1.2694565423215302e-05,
+      "loss": 0.109,
+      "step": 19290
+    },
+    {
+      "epoch": 1.8273054345767847,
+      "grad_norm": 0.3944529592990875,
+      "learning_rate": 1.2690778261692862e-05,
+      "loss": 0.1056,
+      "step": 19300
+    },
+    {
+      "epoch": 1.8282522249573945,
+      "grad_norm": 0.34673944115638733,
+      "learning_rate": 1.2686991100170422e-05,
+      "loss": 0.098,
+      "step": 19310
+    },
+    {
+      "epoch": 1.8291990153380042,
+      "grad_norm": 0.46945884823799133,
+      "learning_rate": 1.2683203938647984e-05,
+      "loss": 0.0998,
+      "step": 19320
+    },
+    {
+      "epoch": 1.830145805718614,
+      "grad_norm": 0.3739577531814575,
+      "learning_rate": 1.2679416777125546e-05,
+      "loss": 0.1151,
+      "step": 19330
+    },
+    {
+      "epoch": 1.8310925960992237,
+      "grad_norm": 0.32211214303970337,
+      "learning_rate": 1.2675629615603106e-05,
+      "loss": 0.122,
+      "step": 19340
+    },
+    {
+      "epoch": 1.8320393864798334,
+      "grad_norm": 0.3556137979030609,
+      "learning_rate": 1.2671842454080668e-05,
+      "loss": 0.1027,
+      "step": 19350
+    },
+    {
+      "epoch": 1.8329861768604432,
+      "grad_norm": 0.3246605098247528,
+      "learning_rate": 1.2668055292558228e-05,
+      "loss": 0.1168,
+      "step": 19360
+    },
+    {
+      "epoch": 1.833932967241053,
+      "grad_norm": 0.3900707960128784,
+      "learning_rate": 1.2664268131035791e-05,
+      "loss": 0.1166,
+      "step": 19370
+    },
+    {
+      "epoch": 1.8348797576216627,
+      "grad_norm": 0.4111650288105011,
+      "learning_rate": 1.2660480969513351e-05,
+      "loss": 0.1055,
+      "step": 19380
+    },
+    {
+      "epoch": 1.8358265480022724,
+      "grad_norm": 0.29402846097946167,
+      "learning_rate": 1.2656693807990911e-05,
+      "loss": 0.1027,
+      "step": 19390
+    },
+    {
+      "epoch": 1.8367733383828821,
+      "grad_norm": 0.4428441822528839,
+      "learning_rate": 1.2652906646468473e-05,
+      "loss": 0.1067,
+      "step": 19400
+    },
+    {
+      "epoch": 1.837720128763492,
+      "grad_norm": 0.40585818886756897,
+      "learning_rate": 1.2649119484946033e-05,
+      "loss": 0.1085,
+      "step": 19410
+    },
+    {
+      "epoch": 1.8386669191441016,
+      "grad_norm": 0.34224626421928406,
+      "learning_rate": 1.2645332323423596e-05,
+      "loss": 0.1152,
+      "step": 19420
+    },
+    {
+      "epoch": 1.8396137095247114,
+      "grad_norm": 0.34518611431121826,
+      "learning_rate": 1.2641545161901156e-05,
+      "loss": 0.1073,
+      "step": 19430
+    },
+    {
+      "epoch": 1.8405604999053211,
+      "grad_norm": 0.4500565826892853,
+      "learning_rate": 1.2637758000378717e-05,
+      "loss": 0.1136,
+      "step": 19440
+    },
+    {
+      "epoch": 1.8415072902859309,
+      "grad_norm": 0.3417593538761139,
+      "learning_rate": 1.2633970838856278e-05,
+      "loss": 0.1154,
+      "step": 19450
+    },
+    {
+      "epoch": 1.8424540806665404,
+      "grad_norm": 0.3833911120891571,
+      "learning_rate": 1.263018367733384e-05,
+      "loss": 0.1125,
+      "step": 19460
+    },
+    {
+      "epoch": 1.8434008710471501,
+      "grad_norm": 0.34328556060791016,
+      "learning_rate": 1.2626396515811402e-05,
+      "loss": 0.1094,
+      "step": 19470
+    },
+    {
+      "epoch": 1.8443476614277599,
+      "grad_norm": 0.3361929953098297,
+      "learning_rate": 1.2622609354288962e-05,
+      "loss": 0.1104,
+      "step": 19480
+    },
+    {
+      "epoch": 1.8452944518083696,
+      "grad_norm": 0.3236779272556305,
+      "learning_rate": 1.2618822192766522e-05,
+      "loss": 0.1153,
+      "step": 19490
+    },
+    {
+      "epoch": 1.8462412421889793,
+      "grad_norm": 0.33372533321380615,
+      "learning_rate": 1.2615035031244084e-05,
+      "loss": 0.1055,
+      "step": 19500
+    },
+    {
+      "epoch": 1.847188032569589,
+      "grad_norm": 0.28622475266456604,
+      "learning_rate": 1.2611247869721645e-05,
+      "loss": 0.1072,
+      "step": 19510
+    },
+    {
+      "epoch": 1.8481348229501988,
+      "grad_norm": 0.45177966356277466,
+      "learning_rate": 1.2607460708199205e-05,
+      "loss": 0.1142,
+      "step": 19520
+    },
+    {
+      "epoch": 1.8490816133308086,
+      "grad_norm": 0.40346699953079224,
+      "learning_rate": 1.2603673546676767e-05,
+      "loss": 0.1147,
+      "step": 19530
+    },
+    {
+      "epoch": 1.8500284037114183,
+      "grad_norm": 0.3747116029262543,
+      "learning_rate": 1.2599886385154327e-05,
+      "loss": 0.1143,
+      "step": 19540
+    },
+    {
+      "epoch": 1.850975194092028,
+      "grad_norm": 0.37933945655822754,
+      "learning_rate": 1.2596099223631887e-05,
+      "loss": 0.1021,
+      "step": 19550
+    },
+    {
+      "epoch": 1.8519219844726378,
+      "grad_norm": 0.34756040573120117,
+      "learning_rate": 1.259231206210945e-05,
+      "loss": 0.1165,
+      "step": 19560
+    },
+    {
+      "epoch": 1.8528687748532475,
+      "grad_norm": 0.32047250866889954,
+      "learning_rate": 1.258852490058701e-05,
+      "loss": 0.1062,
+      "step": 19570
+    },
+    {
+      "epoch": 1.8538155652338573,
+      "grad_norm": 0.3668143153190613,
+      "learning_rate": 1.2584737739064572e-05,
+      "loss": 0.1149,
+      "step": 19580
+    },
+    {
+      "epoch": 1.854762355614467,
+      "grad_norm": 0.3343770205974579,
+      "learning_rate": 1.2580950577542133e-05,
+      "loss": 0.1055,
+      "step": 19590
+    },
+    {
+      "epoch": 1.8557091459950765,
+      "grad_norm": 0.3434257507324219,
+      "learning_rate": 1.2577163416019696e-05,
+      "loss": 0.1035,
+      "step": 19600
+    },
+    {
+      "epoch": 1.8566559363756863,
+      "grad_norm": 0.3178967535495758,
+      "learning_rate": 1.2573376254497256e-05,
+      "loss": 0.0951,
+      "step": 19610
+    },
+    {
+      "epoch": 1.857602726756296,
+      "grad_norm": 0.3695988059043884,
+      "learning_rate": 1.2569589092974816e-05,
+      "loss": 0.1224,
+      "step": 19620
+    },
+    {
+      "epoch": 1.8585495171369057,
+      "grad_norm": 0.3733889162540436,
+      "learning_rate": 1.2565801931452378e-05,
+      "loss": 0.1029,
+      "step": 19630
+    },
+    {
+      "epoch": 1.8594963075175155,
+      "grad_norm": 0.3214036822319031,
+      "learning_rate": 1.2562014769929938e-05,
+      "loss": 0.115,
+      "step": 19640
+    },
+    {
+      "epoch": 1.8604430978981252,
+      "grad_norm": 0.2950606942176819,
+      "learning_rate": 1.2558227608407501e-05,
+      "loss": 0.1078,
+      "step": 19650
+    },
+    {
+      "epoch": 1.861389888278735,
+      "grad_norm": 0.37673529982566833,
+      "learning_rate": 1.2554440446885061e-05,
+      "loss": 0.1129,
+      "step": 19660
+    },
+    {
+      "epoch": 1.8623366786593447,
+      "grad_norm": 0.36980605125427246,
+      "learning_rate": 1.2550653285362621e-05,
+      "loss": 0.1034,
+      "step": 19670
+    },
+    {
+      "epoch": 1.8632834690399545,
+      "grad_norm": 0.2896386384963989,
+      "learning_rate": 1.2546866123840183e-05,
+      "loss": 0.1164,
+      "step": 19680
+    },
+    {
+      "epoch": 1.8642302594205642,
+      "grad_norm": 0.299468457698822,
+      "learning_rate": 1.2543078962317743e-05,
+      "loss": 0.1094,
+      "step": 19690
+    },
+    {
+      "epoch": 1.865177049801174,
+      "grad_norm": 0.4312908351421356,
+      "learning_rate": 1.2539291800795305e-05,
+      "loss": 0.1132,
+      "step": 19700
+    },
+    {
+      "epoch": 1.8661238401817837,
+      "grad_norm": 0.29344403743743896,
+      "learning_rate": 1.2535504639272867e-05,
+      "loss": 0.1007,
+      "step": 19710
+    },
+    {
+      "epoch": 1.8670706305623934,
+      "grad_norm": 0.37645217776298523,
+      "learning_rate": 1.2531717477750427e-05,
+      "loss": 0.1109,
+      "step": 19720
+    },
+    {
+      "epoch": 1.8680174209430032,
+      "grad_norm": 0.40150484442710876,
+      "learning_rate": 1.2527930316227987e-05,
+      "loss": 0.1068,
+      "step": 19730
+    },
+    {
+      "epoch": 1.868964211323613,
+      "grad_norm": 0.3141026198863983,
+      "learning_rate": 1.252414315470555e-05,
+      "loss": 0.1082,
+      "step": 19740
+    },
+    {
+      "epoch": 1.8699110017042226,
+      "grad_norm": 0.4085179269313812,
+      "learning_rate": 1.252035599318311e-05,
+      "loss": 0.1016,
+      "step": 19750
+    },
+    {
+      "epoch": 1.8708577920848324,
+      "grad_norm": 0.33192178606987,
+      "learning_rate": 1.2516568831660672e-05,
+      "loss": 0.1054,
+      "step": 19760
+    },
+    {
+      "epoch": 1.8718045824654421,
+      "grad_norm": 0.3281863033771515,
+      "learning_rate": 1.2512781670138232e-05,
+      "loss": 0.1093,
+      "step": 19770
+    },
+    {
+      "epoch": 1.8727513728460519,
+      "grad_norm": 0.27537137269973755,
+      "learning_rate": 1.2508994508615792e-05,
+      "loss": 0.1078,
+      "step": 19780
+    },
+    {
+      "epoch": 1.8736981632266616,
+      "grad_norm": 0.34461522102355957,
+      "learning_rate": 1.2505207347093356e-05,
+      "loss": 0.1146,
+      "step": 19790
+    },
+    {
+      "epoch": 1.8746449536072713,
+      "grad_norm": 0.3809272050857544,
+      "learning_rate": 1.2501420185570916e-05,
+      "loss": 0.1051,
+      "step": 19800
+    },
+    {
+      "epoch": 1.875591743987881,
+      "grad_norm": 0.4907462000846863,
+      "learning_rate": 1.2497633024048477e-05,
+      "loss": 0.1085,
+      "step": 19810
+    },
+    {
+      "epoch": 1.8765385343684908,
+      "grad_norm": 0.32635852694511414,
+      "learning_rate": 1.2493845862526037e-05,
+      "loss": 0.111,
+      "step": 19820
+    },
+    {
+      "epoch": 1.8774853247491006,
+      "grad_norm": 0.41782793402671814,
+      "learning_rate": 1.2490058701003597e-05,
+      "loss": 0.1049,
+      "step": 19830
+    },
+    {
+      "epoch": 1.8784321151297103,
+      "grad_norm": 0.34212571382522583,
+      "learning_rate": 1.2486271539481161e-05,
+      "loss": 0.1073,
+      "step": 19840
+    },
+    {
+      "epoch": 1.87937890551032,
+      "grad_norm": 0.3190506100654602,
+      "learning_rate": 1.2482484377958721e-05,
+      "loss": 0.0979,
+      "step": 19850
+    },
+    {
+      "epoch": 1.8803256958909298,
+      "grad_norm": 0.34766045212745667,
+      "learning_rate": 1.2478697216436283e-05,
+      "loss": 0.1017,
+      "step": 19860
+    },
+    {
+      "epoch": 1.8812724862715395,
+      "grad_norm": 0.3498804271221161,
+      "learning_rate": 1.2474910054913843e-05,
+      "loss": 0.103,
+      "step": 19870
+    },
+    {
+      "epoch": 1.8822192766521493,
+      "grad_norm": 0.38651737570762634,
+      "learning_rate": 1.2471122893391404e-05,
+      "loss": 0.1176,
+      "step": 19880
+    },
+    {
+      "epoch": 1.883166067032759,
+      "grad_norm": 0.2624821364879608,
+      "learning_rate": 1.2467335731868966e-05,
+      "loss": 0.1112,
+      "step": 19890
+    },
+    {
+      "epoch": 1.8841128574133688,
+      "grad_norm": 0.3932912349700928,
+      "learning_rate": 1.2463548570346526e-05,
+      "loss": 0.1133,
+      "step": 19900
+    },
+    {
+      "epoch": 1.8850596477939785,
+      "grad_norm": 0.3405761420726776,
+      "learning_rate": 1.2459761408824086e-05,
+      "loss": 0.1193,
+      "step": 19910
+    },
+    {
+      "epoch": 1.8860064381745882,
+      "grad_norm": 0.31691908836364746,
+      "learning_rate": 1.2455974247301648e-05,
+      "loss": 0.1105,
+      "step": 19920
+    },
+    {
+      "epoch": 1.886953228555198,
+      "grad_norm": 0.3382716178894043,
+      "learning_rate": 1.245218708577921e-05,
+      "loss": 0.106,
+      "step": 19930
+    },
+    {
+      "epoch": 1.8879000189358077,
+      "grad_norm": 0.3085930645465851,
+      "learning_rate": 1.2448399924256772e-05,
+      "loss": 0.0963,
+      "step": 19940
+    },
+    {
+      "epoch": 1.8888468093164175,
+      "grad_norm": 0.3540954887866974,
+      "learning_rate": 1.2444612762734332e-05,
+      "loss": 0.1054,
+      "step": 19950
+    },
+    {
+      "epoch": 1.8897935996970272,
+      "grad_norm": 0.37571460008621216,
+      "learning_rate": 1.2440825601211892e-05,
+      "loss": 0.1063,
+      "step": 19960
+    },
+    {
+      "epoch": 1.890740390077637,
+      "grad_norm": 0.4311580955982208,
+      "learning_rate": 1.2437038439689453e-05,
+      "loss": 0.1046,
+      "step": 19970
+    },
+    {
+      "epoch": 1.8916871804582467,
+      "grad_norm": 0.4840514659881592,
+      "learning_rate": 1.2433251278167015e-05,
+      "loss": 0.1126,
+      "step": 19980
+    },
+    {
+      "epoch": 1.8926339708388564,
+      "grad_norm": 0.33492323756217957,
+      "learning_rate": 1.2429464116644577e-05,
+      "loss": 0.1036,
+      "step": 19990
+    },
+    {
+      "epoch": 1.8935807612194662,
+      "grad_norm": 0.3419467806816101,
+      "learning_rate": 1.2425676955122137e-05,
+      "loss": 0.1008,
+      "step": 20000
+    },
+    {
+      "epoch": 1.894527551600076,
+      "grad_norm": 0.35411712527275085,
+      "learning_rate": 1.2421889793599697e-05,
+      "loss": 0.0967,
+      "step": 20010
+    },
+    {
+      "epoch": 1.8954743419806854,
+      "grad_norm": 0.508010983467102,
+      "learning_rate": 1.241810263207726e-05,
+      "loss": 0.1017,
+      "step": 20020
+    },
+    {
+      "epoch": 1.8964211323612952,
+      "grad_norm": 0.32599472999572754,
+      "learning_rate": 1.241431547055482e-05,
+      "loss": 0.1,
+      "step": 20030
+    },
+    {
+      "epoch": 1.897367922741905,
+      "grad_norm": 0.401747465133667,
+      "learning_rate": 1.2410528309032382e-05,
+      "loss": 0.1251,
+      "step": 20040
+    },
+    {
+      "epoch": 1.8983147131225147,
+      "grad_norm": 0.37528371810913086,
+      "learning_rate": 1.2406741147509942e-05,
+      "loss": 0.1113,
+      "step": 20050
+    },
+    {
+      "epoch": 1.8992615035031244,
+      "grad_norm": 0.396750271320343,
+      "learning_rate": 1.2402953985987502e-05,
+      "loss": 0.1058,
+      "step": 20060
+    },
+    {
+      "epoch": 1.9002082938837341,
+      "grad_norm": 0.38264715671539307,
+      "learning_rate": 1.2399166824465066e-05,
+      "loss": 0.1173,
+      "step": 20070
+    },
+    {
+      "epoch": 1.9011550842643439,
+      "grad_norm": 0.2985268235206604,
+      "learning_rate": 1.2395379662942626e-05,
+      "loss": 0.1064,
+      "step": 20080
+    },
+    {
+      "epoch": 1.9021018746449536,
+      "grad_norm": 0.40297383069992065,
+      "learning_rate": 1.2391592501420186e-05,
+      "loss": 0.1114,
+      "step": 20090
+    },
+    {
+      "epoch": 1.9030486650255634,
+      "grad_norm": 0.344043493270874,
+      "learning_rate": 1.2387805339897748e-05,
+      "loss": 0.1038,
+      "step": 20100
+    },
+    {
+      "epoch": 1.903995455406173,
+      "grad_norm": 0.35050851106643677,
+      "learning_rate": 1.2384018178375308e-05,
+      "loss": 0.103,
+      "step": 20110
+    },
+    {
+      "epoch": 1.9049422457867828,
+      "grad_norm": 0.35159409046173096,
+      "learning_rate": 1.2380231016852871e-05,
+      "loss": 0.1114,
+      "step": 20120
+    },
+    {
+      "epoch": 1.9058890361673926,
+      "grad_norm": 0.3516301214694977,
+      "learning_rate": 1.2376443855330431e-05,
+      "loss": 0.1026,
+      "step": 20130
+    },
+    {
+      "epoch": 1.9068358265480023,
+      "grad_norm": 0.3457273542881012,
+      "learning_rate": 1.2372656693807991e-05,
+      "loss": 0.104,
+      "step": 20140
+    },
+    {
+      "epoch": 1.9077826169286118,
+      "grad_norm": 0.39187681674957275,
+      "learning_rate": 1.2368869532285553e-05,
+      "loss": 0.1091,
+      "step": 20150
+    },
+    {
+      "epoch": 1.9087294073092216,
+      "grad_norm": 0.41298121213912964,
+      "learning_rate": 1.2365082370763115e-05,
+      "loss": 0.11,
+      "step": 20160
+    },
+    {
+      "epoch": 1.9096761976898313,
+      "grad_norm": 0.47788843512535095,
+      "learning_rate": 1.2361295209240676e-05,
+      "loss": 0.1054,
+      "step": 20170
+    },
+    {
+      "epoch": 1.910622988070441,
+      "grad_norm": 0.45805585384368896,
+      "learning_rate": 1.2357508047718236e-05,
+      "loss": 0.1067,
+      "step": 20180
+    },
+    {
+      "epoch": 1.9115697784510508,
+      "grad_norm": 0.27484065294265747,
+      "learning_rate": 1.2353720886195796e-05,
+      "loss": 0.1078,
+      "step": 20190
+    },
+    {
+      "epoch": 1.9125165688316605,
+      "grad_norm": 0.39492151141166687,
+      "learning_rate": 1.2349933724673358e-05,
+      "loss": 0.1123,
+      "step": 20200
+    },
+    {
+      "epoch": 1.9134633592122703,
+      "grad_norm": 0.29254859685897827,
+      "learning_rate": 1.234614656315092e-05,
+      "loss": 0.1099,
+      "step": 20210
+    },
+    {
+      "epoch": 1.91441014959288,
+      "grad_norm": 0.41775816679000854,
+      "learning_rate": 1.2342359401628482e-05,
+      "loss": 0.106,
+      "step": 20220
+    },
+    {
+      "epoch": 1.9153569399734898,
+      "grad_norm": 0.30593568086624146,
+      "learning_rate": 1.2338572240106042e-05,
+      "loss": 0.1139,
+      "step": 20230
+    },
+    {
+      "epoch": 1.9163037303540995,
+      "grad_norm": 0.38288331031799316,
+      "learning_rate": 1.2334785078583602e-05,
+      "loss": 0.111,
+      "step": 20240
+    },
+    {
+      "epoch": 1.9172505207347093,
+      "grad_norm": 0.3679775297641754,
+      "learning_rate": 1.2330997917061162e-05,
+      "loss": 0.108,
+      "step": 20250
+    },
+    {
+      "epoch": 1.918197311115319,
+      "grad_norm": 0.33158156275749207,
+      "learning_rate": 1.2327210755538725e-05,
+      "loss": 0.1157,
+      "step": 20260
+    },
+    {
+      "epoch": 1.9191441014959287,
+      "grad_norm": 0.32633069157600403,
+      "learning_rate": 1.2323423594016285e-05,
+      "loss": 0.0991,
+      "step": 20270
+    },
+    {
+      "epoch": 1.9200908918765385,
+      "grad_norm": 0.3557127118110657,
+      "learning_rate": 1.2319636432493847e-05,
+      "loss": 0.1048,
+      "step": 20280
+    },
+    {
+      "epoch": 1.9210376822571482,
+      "grad_norm": 0.30199623107910156,
+      "learning_rate": 1.2315849270971407e-05,
+      "loss": 0.1035,
+      "step": 20290
+    },
+    {
+      "epoch": 1.921984472637758,
+      "grad_norm": 0.3319079279899597,
+      "learning_rate": 1.231206210944897e-05,
+      "loss": 0.1095,
+      "step": 20300
+    },
+    {
+      "epoch": 1.9229312630183677,
+      "grad_norm": 0.3735467195510864,
+      "learning_rate": 1.230827494792653e-05,
+      "loss": 0.1107,
+      "step": 20310
+    },
+    {
+      "epoch": 1.9238780533989774,
+      "grad_norm": 0.3348022401332855,
+      "learning_rate": 1.230448778640409e-05,
+      "loss": 0.0992,
+      "step": 20320
+    },
+    {
+      "epoch": 1.9248248437795872,
+      "grad_norm": 0.37244561314582825,
+      "learning_rate": 1.2300700624881652e-05,
+      "loss": 0.1044,
+      "step": 20330
+    },
+    {
+      "epoch": 1.925771634160197,
+      "grad_norm": 0.36787769198417664,
+      "learning_rate": 1.2296913463359212e-05,
+      "loss": 0.0992,
+      "step": 20340
+    },
+    {
+      "epoch": 1.9267184245408067,
+      "grad_norm": 0.4016842842102051,
+      "learning_rate": 1.2293126301836776e-05,
+      "loss": 0.109,
+      "step": 20350
+    },
+    {
+      "epoch": 1.9276652149214164,
+      "grad_norm": 0.39530324935913086,
+      "learning_rate": 1.2289339140314336e-05,
+      "loss": 0.1057,
+      "step": 20360
+    },
+    {
+      "epoch": 1.9286120053020261,
+      "grad_norm": 0.4053460955619812,
+      "learning_rate": 1.2285551978791896e-05,
+      "loss": 0.1048,
+      "step": 20370
+    },
+    {
+      "epoch": 1.9295587956826359,
+      "grad_norm": 0.4953003525733948,
+      "learning_rate": 1.2281764817269458e-05,
+      "loss": 0.1077,
+      "step": 20380
+    },
+    {
+      "epoch": 1.9305055860632456,
+      "grad_norm": 0.4115254878997803,
+      "learning_rate": 1.2277977655747018e-05,
+      "loss": 0.1184,
+      "step": 20390
+    },
+    {
+      "epoch": 1.9314523764438554,
+      "grad_norm": 0.27491551637649536,
+      "learning_rate": 1.2274190494224581e-05,
+      "loss": 0.1114,
+      "step": 20400
+    },
+    {
+      "epoch": 1.932399166824465,
+      "grad_norm": 0.3053814470767975,
+      "learning_rate": 1.2270403332702141e-05,
+      "loss": 0.1023,
+      "step": 20410
+    },
+    {
+      "epoch": 1.9333459572050748,
+      "grad_norm": 0.3862939178943634,
+      "learning_rate": 1.2266616171179701e-05,
+      "loss": 0.11,
+      "step": 20420
+    },
+    {
+      "epoch": 1.9342927475856846,
+      "grad_norm": 0.3790120482444763,
+      "learning_rate": 1.2262829009657261e-05,
+      "loss": 0.1107,
+      "step": 20430
+    },
+    {
+      "epoch": 1.9352395379662943,
+      "grad_norm": 0.35550233721733093,
+      "learning_rate": 1.2259041848134825e-05,
+      "loss": 0.0923,
+      "step": 20440
+    },
+    {
+      "epoch": 1.936186328346904,
+      "grad_norm": 0.3539644777774811,
+      "learning_rate": 1.2255254686612385e-05,
+      "loss": 0.1112,
+      "step": 20450
+    },
+    {
+      "epoch": 1.9371331187275138,
+      "grad_norm": 0.4248740077018738,
+      "learning_rate": 1.2251467525089947e-05,
+      "loss": 0.1124,
+      "step": 20460
+    },
+    {
+      "epoch": 1.9380799091081236,
+      "grad_norm": 0.305600643157959,
+      "learning_rate": 1.2247680363567507e-05,
+      "loss": 0.1069,
+      "step": 20470
+    },
+    {
+      "epoch": 1.9390266994887333,
+      "grad_norm": 0.3527030348777771,
+      "learning_rate": 1.2243893202045067e-05,
+      "loss": 0.1074,
+      "step": 20480
+    },
+    {
+      "epoch": 1.939973489869343,
+      "grad_norm": 0.38150569796562195,
+      "learning_rate": 1.224010604052263e-05,
+      "loss": 0.0973,
+      "step": 20490
+    },
+    {
+      "epoch": 1.9409202802499528,
+      "grad_norm": 0.2797296643257141,
+      "learning_rate": 1.223631887900019e-05,
+      "loss": 0.1099,
+      "step": 20500
+    },
+    {
+      "epoch": 1.9418670706305625,
+      "grad_norm": 0.38830894231796265,
+      "learning_rate": 1.2232531717477752e-05,
+      "loss": 0.1088,
+      "step": 20510
+    },
+    {
+      "epoch": 1.9428138610111723,
+      "grad_norm": 0.3484981954097748,
+      "learning_rate": 1.2228744555955312e-05,
+      "loss": 0.1084,
+      "step": 20520
+    },
+    {
+      "epoch": 1.943760651391782,
+      "grad_norm": 0.45552173256874084,
+      "learning_rate": 1.2224957394432872e-05,
+      "loss": 0.1053,
+      "step": 20530
+    },
+    {
+      "epoch": 1.9447074417723917,
+      "grad_norm": 0.324050635099411,
+      "learning_rate": 1.2221170232910435e-05,
+      "loss": 0.0998,
+      "step": 20540
+    },
+    {
+      "epoch": 1.9456542321530015,
+      "grad_norm": 0.35266685485839844,
+      "learning_rate": 1.2217383071387996e-05,
+      "loss": 0.1083,
+      "step": 20550
+    },
+    {
+      "epoch": 1.9466010225336112,
+      "grad_norm": 0.4260408878326416,
+      "learning_rate": 1.2213595909865557e-05,
+      "loss": 0.1085,
+      "step": 20560
+    },
+    {
+      "epoch": 1.9475478129142207,
+      "grad_norm": 0.3600156009197235,
+      "learning_rate": 1.2209808748343117e-05,
+      "loss": 0.1134,
+      "step": 20570
+    },
+    {
+      "epoch": 1.9484946032948305,
+      "grad_norm": 0.3416011929512024,
+      "learning_rate": 1.220602158682068e-05,
+      "loss": 0.097,
+      "step": 20580
+    },
+    {
+      "epoch": 1.9494413936754402,
+      "grad_norm": 0.3432440459728241,
+      "learning_rate": 1.220223442529824e-05,
+      "loss": 0.1001,
+      "step": 20590
+    },
+    {
+      "epoch": 1.95038818405605,
+      "grad_norm": 0.32206058502197266,
+      "learning_rate": 1.2198447263775801e-05,
+      "loss": 0.1012,
+      "step": 20600
+    },
+    {
+      "epoch": 1.9513349744366597,
+      "grad_norm": 0.382606565952301,
+      "learning_rate": 1.2194660102253361e-05,
+      "loss": 0.1126,
+      "step": 20610
+    },
+    {
+      "epoch": 1.9522817648172694,
+      "grad_norm": 0.3285946249961853,
+      "learning_rate": 1.2190872940730923e-05,
+      "loss": 0.1086,
+      "step": 20620
+    },
+    {
+      "epoch": 1.9532285551978792,
+      "grad_norm": 0.30580756068229675,
+      "learning_rate": 1.2187085779208484e-05,
+      "loss": 0.1067,
+      "step": 20630
+    },
+    {
+      "epoch": 1.954175345578489,
+      "grad_norm": 0.31543785333633423,
+      "learning_rate": 1.2183298617686046e-05,
+      "loss": 0.1095,
+      "step": 20640
+    },
+    {
+      "epoch": 1.9551221359590987,
+      "grad_norm": 0.4250805974006653,
+      "learning_rate": 1.2179511456163606e-05,
+      "loss": 0.1133,
+      "step": 20650
+    },
+    {
+      "epoch": 1.9560689263397084,
+      "grad_norm": 0.3352222442626953,
+      "learning_rate": 1.2175724294641166e-05,
+      "loss": 0.0984,
+      "step": 20660
+    },
+    {
+      "epoch": 1.9570157167203182,
+      "grad_norm": 0.3887184262275696,
+      "learning_rate": 1.2171937133118728e-05,
+      "loss": 0.1153,
+      "step": 20670
+    },
+    {
+      "epoch": 1.957962507100928,
+      "grad_norm": 0.37712177634239197,
+      "learning_rate": 1.216814997159629e-05,
+      "loss": 0.1165,
+      "step": 20680
+    },
+    {
+      "epoch": 1.9589092974815376,
+      "grad_norm": 0.3524229824542999,
+      "learning_rate": 1.2164362810073851e-05,
+      "loss": 0.1024,
+      "step": 20690
+    },
+    {
+      "epoch": 1.9598560878621472,
+      "grad_norm": 0.3581532835960388,
+      "learning_rate": 1.2160575648551412e-05,
+      "loss": 0.1115,
+      "step": 20700
+    },
+    {
+      "epoch": 1.960802878242757,
+      "grad_norm": 0.30578964948654175,
+      "learning_rate": 1.2156788487028972e-05,
+      "loss": 0.1082,
+      "step": 20710
+    },
+    {
+      "epoch": 1.9617496686233666,
+      "grad_norm": 0.3593074381351471,
+      "learning_rate": 1.2153001325506535e-05,
+      "loss": 0.1068,
+      "step": 20720
+    },
+    {
+      "epoch": 1.9626964590039764,
+      "grad_norm": 0.3663315773010254,
+      "learning_rate": 1.2149214163984095e-05,
+      "loss": 0.1089,
+      "step": 20730
+    },
+    {
+      "epoch": 1.9636432493845861,
+      "grad_norm": 0.2964952886104584,
+      "learning_rate": 1.2145427002461657e-05,
+      "loss": 0.1073,
+      "step": 20740
+    },
+    {
+      "epoch": 1.9645900397651959,
+      "grad_norm": 0.38032329082489014,
+      "learning_rate": 1.2141639840939217e-05,
+      "loss": 0.1098,
+      "step": 20750
+    },
+    {
+      "epoch": 1.9655368301458056,
+      "grad_norm": 0.4049619436264038,
+      "learning_rate": 1.2137852679416777e-05,
+      "loss": 0.1162,
+      "step": 20760
+    },
+    {
+      "epoch": 1.9664836205264153,
+      "grad_norm": 0.39563053846359253,
+      "learning_rate": 1.213406551789434e-05,
+      "loss": 0.1091,
+      "step": 20770
+    },
+    {
+      "epoch": 1.967430410907025,
+      "grad_norm": 0.2906529903411865,
+      "learning_rate": 1.21302783563719e-05,
+      "loss": 0.1143,
+      "step": 20780
+    },
+    {
+      "epoch": 1.9683772012876348,
+      "grad_norm": 0.30797553062438965,
+      "learning_rate": 1.212649119484946e-05,
+      "loss": 0.104,
+      "step": 20790
+    },
+    {
+      "epoch": 1.9693239916682446,
+      "grad_norm": 0.3349768817424774,
+      "learning_rate": 1.2122704033327022e-05,
+      "loss": 0.1071,
+      "step": 20800
+    },
+    {
+      "epoch": 1.9702707820488543,
+      "grad_norm": 0.3945980370044708,
+      "learning_rate": 1.2118916871804582e-05,
+      "loss": 0.1191,
+      "step": 20810
+    },
+    {
+      "epoch": 1.971217572429464,
+      "grad_norm": 0.35294216871261597,
+      "learning_rate": 1.2115129710282146e-05,
+      "loss": 0.105,
+      "step": 20820
+    },
+    {
+      "epoch": 1.9721643628100738,
+      "grad_norm": 0.38833561539649963,
+      "learning_rate": 1.2111342548759706e-05,
+      "loss": 0.12,
+      "step": 20830
+    },
+    {
+      "epoch": 1.9731111531906835,
+      "grad_norm": 0.2893434464931488,
+      "learning_rate": 1.2107555387237266e-05,
+      "loss": 0.104,
+      "step": 20840
+    },
+    {
+      "epoch": 1.9740579435712933,
+      "grad_norm": 0.2558499276638031,
+      "learning_rate": 1.2103768225714828e-05,
+      "loss": 0.1068,
+      "step": 20850
+    },
+    {
+      "epoch": 1.975004733951903,
+      "grad_norm": 0.2818835973739624,
+      "learning_rate": 1.209998106419239e-05,
+      "loss": 0.11,
+      "step": 20860
+    },
+    {
+      "epoch": 1.9759515243325128,
+      "grad_norm": 0.29642802476882935,
+      "learning_rate": 1.2096193902669951e-05,
+      "loss": 0.1008,
+      "step": 20870
+    },
+    {
+      "epoch": 1.9768983147131225,
+      "grad_norm": 0.3638819456100464,
+      "learning_rate": 1.2092406741147511e-05,
+      "loss": 0.1125,
+      "step": 20880
+    },
+    {
+      "epoch": 1.9778451050937322,
+      "grad_norm": 0.4232320487499237,
+      "learning_rate": 1.2088619579625071e-05,
+      "loss": 0.1142,
+      "step": 20890
+    },
+    {
+      "epoch": 1.978791895474342,
+      "grad_norm": 0.26880526542663574,
+      "learning_rate": 1.2084832418102633e-05,
+      "loss": 0.1018,
+      "step": 20900
+    },
+    {
+      "epoch": 1.9797386858549517,
+      "grad_norm": 0.35345229506492615,
+      "learning_rate": 1.2081045256580195e-05,
+      "loss": 0.0994,
+      "step": 20910
+    },
+    {
+      "epoch": 1.9806854762355615,
+      "grad_norm": 0.30742210149765015,
+      "learning_rate": 1.2077258095057756e-05,
+      "loss": 0.1117,
+      "step": 20920
+    },
+    {
+      "epoch": 1.9816322666161712,
+      "grad_norm": 0.37625306844711304,
+      "learning_rate": 1.2073470933535316e-05,
+      "loss": 0.106,
+      "step": 20930
+    },
+    {
+      "epoch": 1.982579056996781,
+      "grad_norm": 0.35394927859306335,
+      "learning_rate": 1.2069683772012876e-05,
+      "loss": 0.1196,
+      "step": 20940
+    },
+    {
+      "epoch": 1.9835258473773907,
+      "grad_norm": 0.4539976119995117,
+      "learning_rate": 1.2065896610490438e-05,
+      "loss": 0.1081,
+      "step": 20950
+    },
+    {
+      "epoch": 1.9844726377580004,
+      "grad_norm": 0.37135881185531616,
+      "learning_rate": 1.2062109448968e-05,
+      "loss": 0.1178,
+      "step": 20960
+    },
+    {
+      "epoch": 1.9854194281386102,
+      "grad_norm": 0.29517096281051636,
+      "learning_rate": 1.205832228744556e-05,
+      "loss": 0.1093,
+      "step": 20970
+    },
+    {
+      "epoch": 1.98636621851922,
+      "grad_norm": 0.3522278666496277,
+      "learning_rate": 1.2054535125923122e-05,
+      "loss": 0.1173,
+      "step": 20980
+    },
+    {
+      "epoch": 1.9873130088998296,
+      "grad_norm": 0.3382992744445801,
+      "learning_rate": 1.2050747964400682e-05,
+      "loss": 0.1081,
+      "step": 20990
+    },
+    {
+      "epoch": 1.9882597992804394,
+      "grad_norm": 0.3614194989204407,
+      "learning_rate": 1.2046960802878245e-05,
+      "loss": 0.103,
+      "step": 21000
+    },
+    {
+      "epoch": 1.9892065896610491,
+      "grad_norm": 0.29812687635421753,
+      "learning_rate": 1.2043173641355805e-05,
+      "loss": 0.105,
+      "step": 21010
+    },
+    {
+      "epoch": 1.9901533800416589,
+      "grad_norm": 0.5184694528579712,
+      "learning_rate": 1.2039386479833365e-05,
+      "loss": 0.1009,
+      "step": 21020
+    },
+    {
+      "epoch": 1.9911001704222686,
+      "grad_norm": 0.4298660159111023,
+      "learning_rate": 1.2035599318310927e-05,
+      "loss": 0.1127,
+      "step": 21030
+    },
+    {
+      "epoch": 1.9920469608028784,
+      "grad_norm": 0.32132789492607117,
+      "learning_rate": 1.2031812156788487e-05,
+      "loss": 0.1092,
+      "step": 21040
+    },
+    {
+      "epoch": 1.992993751183488,
+      "grad_norm": 0.4802221953868866,
+      "learning_rate": 1.202802499526605e-05,
+      "loss": 0.1194,
+      "step": 21050
+    },
+    {
+      "epoch": 1.9939405415640978,
+      "grad_norm": 0.32533472776412964,
+      "learning_rate": 1.202423783374361e-05,
+      "loss": 0.0933,
+      "step": 21060
+    },
+    {
+      "epoch": 1.9948873319447076,
+      "grad_norm": 0.3043247163295746,
+      "learning_rate": 1.202045067222117e-05,
+      "loss": 0.1093,
+      "step": 21070
+    },
+    {
+      "epoch": 1.9958341223253173,
+      "grad_norm": 0.36251574754714966,
+      "learning_rate": 1.2016663510698732e-05,
+      "loss": 0.1098,
+      "step": 21080
+    },
+    {
+      "epoch": 1.996780912705927,
+      "grad_norm": 0.40744641423225403,
+      "learning_rate": 1.2012876349176292e-05,
+      "loss": 0.1181,
+      "step": 21090
+    },
+    {
+      "epoch": 1.9977277030865368,
+      "grad_norm": 0.2982351779937744,
+      "learning_rate": 1.2009089187653856e-05,
+      "loss": 0.0982,
+      "step": 21100
+    },
+    {
+      "epoch": 1.9986744934671465,
+      "grad_norm": 0.4201706051826477,
+      "learning_rate": 1.2005302026131416e-05,
+      "loss": 0.1118,
+      "step": 21110
+    },
+    {
+      "epoch": 1.999621283847756,
+      "grad_norm": 0.26378563046455383,
+      "learning_rate": 1.2001514864608976e-05,
+      "loss": 0.1048,
+      "step": 21120
+    },
+    {
+      "epoch": 2.0,
+      "eval_f1_micro": 0.35933516585247743,
+      "eval_loss": 0.11070390790700912,
+      "eval_precision": 0.6185036281623848,
+      "eval_recall": 0.2532268101250577,
+      "eval_runtime": 339.0084,
+      "eval_samples_per_second": 124.613,
+      "eval_steps_per_second": 7.79,
+      "step": 21124
+    },
+    {
+      "epoch": 2.000568074228366,
+      "grad_norm": 0.3789435625076294,
+      "learning_rate": 1.1997727703086538e-05,
+      "loss": 0.1128,
+      "step": 21130
+    },
+    {
+      "epoch": 2.0015148646089758,
+      "grad_norm": 0.3707931637763977,
+      "learning_rate": 1.19939405415641e-05,
+      "loss": 0.1088,
+      "step": 21140
+    },
+    {
+      "epoch": 2.0024616549895855,
+      "grad_norm": 0.30625054240226746,
+      "learning_rate": 1.199015338004166e-05,
+      "loss": 0.0942,
+      "step": 21150
+    },
+    {
+      "epoch": 2.0034084453701952,
+      "grad_norm": 0.28570687770843506,
+      "learning_rate": 1.1986366218519221e-05,
+      "loss": 0.1063,
+      "step": 21160
+    },
+    {
+      "epoch": 2.004355235750805,
+      "grad_norm": 0.313829630613327,
+      "learning_rate": 1.1982579056996781e-05,
+      "loss": 0.0991,
+      "step": 21170
+    },
+    {
+      "epoch": 2.0053020261314143,
+      "grad_norm": 0.2512275278568268,
+      "learning_rate": 1.1978791895474341e-05,
+      "loss": 0.097,
+      "step": 21180
+    },
+    {
+      "epoch": 2.006248816512024,
+      "grad_norm": 0.3135952055454254,
+      "learning_rate": 1.1975004733951905e-05,
+      "loss": 0.1055,
+      "step": 21190
+    },
+    {
+      "epoch": 2.0071956068926338,
+      "grad_norm": 0.35371828079223633,
+      "learning_rate": 1.1971217572429465e-05,
+      "loss": 0.1033,
+      "step": 21200
+    },
+    {
+      "epoch": 2.0081423972732435,
+      "grad_norm": 0.31105470657348633,
+      "learning_rate": 1.1967430410907027e-05,
+      "loss": 0.1014,
+      "step": 21210
+    },
+    {
+      "epoch": 2.0090891876538532,
+      "grad_norm": 0.32313773036003113,
+      "learning_rate": 1.1963643249384587e-05,
+      "loss": 0.094,
+      "step": 21220
+    },
+    {
+      "epoch": 2.010035978034463,
+      "grad_norm": 0.31268981099128723,
+      "learning_rate": 1.1959856087862147e-05,
+      "loss": 0.0951,
+      "step": 21230
+    },
+    {
+      "epoch": 2.0109827684150727,
+      "grad_norm": 0.38261574506759644,
+      "learning_rate": 1.195606892633971e-05,
+      "loss": 0.0923,
+      "step": 21240
+    },
+    {
+      "epoch": 2.0119295587956825,
+      "grad_norm": 0.3849813640117645,
+      "learning_rate": 1.195228176481727e-05,
+      "loss": 0.1121,
+      "step": 21250
+    },
+    {
+      "epoch": 2.012876349176292,
+      "grad_norm": 0.4521833062171936,
+      "learning_rate": 1.1948494603294832e-05,
+      "loss": 0.1121,
+      "step": 21260
+    },
+    {
+      "epoch": 2.013823139556902,
+      "grad_norm": 0.3331480622291565,
+      "learning_rate": 1.1944707441772392e-05,
+      "loss": 0.0995,
+      "step": 21270
+    },
+    {
+      "epoch": 2.0147699299375117,
+      "grad_norm": 0.33220431208610535,
+      "learning_rate": 1.1940920280249955e-05,
+      "loss": 0.1018,
+      "step": 21280
+    },
+    {
+      "epoch": 2.0157167203181214,
+      "grad_norm": 0.37220412492752075,
+      "learning_rate": 1.1937133118727515e-05,
+      "loss": 0.1042,
+      "step": 21290
+    },
+    {
+      "epoch": 2.016663510698731,
+      "grad_norm": 0.3874574303627014,
+      "learning_rate": 1.1933345957205075e-05,
+      "loss": 0.1085,
+      "step": 21300
+    },
+    {
+      "epoch": 2.017610301079341,
+      "grad_norm": 0.3492327928543091,
+      "learning_rate": 1.1929558795682637e-05,
+      "loss": 0.098,
+      "step": 21310
+    },
+    {
+      "epoch": 2.0185570914599507,
+      "grad_norm": 0.3867335617542267,
+      "learning_rate": 1.1925771634160197e-05,
+      "loss": 0.1019,
+      "step": 21320
+    },
+    {
+      "epoch": 2.0195038818405604,
+      "grad_norm": 0.386625200510025,
+      "learning_rate": 1.1921984472637759e-05,
+      "loss": 0.1014,
+      "step": 21330
+    },
+    {
+      "epoch": 2.02045067222117,
+      "grad_norm": 0.3031746447086334,
+      "learning_rate": 1.191819731111532e-05,
+      "loss": 0.1009,
+      "step": 21340
+    },
+    {
+      "epoch": 2.02139746260178,
+      "grad_norm": 0.3026321530342102,
+      "learning_rate": 1.191441014959288e-05,
+      "loss": 0.1019,
+      "step": 21350
+    },
+    {
+      "epoch": 2.0223442529823896,
+      "grad_norm": 0.35251784324645996,
+      "learning_rate": 1.1910622988070441e-05,
+      "loss": 0.095,
+      "step": 21360
+    },
+    {
+      "epoch": 2.0232910433629994,
+      "grad_norm": 0.34639063477516174,
+      "learning_rate": 1.1906835826548003e-05,
+      "loss": 0.1065,
+      "step": 21370
+    },
+    {
+      "epoch": 2.024237833743609,
+      "grad_norm": 0.34346747398376465,
+      "learning_rate": 1.1903048665025564e-05,
+      "loss": 0.1038,
+      "step": 21380
+    },
+    {
+      "epoch": 2.025184624124219,
+      "grad_norm": 0.3798876702785492,
+      "learning_rate": 1.1899261503503126e-05,
+      "loss": 0.1022,
+      "step": 21390
+    },
+    {
+      "epoch": 2.0261314145048286,
+      "grad_norm": 0.4110786020755768,
+      "learning_rate": 1.1895474341980686e-05,
+      "loss": 0.1039,
+      "step": 21400
+    },
+    {
+      "epoch": 2.0270782048854383,
+      "grad_norm": 0.3044571578502655,
+      "learning_rate": 1.1891687180458246e-05,
+      "loss": 0.0952,
+      "step": 21410
+    },
+    {
+      "epoch": 2.028024995266048,
+      "grad_norm": 0.39002007246017456,
+      "learning_rate": 1.188790001893581e-05,
+      "loss": 0.1061,
+      "step": 21420
+    },
+    {
+      "epoch": 2.028971785646658,
+      "grad_norm": 0.3900294899940491,
+      "learning_rate": 1.188411285741337e-05,
+      "loss": 0.11,
+      "step": 21430
+    },
+    {
+      "epoch": 2.0299185760272676,
+      "grad_norm": 0.36899328231811523,
+      "learning_rate": 1.1880325695890931e-05,
+      "loss": 0.0922,
+      "step": 21440
+    },
+    {
+      "epoch": 2.0308653664078773,
+      "grad_norm": 0.35091322660446167,
+      "learning_rate": 1.1876538534368491e-05,
+      "loss": 0.0999,
+      "step": 21450
+    },
+    {
+      "epoch": 2.031812156788487,
+      "grad_norm": 0.40123164653778076,
+      "learning_rate": 1.1872751372846052e-05,
+      "loss": 0.0946,
+      "step": 21460
+    },
+    {
+      "epoch": 2.0327589471690968,
+      "grad_norm": 0.2955148220062256,
+      "learning_rate": 1.1868964211323615e-05,
+      "loss": 0.1001,
+      "step": 21470
+    },
+    {
+      "epoch": 2.0337057375497065,
+      "grad_norm": 0.38906633853912354,
+      "learning_rate": 1.1865177049801175e-05,
+      "loss": 0.1002,
+      "step": 21480
+    },
+    {
+      "epoch": 2.0346525279303163,
+      "grad_norm": 0.46951401233673096,
+      "learning_rate": 1.1861389888278737e-05,
+      "loss": 0.108,
+      "step": 21490
+    },
+    {
+      "epoch": 2.035599318310926,
+      "grad_norm": 0.46195003390312195,
+      "learning_rate": 1.1857602726756297e-05,
+      "loss": 0.1145,
+      "step": 21500
+    },
+    {
+      "epoch": 2.0365461086915357,
+      "grad_norm": 0.4008125364780426,
+      "learning_rate": 1.1853815565233857e-05,
+      "loss": 0.1036,
+      "step": 21510
+    },
+    {
+      "epoch": 2.0374928990721455,
+      "grad_norm": 0.4968140721321106,
+      "learning_rate": 1.185002840371142e-05,
+      "loss": 0.1086,
+      "step": 21520
+    },
+    {
+      "epoch": 2.038439689452755,
+      "grad_norm": 0.39275693893432617,
+      "learning_rate": 1.184624124218898e-05,
+      "loss": 0.109,
+      "step": 21530
+    },
+    {
+      "epoch": 2.039386479833365,
+      "grad_norm": 0.34367573261260986,
+      "learning_rate": 1.184245408066654e-05,
+      "loss": 0.1009,
+      "step": 21540
+    },
+    {
+      "epoch": 2.0403332702139747,
+      "grad_norm": 0.4184512794017792,
+      "learning_rate": 1.1838666919144102e-05,
+      "loss": 0.1024,
+      "step": 21550
+    },
+    {
+      "epoch": 2.0412800605945844,
+      "grad_norm": 0.4419008493423462,
+      "learning_rate": 1.1834879757621664e-05,
+      "loss": 0.1033,
+      "step": 21560
+    },
+    {
+      "epoch": 2.042226850975194,
+      "grad_norm": 0.37311434745788574,
+      "learning_rate": 1.1831092596099226e-05,
+      "loss": 0.0988,
+      "step": 21570
+    },
+    {
+      "epoch": 2.043173641355804,
+      "grad_norm": 0.39084023237228394,
+      "learning_rate": 1.1827305434576786e-05,
+      "loss": 0.0997,
+      "step": 21580
+    },
+    {
+      "epoch": 2.0441204317364137,
+      "grad_norm": 0.3225145637989044,
+      "learning_rate": 1.1823518273054346e-05,
+      "loss": 0.1041,
+      "step": 21590
+    },
+    {
+      "epoch": 2.0450672221170234,
+      "grad_norm": 0.44645923376083374,
+      "learning_rate": 1.1819731111531907e-05,
+      "loss": 0.1047,
+      "step": 21600
+    },
+    {
+      "epoch": 2.046014012497633,
+      "grad_norm": 0.39939242601394653,
+      "learning_rate": 1.181594395000947e-05,
+      "loss": 0.097,
+      "step": 21610
+    },
+    {
+      "epoch": 2.046960802878243,
+      "grad_norm": 0.41991719603538513,
+      "learning_rate": 1.1812156788487031e-05,
+      "loss": 0.1054,
+      "step": 21620
+    },
+    {
+      "epoch": 2.0479075932588526,
+      "grad_norm": 0.4075891971588135,
+      "learning_rate": 1.1808369626964591e-05,
+      "loss": 0.0975,
+      "step": 21630
+    },
+    {
+      "epoch": 2.0488543836394624,
+      "grad_norm": 0.3350875675678253,
+      "learning_rate": 1.1804582465442151e-05,
+      "loss": 0.0968,
+      "step": 21640
+    },
+    {
+      "epoch": 2.049801174020072,
+      "grad_norm": 0.33716604113578796,
+      "learning_rate": 1.1800795303919713e-05,
+      "loss": 0.1,
+      "step": 21650
+    },
+    {
+      "epoch": 2.050747964400682,
+      "grad_norm": 0.37090936303138733,
+      "learning_rate": 1.1797008142397275e-05,
+      "loss": 0.105,
+      "step": 21660
+    },
+    {
+      "epoch": 2.0516947547812916,
+      "grad_norm": 0.33265426754951477,
+      "learning_rate": 1.1793220980874836e-05,
+      "loss": 0.1001,
+      "step": 21670
+    },
+    {
+      "epoch": 2.0526415451619013,
+      "grad_norm": 0.3445303440093994,
+      "learning_rate": 1.1789433819352396e-05,
+      "loss": 0.0987,
+      "step": 21680
+    },
+    {
+      "epoch": 2.053588335542511,
+      "grad_norm": 0.42871901392936707,
+      "learning_rate": 1.1785646657829956e-05,
+      "loss": 0.0973,
+      "step": 21690
+    },
+    {
+      "epoch": 2.054535125923121,
+      "grad_norm": 0.3626950979232788,
+      "learning_rate": 1.178185949630752e-05,
+      "loss": 0.1076,
+      "step": 21700
+    },
+    {
+      "epoch": 2.0554819163037306,
+      "grad_norm": 0.4158441126346588,
+      "learning_rate": 1.177807233478508e-05,
+      "loss": 0.1075,
+      "step": 21710
+    },
+    {
+      "epoch": 2.0564287066843403,
+      "grad_norm": 0.43088018894195557,
+      "learning_rate": 1.177428517326264e-05,
+      "loss": 0.0975,
+      "step": 21720
+    },
+    {
+      "epoch": 2.0573754970649496,
+      "grad_norm": 0.3257809281349182,
+      "learning_rate": 1.1770498011740202e-05,
+      "loss": 0.1001,
+      "step": 21730
+    },
+    {
+      "epoch": 2.0583222874455593,
+      "grad_norm": 0.3736395537853241,
+      "learning_rate": 1.1766710850217762e-05,
+      "loss": 0.0882,
+      "step": 21740
+    },
+    {
+      "epoch": 2.059269077826169,
+      "grad_norm": 0.3264060318470001,
+      "learning_rate": 1.1762923688695325e-05,
+      "loss": 0.1097,
+      "step": 21750
+    },
+    {
+      "epoch": 2.060215868206779,
+      "grad_norm": 0.4029650092124939,
+      "learning_rate": 1.1759136527172885e-05,
+      "loss": 0.105,
+      "step": 21760
+    },
+    {
+      "epoch": 2.0611626585873886,
+      "grad_norm": 0.4287700653076172,
+      "learning_rate": 1.1755349365650445e-05,
+      "loss": 0.1034,
+      "step": 21770
+    },
+    {
+      "epoch": 2.0621094489679983,
+      "grad_norm": 0.43112558126449585,
+      "learning_rate": 1.1751562204128007e-05,
+      "loss": 0.1001,
+      "step": 21780
+    },
+    {
+      "epoch": 2.063056239348608,
+      "grad_norm": 0.34501007199287415,
+      "learning_rate": 1.1747775042605567e-05,
+      "loss": 0.0823,
+      "step": 21790
+    },
+    {
+      "epoch": 2.064003029729218,
+      "grad_norm": 0.36909836530685425,
+      "learning_rate": 1.174398788108313e-05,
+      "loss": 0.0943,
+      "step": 21800
+    },
+    {
+      "epoch": 2.0649498201098275,
+      "grad_norm": 0.3652644753456116,
+      "learning_rate": 1.174020071956069e-05,
+      "loss": 0.0984,
+      "step": 21810
+    },
+    {
+      "epoch": 2.0658966104904373,
+      "grad_norm": 0.40247103571891785,
+      "learning_rate": 1.173641355803825e-05,
+      "loss": 0.1,
+      "step": 21820
+    },
+    {
+      "epoch": 2.066843400871047,
+      "grad_norm": 0.437529057264328,
+      "learning_rate": 1.1732626396515812e-05,
+      "loss": 0.0996,
+      "step": 21830
+    },
+    {
+      "epoch": 2.0677901912516568,
+      "grad_norm": 0.4239822030067444,
+      "learning_rate": 1.1728839234993374e-05,
+      "loss": 0.1022,
+      "step": 21840
+    },
+    {
+      "epoch": 2.0687369816322665,
+      "grad_norm": 0.3921540677547455,
+      "learning_rate": 1.1725052073470936e-05,
+      "loss": 0.1009,
+      "step": 21850
+    },
+    {
+      "epoch": 2.0696837720128762,
+      "grad_norm": 0.32198435068130493,
+      "learning_rate": 1.1721264911948496e-05,
+      "loss": 0.0923,
+      "step": 21860
+    },
+    {
+      "epoch": 2.070630562393486,
+      "grad_norm": 0.4083239734172821,
+      "learning_rate": 1.1717477750426056e-05,
+      "loss": 0.1084,
+      "step": 21870
+    },
+    {
+      "epoch": 2.0715773527740957,
+      "grad_norm": 0.47515833377838135,
+      "learning_rate": 1.1713690588903616e-05,
+      "loss": 0.0953,
+      "step": 21880
+    },
+    {
+      "epoch": 2.0725241431547055,
+      "grad_norm": 0.3879965841770172,
+      "learning_rate": 1.170990342738118e-05,
+      "loss": 0.0956,
+      "step": 21890
+    },
+    {
+      "epoch": 2.073470933535315,
+      "grad_norm": 0.34762468934059143,
+      "learning_rate": 1.170611626585874e-05,
+      "loss": 0.1116,
+      "step": 21900
+    },
+    {
+      "epoch": 2.074417723915925,
+      "grad_norm": 0.3374441862106323,
+      "learning_rate": 1.1702329104336301e-05,
+      "loss": 0.1126,
+      "step": 21910
+    },
+    {
+      "epoch": 2.0753645142965347,
+      "grad_norm": 0.3270218074321747,
+      "learning_rate": 1.1698541942813861e-05,
+      "loss": 0.0973,
+      "step": 21920
+    },
+    {
+      "epoch": 2.0763113046771444,
+      "grad_norm": 0.39837753772735596,
+      "learning_rate": 1.1694754781291421e-05,
+      "loss": 0.1041,
+      "step": 21930
+    },
+    {
+      "epoch": 2.077258095057754,
+      "grad_norm": 0.3634657561779022,
+      "learning_rate": 1.1690967619768985e-05,
+      "loss": 0.1133,
+      "step": 21940
+    },
+    {
+      "epoch": 2.078204885438364,
+      "grad_norm": 0.41098886728286743,
+      "learning_rate": 1.1687180458246545e-05,
+      "loss": 0.1103,
+      "step": 21950
+    },
+    {
+      "epoch": 2.0791516758189736,
+      "grad_norm": 0.3962465822696686,
+      "learning_rate": 1.1683393296724107e-05,
+      "loss": 0.0996,
+      "step": 21960
+    },
+    {
+      "epoch": 2.0800984661995834,
+      "grad_norm": 0.34035059809684753,
+      "learning_rate": 1.1679606135201667e-05,
+      "loss": 0.0988,
+      "step": 21970
+    },
+    {
+      "epoch": 2.081045256580193,
+      "grad_norm": 0.5173091888427734,
+      "learning_rate": 1.167581897367923e-05,
+      "loss": 0.1018,
+      "step": 21980
+    },
+    {
+      "epoch": 2.081992046960803,
+      "grad_norm": 0.4123232662677765,
+      "learning_rate": 1.167203181215679e-05,
+      "loss": 0.102,
+      "step": 21990
+    },
+    {
+      "epoch": 2.0829388373414126,
+      "grad_norm": 0.4414355158805847,
+      "learning_rate": 1.166824465063435e-05,
+      "loss": 0.1068,
+      "step": 22000
+    },
+    {
+      "epoch": 2.0838856277220223,
+      "grad_norm": 0.4595598578453064,
+      "learning_rate": 1.1664457489111912e-05,
+      "loss": 0.1072,
+      "step": 22010
+    },
+    {
+      "epoch": 2.084832418102632,
+      "grad_norm": 0.33270683884620667,
+      "learning_rate": 1.1660670327589472e-05,
+      "loss": 0.1039,
+      "step": 22020
+    },
+    {
+      "epoch": 2.085779208483242,
+      "grad_norm": 0.2970442771911621,
+      "learning_rate": 1.1656883166067035e-05,
+      "loss": 0.0969,
+      "step": 22030
+    },
+    {
+      "epoch": 2.0867259988638516,
+      "grad_norm": 0.3258838355541229,
+      "learning_rate": 1.1653096004544595e-05,
+      "loss": 0.097,
+      "step": 22040
+    },
+    {
+      "epoch": 2.0876727892444613,
+      "grad_norm": 0.3754783272743225,
+      "learning_rate": 1.1649308843022155e-05,
+      "loss": 0.1031,
+      "step": 22050
+    },
+    {
+      "epoch": 2.088619579625071,
+      "grad_norm": 0.38050562143325806,
+      "learning_rate": 1.1645521681499715e-05,
+      "loss": 0.1027,
+      "step": 22060
+    },
+    {
+      "epoch": 2.089566370005681,
+      "grad_norm": 0.2980318069458008,
+      "learning_rate": 1.1641734519977277e-05,
+      "loss": 0.0978,
+      "step": 22070
+    },
+    {
+      "epoch": 2.0905131603862905,
+      "grad_norm": 0.4399225115776062,
+      "learning_rate": 1.1637947358454839e-05,
+      "loss": 0.1075,
+      "step": 22080
+    },
+    {
+      "epoch": 2.0914599507669003,
+      "grad_norm": 0.4211890697479248,
+      "learning_rate": 1.16341601969324e-05,
+      "loss": 0.1011,
+      "step": 22090
+    },
+    {
+      "epoch": 2.09240674114751,
+      "grad_norm": 0.4464966058731079,
+      "learning_rate": 1.163037303540996e-05,
+      "loss": 0.1032,
+      "step": 22100
+    },
+    {
+      "epoch": 2.0933535315281198,
+      "grad_norm": 0.40521979331970215,
+      "learning_rate": 1.162658587388752e-05,
+      "loss": 0.1041,
+      "step": 22110
+    },
+    {
+      "epoch": 2.0943003219087295,
+      "grad_norm": 0.33831852674484253,
+      "learning_rate": 1.1622798712365084e-05,
+      "loss": 0.102,
+      "step": 22120
+    },
+    {
+      "epoch": 2.0952471122893392,
+      "grad_norm": 0.3584233522415161,
+      "learning_rate": 1.1619011550842644e-05,
+      "loss": 0.1055,
+      "step": 22130
+    },
+    {
+      "epoch": 2.096193902669949,
+      "grad_norm": 0.3828630745410919,
+      "learning_rate": 1.1615224389320206e-05,
+      "loss": 0.115,
+      "step": 22140
+    },
+    {
+      "epoch": 2.0971406930505587,
+      "grad_norm": 0.4157094359397888,
+      "learning_rate": 1.1611437227797766e-05,
+      "loss": 0.1081,
+      "step": 22150
+    },
+    {
+      "epoch": 2.0980874834311685,
+      "grad_norm": 0.39398008584976196,
+      "learning_rate": 1.1607650066275326e-05,
+      "loss": 0.0993,
+      "step": 22160
+    },
+    {
+      "epoch": 2.099034273811778,
+      "grad_norm": 0.4000839591026306,
+      "learning_rate": 1.160386290475289e-05,
+      "loss": 0.0977,
+      "step": 22170
+    },
+    {
+      "epoch": 2.099981064192388,
+      "grad_norm": 0.39173951745033264,
+      "learning_rate": 1.160007574323045e-05,
+      "loss": 0.0993,
+      "step": 22180
+    },
+    {
+      "epoch": 2.1009278545729977,
+      "grad_norm": 0.43835166096687317,
+      "learning_rate": 1.1596288581708011e-05,
+      "loss": 0.1078,
+      "step": 22190
+    },
+    {
+      "epoch": 2.1018746449536074,
+      "grad_norm": 0.383684903383255,
+      "learning_rate": 1.1592501420185571e-05,
+      "loss": 0.0987,
+      "step": 22200
+    },
+    {
+      "epoch": 2.102821435334217,
+      "grad_norm": 0.3146076798439026,
+      "learning_rate": 1.1588714258663131e-05,
+      "loss": 0.0955,
+      "step": 22210
+    },
+    {
+      "epoch": 2.103768225714827,
+      "grad_norm": 0.4743645191192627,
+      "learning_rate": 1.1584927097140695e-05,
+      "loss": 0.1029,
+      "step": 22220
+    },
+    {
+      "epoch": 2.1047150160954367,
+      "grad_norm": 0.4409228563308716,
+      "learning_rate": 1.1581139935618255e-05,
+      "loss": 0.1088,
+      "step": 22230
+    },
+    {
+      "epoch": 2.1056618064760464,
+      "grad_norm": 0.36643436551094055,
+      "learning_rate": 1.1577352774095815e-05,
+      "loss": 0.0978,
+      "step": 22240
+    },
+    {
+      "epoch": 2.106608596856656,
+      "grad_norm": 0.33801764249801636,
+      "learning_rate": 1.1573565612573377e-05,
+      "loss": 0.0965,
+      "step": 22250
+    },
+    {
+      "epoch": 2.107555387237266,
+      "grad_norm": 0.3886876404285431,
+      "learning_rate": 1.1569778451050938e-05,
+      "loss": 0.1043,
+      "step": 22260
+    },
+    {
+      "epoch": 2.1085021776178756,
+      "grad_norm": 0.36893683671951294,
+      "learning_rate": 1.15659912895285e-05,
+      "loss": 0.1088,
+      "step": 22270
+    },
+    {
+      "epoch": 2.1094489679984854,
+      "grad_norm": 0.4243239760398865,
+      "learning_rate": 1.156220412800606e-05,
+      "loss": 0.1071,
+      "step": 22280
+    },
+    {
+      "epoch": 2.1103957583790947,
+      "grad_norm": 0.38450294733047485,
+      "learning_rate": 1.155841696648362e-05,
+      "loss": 0.1044,
+      "step": 22290
+    },
+    {
+      "epoch": 2.1113425487597044,
+      "grad_norm": 0.4414058327674866,
+      "learning_rate": 1.1554629804961182e-05,
+      "loss": 0.0988,
+      "step": 22300
+    },
+    {
+      "epoch": 2.112289339140314,
+      "grad_norm": 0.5164914131164551,
+      "learning_rate": 1.1550842643438744e-05,
+      "loss": 0.1067,
+      "step": 22310
+    },
+    {
+      "epoch": 2.113236129520924,
+      "grad_norm": 0.5250256657600403,
+      "learning_rate": 1.1547055481916306e-05,
+      "loss": 0.0878,
+      "step": 22320
+    },
+    {
+      "epoch": 2.1141829199015336,
+      "grad_norm": 0.3771630823612213,
+      "learning_rate": 1.1543268320393866e-05,
+      "loss": 0.0995,
+      "step": 22330
+    },
+    {
+      "epoch": 2.1151297102821434,
+      "grad_norm": 0.3611592948436737,
+      "learning_rate": 1.1539481158871426e-05,
+      "loss": 0.0974,
+      "step": 22340
+    },
+    {
+      "epoch": 2.116076500662753,
+      "grad_norm": 0.3579205572605133,
+      "learning_rate": 1.1535693997348989e-05,
+      "loss": 0.0936,
+      "step": 22350
+    },
+    {
+      "epoch": 2.117023291043363,
+      "grad_norm": 0.3400232493877411,
+      "learning_rate": 1.1531906835826549e-05,
+      "loss": 0.1052,
+      "step": 22360
+    },
+    {
+      "epoch": 2.1179700814239726,
+      "grad_norm": 0.404937744140625,
+      "learning_rate": 1.1528119674304111e-05,
+      "loss": 0.0953,
+      "step": 22370
+    },
+    {
+      "epoch": 2.1189168718045823,
+      "grad_norm": 0.37508904933929443,
+      "learning_rate": 1.1524332512781671e-05,
+      "loss": 0.0947,
+      "step": 22380
+    },
+    {
+      "epoch": 2.119863662185192,
+      "grad_norm": 0.3533603847026825,
+      "learning_rate": 1.1520545351259231e-05,
+      "loss": 0.0974,
+      "step": 22390
+    },
+    {
+      "epoch": 2.120810452565802,
+      "grad_norm": 0.5213316082954407,
+      "learning_rate": 1.1516758189736794e-05,
+      "loss": 0.1033,
+      "step": 22400
+    },
+    {
+      "epoch": 2.1217572429464115,
+      "grad_norm": 0.33669185638427734,
+      "learning_rate": 1.1512971028214354e-05,
+      "loss": 0.1032,
+      "step": 22410
+    },
+    {
+      "epoch": 2.1227040333270213,
+      "grad_norm": 0.40858665108680725,
+      "learning_rate": 1.1509183866691915e-05,
+      "loss": 0.0935,
+      "step": 22420
+    },
+    {
+      "epoch": 2.123650823707631,
+      "grad_norm": 0.4377613067626953,
+      "learning_rate": 1.1505396705169476e-05,
+      "loss": 0.1039,
+      "step": 22430
+    },
+    {
+      "epoch": 2.1245976140882408,
+      "grad_norm": 0.4569412171840668,
+      "learning_rate": 1.1501609543647036e-05,
+      "loss": 0.1047,
+      "step": 22440
+    },
+    {
+      "epoch": 2.1255444044688505,
+      "grad_norm": 0.33584511280059814,
+      "learning_rate": 1.14978223821246e-05,
+      "loss": 0.1045,
+      "step": 22450
+    },
+    {
+      "epoch": 2.1264911948494603,
+      "grad_norm": 0.3816305994987488,
+      "learning_rate": 1.149403522060216e-05,
+      "loss": 0.0912,
+      "step": 22460
+    },
+    {
+      "epoch": 2.12743798523007,
+      "grad_norm": 0.39922207593917847,
+      "learning_rate": 1.149024805907972e-05,
+      "loss": 0.1083,
+      "step": 22470
+    },
+    {
+      "epoch": 2.1283847756106797,
+      "grad_norm": 0.3402602970600128,
+      "learning_rate": 1.1486460897557282e-05,
+      "loss": 0.1036,
+      "step": 22480
+    },
+    {
+      "epoch": 2.1293315659912895,
+      "grad_norm": 0.45204728841781616,
+      "learning_rate": 1.1482673736034843e-05,
+      "loss": 0.1006,
+      "step": 22490
+    },
+    {
+      "epoch": 2.130278356371899,
+      "grad_norm": 0.3794861435890198,
+      "learning_rate": 1.1478886574512405e-05,
+      "loss": 0.1013,
+      "step": 22500
+    },
+    {
+      "epoch": 2.131225146752509,
+      "grad_norm": 0.341086745262146,
+      "learning_rate": 1.1475099412989965e-05,
+      "loss": 0.1117,
+      "step": 22510
+    },
+    {
+      "epoch": 2.1321719371331187,
+      "grad_norm": 0.3587258756160736,
+      "learning_rate": 1.1471312251467525e-05,
+      "loss": 0.099,
+      "step": 22520
+    },
+    {
+      "epoch": 2.1331187275137284,
+      "grad_norm": 0.267326295375824,
+      "learning_rate": 1.1467525089945087e-05,
+      "loss": 0.0919,
+      "step": 22530
+    },
+    {
+      "epoch": 2.134065517894338,
+      "grad_norm": 0.3860250413417816,
+      "learning_rate": 1.1463737928422649e-05,
+      "loss": 0.1005,
+      "step": 22540
+    },
+    {
+      "epoch": 2.135012308274948,
+      "grad_norm": 0.4012054204940796,
+      "learning_rate": 1.145995076690021e-05,
+      "loss": 0.1085,
+      "step": 22550
+    },
+    {
+      "epoch": 2.1359590986555577,
+      "grad_norm": 0.40226098895072937,
+      "learning_rate": 1.145616360537777e-05,
+      "loss": 0.0968,
+      "step": 22560
+    },
+    {
+      "epoch": 2.1369058890361674,
+      "grad_norm": 0.3890402615070343,
+      "learning_rate": 1.145237644385533e-05,
+      "loss": 0.0997,
+      "step": 22570
+    },
+    {
+      "epoch": 2.137852679416777,
+      "grad_norm": 0.3543037176132202,
+      "learning_rate": 1.1448589282332892e-05,
+      "loss": 0.1072,
+      "step": 22580
+    },
+    {
+      "epoch": 2.138799469797387,
+      "grad_norm": 0.3767646253108978,
+      "learning_rate": 1.1444802120810454e-05,
+      "loss": 0.1018,
+      "step": 22590
+    },
+    {
+      "epoch": 2.1397462601779966,
+      "grad_norm": 0.5453084707260132,
+      "learning_rate": 1.1441014959288014e-05,
+      "loss": 0.1013,
+      "step": 22600
+    },
+    {
+      "epoch": 2.1406930505586064,
+      "grad_norm": 0.434987336397171,
+      "learning_rate": 1.1437227797765576e-05,
+      "loss": 0.0935,
+      "step": 22610
+    },
+    {
+      "epoch": 2.141639840939216,
+      "grad_norm": 0.3639536201953888,
+      "learning_rate": 1.1433440636243136e-05,
+      "loss": 0.1021,
+      "step": 22620
+    },
+    {
+      "epoch": 2.142586631319826,
+      "grad_norm": 0.4960155487060547,
+      "learning_rate": 1.14296534747207e-05,
+      "loss": 0.0851,
+      "step": 22630
+    },
+    {
+      "epoch": 2.1435334217004356,
+      "grad_norm": 0.47204598784446716,
+      "learning_rate": 1.142586631319826e-05,
+      "loss": 0.1004,
+      "step": 22640
+    },
+    {
+      "epoch": 2.1444802120810453,
+      "grad_norm": 0.3582194745540619,
+      "learning_rate": 1.142207915167582e-05,
+      "loss": 0.1055,
+      "step": 22650
+    },
+    {
+      "epoch": 2.145427002461655,
+      "grad_norm": 0.3919258713722229,
+      "learning_rate": 1.1418291990153381e-05,
+      "loss": 0.1082,
+      "step": 22660
+    },
+    {
+      "epoch": 2.146373792842265,
+      "grad_norm": 0.43927210569381714,
+      "learning_rate": 1.1414504828630941e-05,
+      "loss": 0.1002,
+      "step": 22670
+    },
+    {
+      "epoch": 2.1473205832228746,
+      "grad_norm": 0.371798038482666,
+      "learning_rate": 1.1410717667108505e-05,
+      "loss": 0.0947,
+      "step": 22680
+    },
+    {
+      "epoch": 2.1482673736034843,
+      "grad_norm": 0.3531559705734253,
+      "learning_rate": 1.1406930505586065e-05,
+      "loss": 0.0998,
+      "step": 22690
+    },
+    {
+      "epoch": 2.149214163984094,
+      "grad_norm": 0.36792996525764465,
+      "learning_rate": 1.1403143344063625e-05,
+      "loss": 0.0975,
+      "step": 22700
+    },
+    {
+      "epoch": 2.150160954364704,
+      "grad_norm": 0.48282304406166077,
+      "learning_rate": 1.1399356182541186e-05,
+      "loss": 0.1017,
+      "step": 22710
+    },
+    {
+      "epoch": 2.1511077447453135,
+      "grad_norm": 0.3662376403808594,
+      "learning_rate": 1.1395569021018746e-05,
+      "loss": 0.0957,
+      "step": 22720
+    },
+    {
+      "epoch": 2.1520545351259233,
+      "grad_norm": 0.46066606044769287,
+      "learning_rate": 1.139178185949631e-05,
+      "loss": 0.108,
+      "step": 22730
+    },
+    {
+      "epoch": 2.153001325506533,
+      "grad_norm": 0.3750786781311035,
+      "learning_rate": 1.138799469797387e-05,
+      "loss": 0.104,
+      "step": 22740
+    },
+    {
+      "epoch": 2.1539481158871427,
+      "grad_norm": 0.40401408076286316,
+      "learning_rate": 1.138420753645143e-05,
+      "loss": 0.0964,
+      "step": 22750
+    },
+    {
+      "epoch": 2.1548949062677525,
+      "grad_norm": 0.4238393008708954,
+      "learning_rate": 1.1380420374928992e-05,
+      "loss": 0.0976,
+      "step": 22760
+    },
+    {
+      "epoch": 2.1558416966483622,
+      "grad_norm": 0.43456730246543884,
+      "learning_rate": 1.1376633213406554e-05,
+      "loss": 0.1018,
+      "step": 22770
+    },
+    {
+      "epoch": 2.156788487028972,
+      "grad_norm": 0.36559218168258667,
+      "learning_rate": 1.1372846051884114e-05,
+      "loss": 0.1044,
+      "step": 22780
+    },
+    {
+      "epoch": 2.1577352774095817,
+      "grad_norm": 0.41886982321739197,
+      "learning_rate": 1.1369058890361675e-05,
+      "loss": 0.1046,
+      "step": 22790
+    },
+    {
+      "epoch": 2.1586820677901915,
+      "grad_norm": 0.47367626428604126,
+      "learning_rate": 1.1365271728839235e-05,
+      "loss": 0.1025,
+      "step": 22800
+    },
+    {
+      "epoch": 2.159628858170801,
+      "grad_norm": 0.38137686252593994,
+      "learning_rate": 1.1361484567316795e-05,
+      "loss": 0.1115,
+      "step": 22810
+    },
+    {
+      "epoch": 2.1605756485514105,
+      "grad_norm": 0.44069352746009827,
+      "learning_rate": 1.1357697405794359e-05,
+      "loss": 0.1125,
+      "step": 22820
+    },
+    {
+      "epoch": 2.1615224389320202,
+      "grad_norm": 0.39172229170799255,
+      "learning_rate": 1.1353910244271919e-05,
+      "loss": 0.099,
+      "step": 22830
+    },
+    {
+      "epoch": 2.16246922931263,
+      "grad_norm": 0.3726031184196472,
+      "learning_rate": 1.135012308274948e-05,
+      "loss": 0.1056,
+      "step": 22840
+    },
+    {
+      "epoch": 2.1634160196932397,
+      "grad_norm": 0.3669355511665344,
+      "learning_rate": 1.134633592122704e-05,
+      "loss": 0.0933,
+      "step": 22850
+    },
+    {
+      "epoch": 2.1643628100738495,
+      "grad_norm": 0.34460747241973877,
+      "learning_rate": 1.13425487597046e-05,
+      "loss": 0.1067,
+      "step": 22860
+    },
+    {
+      "epoch": 2.165309600454459,
+      "grad_norm": 0.35513612627983093,
+      "learning_rate": 1.1338761598182164e-05,
+      "loss": 0.1112,
+      "step": 22870
+    },
+    {
+      "epoch": 2.166256390835069,
+      "grad_norm": 0.33671122789382935,
+      "learning_rate": 1.1334974436659724e-05,
+      "loss": 0.1029,
+      "step": 22880
+    },
+    {
+      "epoch": 2.1672031812156787,
+      "grad_norm": 0.35488536953926086,
+      "learning_rate": 1.1331187275137286e-05,
+      "loss": 0.1062,
+      "step": 22890
+    },
+    {
+      "epoch": 2.1681499715962884,
+      "grad_norm": 0.49665024876594543,
+      "learning_rate": 1.1327400113614846e-05,
+      "loss": 0.1067,
+      "step": 22900
+    },
+    {
+      "epoch": 2.169096761976898,
+      "grad_norm": 0.326875776052475,
+      "learning_rate": 1.132361295209241e-05,
+      "loss": 0.1,
+      "step": 22910
+    },
+    {
+      "epoch": 2.170043552357508,
+      "grad_norm": 0.4280208945274353,
+      "learning_rate": 1.131982579056997e-05,
+      "loss": 0.0971,
+      "step": 22920
+    },
+    {
+      "epoch": 2.1709903427381176,
+      "grad_norm": 0.3841555714607239,
+      "learning_rate": 1.131603862904753e-05,
+      "loss": 0.1064,
+      "step": 22930
+    },
+    {
+      "epoch": 2.1719371331187274,
+      "grad_norm": 0.5070419311523438,
+      "learning_rate": 1.1312251467525091e-05,
+      "loss": 0.101,
+      "step": 22940
+    },
+    {
+      "epoch": 2.172883923499337,
+      "grad_norm": 0.3213464915752411,
+      "learning_rate": 1.1308464306002651e-05,
+      "loss": 0.1038,
+      "step": 22950
+    },
+    {
+      "epoch": 2.173830713879947,
+      "grad_norm": 0.33351173996925354,
+      "learning_rate": 1.1304677144480213e-05,
+      "loss": 0.1018,
+      "step": 22960
+    },
+    {
+      "epoch": 2.1747775042605566,
+      "grad_norm": 0.4228649139404297,
+      "learning_rate": 1.1300889982957775e-05,
+      "loss": 0.0955,
+      "step": 22970
+    },
+    {
+      "epoch": 2.1757242946411663,
+      "grad_norm": 0.3855364918708801,
+      "learning_rate": 1.1297102821435335e-05,
+      "loss": 0.104,
+      "step": 22980
+    },
+    {
+      "epoch": 2.176671085021776,
+      "grad_norm": 0.340480238199234,
+      "learning_rate": 1.1293315659912895e-05,
+      "loss": 0.0951,
+      "step": 22990
+    },
+    {
+      "epoch": 2.177617875402386,
+      "grad_norm": 0.38249287009239197,
+      "learning_rate": 1.1289528498390457e-05,
+      "loss": 0.1,
+      "step": 23000
+    },
+    {
+      "epoch": 2.1785646657829956,
+      "grad_norm": 0.40845736861228943,
+      "learning_rate": 1.1285741336868018e-05,
+      "loss": 0.1086,
+      "step": 23010
+    },
+    {
+      "epoch": 2.1795114561636053,
+      "grad_norm": 0.34934353828430176,
+      "learning_rate": 1.128195417534558e-05,
+      "loss": 0.0902,
+      "step": 23020
+    },
+    {
+      "epoch": 2.180458246544215,
+      "grad_norm": 0.3424789607524872,
+      "learning_rate": 1.127816701382314e-05,
+      "loss": 0.101,
+      "step": 23030
+    },
+    {
+      "epoch": 2.181405036924825,
+      "grad_norm": 0.4287693500518799,
+      "learning_rate": 1.12743798523007e-05,
+      "loss": 0.0964,
+      "step": 23040
+    },
+    {
+      "epoch": 2.1823518273054345,
+      "grad_norm": 0.40323200821876526,
+      "learning_rate": 1.1270592690778264e-05,
+      "loss": 0.105,
+      "step": 23050
+    },
+    {
+      "epoch": 2.1832986176860443,
+      "grad_norm": 0.4015294015407562,
+      "learning_rate": 1.1266805529255824e-05,
+      "loss": 0.1071,
+      "step": 23060
+    },
+    {
+      "epoch": 2.184245408066654,
+      "grad_norm": 0.33209195733070374,
+      "learning_rate": 1.1263018367733385e-05,
+      "loss": 0.1095,
+      "step": 23070
+    },
+    {
+      "epoch": 2.1851921984472638,
+      "grad_norm": 0.32889801263809204,
+      "learning_rate": 1.1259231206210946e-05,
+      "loss": 0.1038,
+      "step": 23080
+    },
+    {
+      "epoch": 2.1861389888278735,
+      "grad_norm": 0.4167686104774475,
+      "learning_rate": 1.1255444044688506e-05,
+      "loss": 0.1093,
+      "step": 23090
+    },
+    {
+      "epoch": 2.1870857792084832,
+      "grad_norm": 0.3739524483680725,
+      "learning_rate": 1.1251656883166069e-05,
+      "loss": 0.0927,
+      "step": 23100
+    },
+    {
+      "epoch": 2.188032569589093,
+      "grad_norm": 0.31450697779655457,
+      "learning_rate": 1.1247869721643629e-05,
+      "loss": 0.0995,
+      "step": 23110
+    },
+    {
+      "epoch": 2.1889793599697027,
+      "grad_norm": 0.4176722466945648,
+      "learning_rate": 1.124408256012119e-05,
+      "loss": 0.1056,
+      "step": 23120
+    },
+    {
+      "epoch": 2.1899261503503125,
+      "grad_norm": 0.4337596595287323,
+      "learning_rate": 1.1240295398598751e-05,
+      "loss": 0.0984,
+      "step": 23130
+    },
+    {
+      "epoch": 2.190872940730922,
+      "grad_norm": 0.42897704243659973,
+      "learning_rate": 1.1236508237076311e-05,
+      "loss": 0.112,
+      "step": 23140
+    },
+    {
+      "epoch": 2.191819731111532,
+      "grad_norm": 0.3697352111339569,
+      "learning_rate": 1.1232721075553874e-05,
+      "loss": 0.1104,
+      "step": 23150
+    },
+    {
+      "epoch": 2.1927665214921417,
+      "grad_norm": 0.38484689593315125,
+      "learning_rate": 1.1228933914031434e-05,
+      "loss": 0.1043,
+      "step": 23160
+    },
+    {
+      "epoch": 2.1937133118727514,
+      "grad_norm": 0.4187999367713928,
+      "learning_rate": 1.1225146752508994e-05,
+      "loss": 0.0929,
+      "step": 23170
+    },
+    {
+      "epoch": 2.194660102253361,
+      "grad_norm": 0.38299185037612915,
+      "learning_rate": 1.1221359590986556e-05,
+      "loss": 0.0971,
+      "step": 23180
+    },
+    {
+      "epoch": 2.195606892633971,
+      "grad_norm": 0.3589703142642975,
+      "learning_rate": 1.1217572429464118e-05,
+      "loss": 0.1065,
+      "step": 23190
+    },
+    {
+      "epoch": 2.1965536830145806,
+      "grad_norm": 0.37426868081092834,
+      "learning_rate": 1.121378526794168e-05,
+      "loss": 0.1031,
+      "step": 23200
+    },
+    {
+      "epoch": 2.1975004733951904,
+      "grad_norm": 0.40484800934791565,
+      "learning_rate": 1.120999810641924e-05,
+      "loss": 0.0995,
+      "step": 23210
+    },
+    {
+      "epoch": 2.1984472637758,
+      "grad_norm": 0.355503112077713,
+      "learning_rate": 1.12062109448968e-05,
+      "loss": 0.0865,
+      "step": 23220
+    },
+    {
+      "epoch": 2.19939405415641,
+      "grad_norm": 0.32087182998657227,
+      "learning_rate": 1.1202423783374362e-05,
+      "loss": 0.0965,
+      "step": 23230
+    },
+    {
+      "epoch": 2.2003408445370196,
+      "grad_norm": 0.4689808189868927,
+      "learning_rate": 1.1198636621851923e-05,
+      "loss": 0.1116,
+      "step": 23240
+    },
+    {
+      "epoch": 2.2012876349176294,
+      "grad_norm": 0.4498715400695801,
+      "learning_rate": 1.1194849460329485e-05,
+      "loss": 0.0951,
+      "step": 23250
+    },
+    {
+      "epoch": 2.202234425298239,
+      "grad_norm": 0.41492167115211487,
+      "learning_rate": 1.1191062298807045e-05,
+      "loss": 0.1072,
+      "step": 23260
+    },
+    {
+      "epoch": 2.203181215678849,
+      "grad_norm": 0.3707394599914551,
+      "learning_rate": 1.1187275137284605e-05,
+      "loss": 0.1012,
+      "step": 23270
+    },
+    {
+      "epoch": 2.2041280060594586,
+      "grad_norm": 0.29027289152145386,
+      "learning_rate": 1.1183487975762167e-05,
+      "loss": 0.1106,
+      "step": 23280
+    },
+    {
+      "epoch": 2.2050747964400683,
+      "grad_norm": 0.4856210947036743,
+      "learning_rate": 1.1179700814239729e-05,
+      "loss": 0.1114,
+      "step": 23290
+    },
+    {
+      "epoch": 2.206021586820678,
+      "grad_norm": 0.3140472173690796,
+      "learning_rate": 1.117591365271729e-05,
+      "loss": 0.1064,
+      "step": 23300
+    },
+    {
+      "epoch": 2.206968377201288,
+      "grad_norm": 0.33930498361587524,
+      "learning_rate": 1.117212649119485e-05,
+      "loss": 0.0924,
+      "step": 23310
+    },
+    {
+      "epoch": 2.2079151675818975,
+      "grad_norm": 0.37622299790382385,
+      "learning_rate": 1.116833932967241e-05,
+      "loss": 0.1065,
+      "step": 23320
+    },
+    {
+      "epoch": 2.2088619579625073,
+      "grad_norm": 0.3196406066417694,
+      "learning_rate": 1.1164552168149974e-05,
+      "loss": 0.1021,
+      "step": 23330
+    },
+    {
+      "epoch": 2.209808748343117,
+      "grad_norm": 0.5100225210189819,
+      "learning_rate": 1.1160765006627534e-05,
+      "loss": 0.0962,
+      "step": 23340
+    },
+    {
+      "epoch": 2.2107555387237268,
+      "grad_norm": 0.3671920895576477,
+      "learning_rate": 1.1156977845105094e-05,
+      "loss": 0.1023,
+      "step": 23350
+    },
+    {
+      "epoch": 2.2117023291043365,
+      "grad_norm": 0.45640867948532104,
+      "learning_rate": 1.1153190683582656e-05,
+      "loss": 0.1033,
+      "step": 23360
+    },
+    {
+      "epoch": 2.2126491194849462,
+      "grad_norm": 0.36683550477027893,
+      "learning_rate": 1.1149403522060216e-05,
+      "loss": 0.1012,
+      "step": 23370
+    },
+    {
+      "epoch": 2.213595909865556,
+      "grad_norm": 0.30357277393341064,
+      "learning_rate": 1.114561636053778e-05,
+      "loss": 0.1083,
+      "step": 23380
+    },
+    {
+      "epoch": 2.2145427002461657,
+      "grad_norm": 0.4162294268608093,
+      "learning_rate": 1.114182919901534e-05,
+      "loss": 0.0986,
+      "step": 23390
+    },
+    {
+      "epoch": 2.215489490626775,
+      "grad_norm": 0.45273682475090027,
+      "learning_rate": 1.11380420374929e-05,
+      "loss": 0.1074,
+      "step": 23400
+    },
+    {
+      "epoch": 2.2164362810073848,
+      "grad_norm": 0.36101433634757996,
+      "learning_rate": 1.1134254875970461e-05,
+      "loss": 0.1018,
+      "step": 23410
+    },
+    {
+      "epoch": 2.2173830713879945,
+      "grad_norm": 0.3455662727355957,
+      "learning_rate": 1.1130467714448021e-05,
+      "loss": 0.1075,
+      "step": 23420
+    },
+    {
+      "epoch": 2.2183298617686042,
+      "grad_norm": 0.34743842482566833,
+      "learning_rate": 1.1126680552925585e-05,
+      "loss": 0.0967,
+      "step": 23430
+    },
+    {
+      "epoch": 2.219276652149214,
+      "grad_norm": 0.560983419418335,
+      "learning_rate": 1.1122893391403145e-05,
+      "loss": 0.1049,
+      "step": 23440
+    },
+    {
+      "epoch": 2.2202234425298237,
+      "grad_norm": 0.4330439865589142,
+      "learning_rate": 1.1119106229880705e-05,
+      "loss": 0.1065,
+      "step": 23450
+    },
+    {
+      "epoch": 2.2211702329104335,
+      "grad_norm": 0.32083675265312195,
+      "learning_rate": 1.1115319068358266e-05,
+      "loss": 0.0957,
+      "step": 23460
+    },
+    {
+      "epoch": 2.222117023291043,
+      "grad_norm": 0.4112713634967804,
+      "learning_rate": 1.1111531906835828e-05,
+      "loss": 0.1034,
+      "step": 23470
+    },
+    {
+      "epoch": 2.223063813671653,
+      "grad_norm": 0.3654922842979431,
+      "learning_rate": 1.110774474531339e-05,
+      "loss": 0.095,
+      "step": 23480
+    },
+    {
+      "epoch": 2.2240106040522627,
+      "grad_norm": 0.3552573621273041,
+      "learning_rate": 1.110395758379095e-05,
+      "loss": 0.1026,
+      "step": 23490
+    },
+    {
+      "epoch": 2.2249573944328724,
+      "grad_norm": 0.34025436639785767,
+      "learning_rate": 1.110017042226851e-05,
+      "loss": 0.0899,
+      "step": 23500
+    },
+    {
+      "epoch": 2.225904184813482,
+      "grad_norm": 0.4973716139793396,
+      "learning_rate": 1.109638326074607e-05,
+      "loss": 0.1019,
+      "step": 23510
+    },
+    {
+      "epoch": 2.226850975194092,
+      "grad_norm": 0.4082499146461487,
+      "learning_rate": 1.1092596099223633e-05,
+      "loss": 0.1037,
+      "step": 23520
+    },
+    {
+      "epoch": 2.2277977655747017,
+      "grad_norm": 0.4043087959289551,
+      "learning_rate": 1.1088808937701194e-05,
+      "loss": 0.1039,
+      "step": 23530
+    },
+    {
+      "epoch": 2.2287445559553114,
+      "grad_norm": 0.38824453949928284,
+      "learning_rate": 1.1085021776178755e-05,
+      "loss": 0.1082,
+      "step": 23540
+    },
+    {
+      "epoch": 2.229691346335921,
+      "grad_norm": 0.39685913920402527,
+      "learning_rate": 1.1081234614656315e-05,
+      "loss": 0.1018,
+      "step": 23550
+    },
+    {
+      "epoch": 2.230638136716531,
+      "grad_norm": 0.3705897927284241,
+      "learning_rate": 1.1077447453133875e-05,
+      "loss": 0.0942,
+      "step": 23560
+    },
+    {
+      "epoch": 2.2315849270971406,
+      "grad_norm": 0.34453701972961426,
+      "learning_rate": 1.1073660291611439e-05,
+      "loss": 0.0989,
+      "step": 23570
+    },
+    {
+      "epoch": 2.2325317174777504,
+      "grad_norm": 0.3718234896659851,
+      "learning_rate": 1.1069873130088999e-05,
+      "loss": 0.1031,
+      "step": 23580
+    },
+    {
+      "epoch": 2.23347850785836,
+      "grad_norm": 0.41039136052131653,
+      "learning_rate": 1.106608596856656e-05,
+      "loss": 0.1032,
+      "step": 23590
+    },
+    {
+      "epoch": 2.23442529823897,
+      "grad_norm": 0.3343423008918762,
+      "learning_rate": 1.106229880704412e-05,
+      "loss": 0.0923,
+      "step": 23600
+    },
+    {
+      "epoch": 2.2353720886195796,
+      "grad_norm": 0.47895216941833496,
+      "learning_rate": 1.1058511645521684e-05,
+      "loss": 0.0983,
+      "step": 23610
+    },
+    {
+      "epoch": 2.2363188790001893,
+      "grad_norm": 0.4448593258857727,
+      "learning_rate": 1.1054724483999244e-05,
+      "loss": 0.1053,
+      "step": 23620
+    },
+    {
+      "epoch": 2.237265669380799,
+      "grad_norm": 0.3424699008464813,
+      "learning_rate": 1.1050937322476804e-05,
+      "loss": 0.1031,
+      "step": 23630
+    },
+    {
+      "epoch": 2.238212459761409,
+      "grad_norm": 0.39542269706726074,
+      "learning_rate": 1.1047150160954366e-05,
+      "loss": 0.1019,
+      "step": 23640
+    },
+    {
+      "epoch": 2.2391592501420186,
+      "grad_norm": 0.46471700072288513,
+      "learning_rate": 1.1043362999431926e-05,
+      "loss": 0.1024,
+      "step": 23650
+    },
+    {
+      "epoch": 2.2401060405226283,
+      "grad_norm": 0.3723284900188446,
+      "learning_rate": 1.103957583790949e-05,
+      "loss": 0.1063,
+      "step": 23660
+    },
+    {
+      "epoch": 2.241052830903238,
+      "grad_norm": 0.43616223335266113,
+      "learning_rate": 1.103578867638705e-05,
+      "loss": 0.0981,
+      "step": 23670
+    },
+    {
+      "epoch": 2.2419996212838478,
+      "grad_norm": 0.48921120166778564,
+      "learning_rate": 1.103200151486461e-05,
+      "loss": 0.0993,
+      "step": 23680
+    },
+    {
+      "epoch": 2.2429464116644575,
+      "grad_norm": 0.38076233863830566,
+      "learning_rate": 1.102821435334217e-05,
+      "loss": 0.0946,
+      "step": 23690
+    },
+    {
+      "epoch": 2.2438932020450673,
+      "grad_norm": 0.4395969808101654,
+      "learning_rate": 1.1024427191819731e-05,
+      "loss": 0.1043,
+      "step": 23700
+    },
+    {
+      "epoch": 2.244839992425677,
+      "grad_norm": 0.4211001396179199,
+      "learning_rate": 1.1020640030297293e-05,
+      "loss": 0.1045,
+      "step": 23710
+    },
+    {
+      "epoch": 2.2457867828062867,
+      "grad_norm": 0.47736597061157227,
+      "learning_rate": 1.1016852868774855e-05,
+      "loss": 0.0977,
+      "step": 23720
+    },
+    {
+      "epoch": 2.2467335731868965,
+      "grad_norm": 0.37812915444374084,
+      "learning_rate": 1.1013065707252415e-05,
+      "loss": 0.1043,
+      "step": 23730
+    },
+    {
+      "epoch": 2.2476803635675062,
+      "grad_norm": 0.3703843057155609,
+      "learning_rate": 1.1009278545729975e-05,
+      "loss": 0.0907,
+      "step": 23740
+    },
+    {
+      "epoch": 2.248627153948116,
+      "grad_norm": 0.5733705759048462,
+      "learning_rate": 1.1005491384207538e-05,
+      "loss": 0.1081,
+      "step": 23750
+    },
+    {
+      "epoch": 2.2495739443287257,
+      "grad_norm": 0.6033251881599426,
+      "learning_rate": 1.1001704222685098e-05,
+      "loss": 0.0984,
+      "step": 23760
+    },
+    {
+      "epoch": 2.2505207347093354,
+      "grad_norm": 0.39114996790885925,
+      "learning_rate": 1.099791706116266e-05,
+      "loss": 0.0993,
+      "step": 23770
+    },
+    {
+      "epoch": 2.251467525089945,
+      "grad_norm": 0.19909757375717163,
+      "learning_rate": 1.099412989964022e-05,
+      "loss": 0.0931,
+      "step": 23780
+    },
+    {
+      "epoch": 2.252414315470555,
+      "grad_norm": 0.36311134696006775,
+      "learning_rate": 1.099034273811778e-05,
+      "loss": 0.0998,
+      "step": 23790
+    },
+    {
+      "epoch": 2.2533611058511647,
+      "grad_norm": 0.392431378364563,
+      "learning_rate": 1.0986555576595344e-05,
+      "loss": 0.111,
+      "step": 23800
+    },
+    {
+      "epoch": 2.2543078962317744,
+      "grad_norm": 0.46047669649124146,
+      "learning_rate": 1.0982768415072904e-05,
+      "loss": 0.106,
+      "step": 23810
+    },
+    {
+      "epoch": 2.255254686612384,
+      "grad_norm": 0.42195338010787964,
+      "learning_rate": 1.0978981253550465e-05,
+      "loss": 0.1035,
+      "step": 23820
+    },
+    {
+      "epoch": 2.256201476992994,
+      "grad_norm": 0.3718126118183136,
+      "learning_rate": 1.0975194092028025e-05,
+      "loss": 0.1002,
+      "step": 23830
+    },
+    {
+      "epoch": 2.2571482673736036,
+      "grad_norm": 0.42515698075294495,
+      "learning_rate": 1.0971406930505586e-05,
+      "loss": 0.0956,
+      "step": 23840
+    },
+    {
+      "epoch": 2.2580950577542134,
+      "grad_norm": 0.3577660620212555,
+      "learning_rate": 1.0967619768983149e-05,
+      "loss": 0.0911,
+      "step": 23850
+    },
+    {
+      "epoch": 2.259041848134823,
+      "grad_norm": 0.3810538649559021,
+      "learning_rate": 1.0963832607460709e-05,
+      "loss": 0.0925,
+      "step": 23860
+    },
+    {
+      "epoch": 2.259988638515433,
+      "grad_norm": 0.3717814087867737,
+      "learning_rate": 1.0960045445938269e-05,
+      "loss": 0.1065,
+      "step": 23870
+    },
+    {
+      "epoch": 2.2609354288960426,
+      "grad_norm": 0.3299460709095001,
+      "learning_rate": 1.095625828441583e-05,
+      "loss": 0.101,
+      "step": 23880
+    },
+    {
+      "epoch": 2.2618822192766523,
+      "grad_norm": 0.44920241832733154,
+      "learning_rate": 1.0952471122893393e-05,
+      "loss": 0.0995,
+      "step": 23890
+    },
+    {
+      "epoch": 2.2628290096572616,
+      "grad_norm": 0.475917786359787,
+      "learning_rate": 1.0948683961370954e-05,
+      "loss": 0.1048,
+      "step": 23900
+    },
+    {
+      "epoch": 2.2637758000378714,
+      "grad_norm": 0.34341323375701904,
+      "learning_rate": 1.0944896799848514e-05,
+      "loss": 0.0944,
+      "step": 23910
+    },
+    {
+      "epoch": 2.264722590418481,
+      "grad_norm": 0.4694386124610901,
+      "learning_rate": 1.0941109638326074e-05,
+      "loss": 0.1032,
+      "step": 23920
+    },
+    {
+      "epoch": 2.265669380799091,
+      "grad_norm": 0.3619329631328583,
+      "learning_rate": 1.0937322476803636e-05,
+      "loss": 0.1046,
+      "step": 23930
+    },
+    {
+      "epoch": 2.2666161711797006,
+      "grad_norm": 0.4530966877937317,
+      "learning_rate": 1.0933535315281198e-05,
+      "loss": 0.1078,
+      "step": 23940
+    },
+    {
+      "epoch": 2.2675629615603103,
+      "grad_norm": 0.44397875666618347,
+      "learning_rate": 1.092974815375876e-05,
+      "loss": 0.1055,
+      "step": 23950
+    },
+    {
+      "epoch": 2.26850975194092,
+      "grad_norm": 0.3822941482067108,
+      "learning_rate": 1.092596099223632e-05,
+      "loss": 0.0984,
+      "step": 23960
+    },
+    {
+      "epoch": 2.26945654232153,
+      "grad_norm": 0.41480833292007446,
+      "learning_rate": 1.092217383071388e-05,
+      "loss": 0.0986,
+      "step": 23970
+    },
+    {
+      "epoch": 2.2704033327021396,
+      "grad_norm": 0.44778960943222046,
+      "learning_rate": 1.0918386669191441e-05,
+      "loss": 0.0938,
+      "step": 23980
+    },
+    {
+      "epoch": 2.2713501230827493,
+      "grad_norm": 0.36493951082229614,
+      "learning_rate": 1.0914599507669003e-05,
+      "loss": 0.114,
+      "step": 23990
+    },
+    {
+      "epoch": 2.272296913463359,
+      "grad_norm": 0.3928702175617218,
+      "learning_rate": 1.0910812346146565e-05,
+      "loss": 0.099,
+      "step": 24000
+    },
+    {
+      "epoch": 2.273243703843969,
+      "grad_norm": 0.3912307024002075,
+      "learning_rate": 1.0907025184624125e-05,
+      "loss": 0.1041,
+      "step": 24010
+    },
+    {
+      "epoch": 2.2741904942245785,
+      "grad_norm": 0.4148852229118347,
+      "learning_rate": 1.0903238023101685e-05,
+      "loss": 0.1105,
+      "step": 24020
+    },
+    {
+      "epoch": 2.2751372846051883,
+      "grad_norm": 0.35756590962409973,
+      "learning_rate": 1.0899450861579249e-05,
+      "loss": 0.1032,
+      "step": 24030
+    },
+    {
+      "epoch": 2.276084074985798,
+      "grad_norm": 0.3831791281700134,
+      "learning_rate": 1.0895663700056809e-05,
+      "loss": 0.1013,
+      "step": 24040
+    },
+    {
+      "epoch": 2.2770308653664078,
+      "grad_norm": 0.40133121609687805,
+      "learning_rate": 1.0891876538534369e-05,
+      "loss": 0.1076,
+      "step": 24050
+    },
+    {
+      "epoch": 2.2779776557470175,
+      "grad_norm": 0.38974645733833313,
+      "learning_rate": 1.088808937701193e-05,
+      "loss": 0.102,
+      "step": 24060
+    },
+    {
+      "epoch": 2.2789244461276272,
+      "grad_norm": 0.3199298679828644,
+      "learning_rate": 1.088430221548949e-05,
+      "loss": 0.1021,
+      "step": 24070
+    },
+    {
+      "epoch": 2.279871236508237,
+      "grad_norm": 0.3961081802845001,
+      "learning_rate": 1.0880515053967054e-05,
+      "loss": 0.1049,
+      "step": 24080
+    },
+    {
+      "epoch": 2.2808180268888467,
+      "grad_norm": 0.41309094429016113,
+      "learning_rate": 1.0876727892444614e-05,
+      "loss": 0.102,
+      "step": 24090
+    },
+    {
+      "epoch": 2.2817648172694565,
+      "grad_norm": 0.465978741645813,
+      "learning_rate": 1.0872940730922174e-05,
+      "loss": 0.1078,
+      "step": 24100
+    },
+    {
+      "epoch": 2.282711607650066,
+      "grad_norm": 0.4430433511734009,
+      "learning_rate": 1.0869153569399736e-05,
+      "loss": 0.1085,
+      "step": 24110
+    },
+    {
+      "epoch": 2.283658398030676,
+      "grad_norm": 0.30565348267555237,
+      "learning_rate": 1.0865366407877296e-05,
+      "loss": 0.0987,
+      "step": 24120
+    },
+    {
+      "epoch": 2.2846051884112857,
+      "grad_norm": 0.39248156547546387,
+      "learning_rate": 1.086157924635486e-05,
+      "loss": 0.1028,
+      "step": 24130
+    },
+    {
+      "epoch": 2.2855519787918954,
+      "grad_norm": 0.2839440405368805,
+      "learning_rate": 1.085779208483242e-05,
+      "loss": 0.1027,
+      "step": 24140
+    },
+    {
+      "epoch": 2.286498769172505,
+      "grad_norm": 0.4470391571521759,
+      "learning_rate": 1.085400492330998e-05,
+      "loss": 0.1004,
+      "step": 24150
+    },
+    {
+      "epoch": 2.287445559553115,
+      "grad_norm": 0.388384610414505,
+      "learning_rate": 1.0850217761787541e-05,
+      "loss": 0.1032,
+      "step": 24160
+    },
+    {
+      "epoch": 2.2883923499337246,
+      "grad_norm": 0.4492109417915344,
+      "learning_rate": 1.0846430600265103e-05,
+      "loss": 0.0996,
+      "step": 24170
+    },
+    {
+      "epoch": 2.2893391403143344,
+      "grad_norm": 0.5113523602485657,
+      "learning_rate": 1.0842643438742664e-05,
+      "loss": 0.1005,
+      "step": 24180
+    },
+    {
+      "epoch": 2.290285930694944,
+      "grad_norm": 0.385354220867157,
+      "learning_rate": 1.0838856277220225e-05,
+      "loss": 0.0923,
+      "step": 24190
+    },
+    {
+      "epoch": 2.291232721075554,
+      "grad_norm": 0.32123634219169617,
+      "learning_rate": 1.0835069115697785e-05,
+      "loss": 0.0998,
+      "step": 24200
+    },
+    {
+      "epoch": 2.2921795114561636,
+      "grad_norm": 0.343923956155777,
+      "learning_rate": 1.0831281954175346e-05,
+      "loss": 0.1069,
+      "step": 24210
+    },
+    {
+      "epoch": 2.2931263018367734,
+      "grad_norm": 0.4596140682697296,
+      "learning_rate": 1.0827494792652908e-05,
+      "loss": 0.1073,
+      "step": 24220
+    },
+    {
+      "epoch": 2.294073092217383,
+      "grad_norm": 0.41971439123153687,
+      "learning_rate": 1.0823707631130468e-05,
+      "loss": 0.1001,
+      "step": 24230
+    },
+    {
+      "epoch": 2.295019882597993,
+      "grad_norm": 0.42049020528793335,
+      "learning_rate": 1.081992046960803e-05,
+      "loss": 0.1014,
+      "step": 24240
+    },
+    {
+      "epoch": 2.2959666729786026,
+      "grad_norm": 0.3357348144054413,
+      "learning_rate": 1.081613330808559e-05,
+      "loss": 0.1056,
+      "step": 24250
+    },
+    {
+      "epoch": 2.2969134633592123,
+      "grad_norm": 0.4161593019962311,
+      "learning_rate": 1.081234614656315e-05,
+      "loss": 0.0925,
+      "step": 24260
+    },
+    {
+      "epoch": 2.297860253739822,
+      "grad_norm": 0.4063127934932709,
+      "learning_rate": 1.0808558985040713e-05,
+      "loss": 0.1007,
+      "step": 24270
+    },
+    {
+      "epoch": 2.298807044120432,
+      "grad_norm": 0.4100823402404785,
+      "learning_rate": 1.0804771823518273e-05,
+      "loss": 0.0959,
+      "step": 24280
+    },
+    {
+      "epoch": 2.2997538345010415,
+      "grad_norm": 0.37574973702430725,
+      "learning_rate": 1.0800984661995835e-05,
+      "loss": 0.0992,
+      "step": 24290
+    },
+    {
+      "epoch": 2.3007006248816513,
+      "grad_norm": 0.3953949213027954,
+      "learning_rate": 1.0797197500473395e-05,
+      "loss": 0.0973,
+      "step": 24300
+    },
+    {
+      "epoch": 2.301647415262261,
+      "grad_norm": 0.41357365250587463,
+      "learning_rate": 1.0793410338950959e-05,
+      "loss": 0.0964,
+      "step": 24310
+    },
+    {
+      "epoch": 2.3025942056428708,
+      "grad_norm": 0.5206488966941833,
+      "learning_rate": 1.0789623177428519e-05,
+      "loss": 0.0944,
+      "step": 24320
+    },
+    {
+      "epoch": 2.3035409960234805,
+      "grad_norm": 0.4037434458732605,
+      "learning_rate": 1.0785836015906079e-05,
+      "loss": 0.1056,
+      "step": 24330
+    },
+    {
+      "epoch": 2.3044877864040902,
+      "grad_norm": 0.3638472557067871,
+      "learning_rate": 1.078204885438364e-05,
+      "loss": 0.105,
+      "step": 24340
+    },
+    {
+      "epoch": 2.3054345767847,
+      "grad_norm": 0.31433483958244324,
+      "learning_rate": 1.07782616928612e-05,
+      "loss": 0.0918,
+      "step": 24350
+    },
+    {
+      "epoch": 2.3063813671653097,
+      "grad_norm": 0.4167768359184265,
+      "learning_rate": 1.0774474531338764e-05,
+      "loss": 0.1058,
+      "step": 24360
+    },
+    {
+      "epoch": 2.3073281575459195,
+      "grad_norm": 0.3920457661151886,
+      "learning_rate": 1.0770687369816324e-05,
+      "loss": 0.1039,
+      "step": 24370
+    },
+    {
+      "epoch": 2.308274947926529,
+      "grad_norm": 0.40075308084487915,
+      "learning_rate": 1.0766900208293884e-05,
+      "loss": 0.1016,
+      "step": 24380
+    },
+    {
+      "epoch": 2.309221738307139,
+      "grad_norm": 0.33713826537132263,
+      "learning_rate": 1.0763113046771446e-05,
+      "loss": 0.1019,
+      "step": 24390
+    },
+    {
+      "epoch": 2.3101685286877487,
+      "grad_norm": 0.3890605568885803,
+      "learning_rate": 1.0759325885249006e-05,
+      "loss": 0.1052,
+      "step": 24400
+    },
+    {
+      "epoch": 2.3111153190683584,
+      "grad_norm": 0.39935365319252014,
+      "learning_rate": 1.0755538723726568e-05,
+      "loss": 0.1035,
+      "step": 24410
+    },
+    {
+      "epoch": 2.312062109448968,
+      "grad_norm": 0.45820632576942444,
+      "learning_rate": 1.075175156220413e-05,
+      "loss": 0.1042,
+      "step": 24420
+    },
+    {
+      "epoch": 2.313008899829578,
+      "grad_norm": 0.39744850993156433,
+      "learning_rate": 1.074796440068169e-05,
+      "loss": 0.107,
+      "step": 24430
+    },
+    {
+      "epoch": 2.3139556902101877,
+      "grad_norm": 0.3054029643535614,
+      "learning_rate": 1.074417723915925e-05,
+      "loss": 0.0961,
+      "step": 24440
+    },
+    {
+      "epoch": 2.3149024805907974,
+      "grad_norm": 0.39057713747024536,
+      "learning_rate": 1.0740390077636813e-05,
+      "loss": 0.0982,
+      "step": 24450
+    },
+    {
+      "epoch": 2.315849270971407,
+      "grad_norm": 0.3561437726020813,
+      "learning_rate": 1.0736602916114373e-05,
+      "loss": 0.1011,
+      "step": 24460
+    },
+    {
+      "epoch": 2.316796061352017,
+      "grad_norm": 0.5833640694618225,
+      "learning_rate": 1.0732815754591935e-05,
+      "loss": 0.0979,
+      "step": 24470
+    },
+    {
+      "epoch": 2.3177428517326266,
+      "grad_norm": 0.37024566531181335,
+      "learning_rate": 1.0729028593069495e-05,
+      "loss": 0.097,
+      "step": 24480
+    },
+    {
+      "epoch": 2.3186896421132364,
+      "grad_norm": 0.4453885853290558,
+      "learning_rate": 1.0725241431547055e-05,
+      "loss": 0.0997,
+      "step": 24490
+    },
+    {
+      "epoch": 2.319636432493846,
+      "grad_norm": 0.3976321220397949,
+      "learning_rate": 1.0721454270024618e-05,
+      "loss": 0.1111,
+      "step": 24500
+    },
+    {
+      "epoch": 2.320583222874456,
+      "grad_norm": 0.4668462872505188,
+      "learning_rate": 1.0717667108502178e-05,
+      "loss": 0.1026,
+      "step": 24510
+    },
+    {
+      "epoch": 2.321530013255065,
+      "grad_norm": 0.39205145835876465,
+      "learning_rate": 1.071387994697974e-05,
+      "loss": 0.0995,
+      "step": 24520
+    },
+    {
+      "epoch": 2.322476803635675,
+      "grad_norm": 0.5295231342315674,
+      "learning_rate": 1.07100927854573e-05,
+      "loss": 0.0996,
+      "step": 24530
+    },
+    {
+      "epoch": 2.3234235940162846,
+      "grad_norm": 0.41702479124069214,
+      "learning_rate": 1.070630562393486e-05,
+      "loss": 0.0925,
+      "step": 24540
+    },
+    {
+      "epoch": 2.3243703843968944,
+      "grad_norm": 0.41526830196380615,
+      "learning_rate": 1.0702518462412424e-05,
+      "loss": 0.1025,
+      "step": 24550
+    },
+    {
+      "epoch": 2.325317174777504,
+      "grad_norm": 0.5049123167991638,
+      "learning_rate": 1.0698731300889984e-05,
+      "loss": 0.1003,
+      "step": 24560
+    },
+    {
+      "epoch": 2.326263965158114,
+      "grad_norm": 0.35912126302719116,
+      "learning_rate": 1.0694944139367545e-05,
+      "loss": 0.0987,
+      "step": 24570
+    },
+    {
+      "epoch": 2.3272107555387236,
+      "grad_norm": 0.4618116319179535,
+      "learning_rate": 1.0691156977845105e-05,
+      "loss": 0.0898,
+      "step": 24580
+    },
+    {
+      "epoch": 2.3281575459193333,
+      "grad_norm": 0.6204038262367249,
+      "learning_rate": 1.0687369816322667e-05,
+      "loss": 0.1065,
+      "step": 24590
+    },
+    {
+      "epoch": 2.329104336299943,
+      "grad_norm": 0.43588438630104065,
+      "learning_rate": 1.0683582654800229e-05,
+      "loss": 0.0994,
+      "step": 24600
+    },
+    {
+      "epoch": 2.330051126680553,
+      "grad_norm": 0.4884413480758667,
+      "learning_rate": 1.0679795493277789e-05,
+      "loss": 0.1034,
+      "step": 24610
+    },
+    {
+      "epoch": 2.3309979170611625,
+      "grad_norm": 0.4790765643119812,
+      "learning_rate": 1.0676008331755349e-05,
+      "loss": 0.103,
+      "step": 24620
+    },
+    {
+      "epoch": 2.3319447074417723,
+      "grad_norm": 0.4357077181339264,
+      "learning_rate": 1.067222117023291e-05,
+      "loss": 0.1067,
+      "step": 24630
+    },
+    {
+      "epoch": 2.332891497822382,
+      "grad_norm": 0.4674258530139923,
+      "learning_rate": 1.0668434008710473e-05,
+      "loss": 0.1077,
+      "step": 24640
+    },
+    {
+      "epoch": 2.3338382882029918,
+      "grad_norm": 0.3695948123931885,
+      "learning_rate": 1.0664646847188034e-05,
+      "loss": 0.1057,
+      "step": 24650
+    },
+    {
+      "epoch": 2.3347850785836015,
+      "grad_norm": 0.4131108522415161,
+      "learning_rate": 1.0660859685665594e-05,
+      "loss": 0.0913,
+      "step": 24660
+    },
+    {
+      "epoch": 2.3357318689642113,
+      "grad_norm": 0.42493703961372375,
+      "learning_rate": 1.0657072524143154e-05,
+      "loss": 0.1159,
+      "step": 24670
+    },
+    {
+      "epoch": 2.336678659344821,
+      "grad_norm": 0.33579689264297485,
+      "learning_rate": 1.0653285362620716e-05,
+      "loss": 0.1058,
+      "step": 24680
+    },
+    {
+      "epoch": 2.3376254497254307,
+      "grad_norm": 0.32977592945098877,
+      "learning_rate": 1.0649498201098278e-05,
+      "loss": 0.1109,
+      "step": 24690
+    },
+    {
+      "epoch": 2.3385722401060405,
+      "grad_norm": 0.5081207752227783,
+      "learning_rate": 1.064571103957584e-05,
+      "loss": 0.1073,
+      "step": 24700
+    },
+    {
+      "epoch": 2.33951903048665,
+      "grad_norm": 0.4069708585739136,
+      "learning_rate": 1.06419238780534e-05,
+      "loss": 0.0989,
+      "step": 24710
+    },
+    {
+      "epoch": 2.34046582086726,
+      "grad_norm": 0.522319495677948,
+      "learning_rate": 1.063813671653096e-05,
+      "loss": 0.1065,
+      "step": 24720
+    },
+    {
+      "epoch": 2.3414126112478697,
+      "grad_norm": 0.4592416286468506,
+      "learning_rate": 1.0634349555008523e-05,
+      "loss": 0.099,
+      "step": 24730
+    },
+    {
+      "epoch": 2.3423594016284794,
+      "grad_norm": 0.4711533486843109,
+      "learning_rate": 1.0630562393486083e-05,
+      "loss": 0.1073,
+      "step": 24740
+    },
+    {
+      "epoch": 2.343306192009089,
+      "grad_norm": 0.47300341725349426,
+      "learning_rate": 1.0626775231963645e-05,
+      "loss": 0.0981,
+      "step": 24750
+    },
+    {
+      "epoch": 2.344252982389699,
+      "grad_norm": 0.43047043681144714,
+      "learning_rate": 1.0622988070441205e-05,
+      "loss": 0.1011,
+      "step": 24760
+    },
+    {
+      "epoch": 2.3451997727703087,
+      "grad_norm": 0.44232434034347534,
+      "learning_rate": 1.0619200908918765e-05,
+      "loss": 0.1123,
+      "step": 24770
+    },
+    {
+      "epoch": 2.3461465631509184,
+      "grad_norm": 0.37191641330718994,
+      "learning_rate": 1.0615413747396328e-05,
+      "loss": 0.1089,
+      "step": 24780
+    },
+    {
+      "epoch": 2.347093353531528,
+      "grad_norm": 0.3949145972728729,
+      "learning_rate": 1.0611626585873888e-05,
+      "loss": 0.0954,
+      "step": 24790
+    },
+    {
+      "epoch": 2.348040143912138,
+      "grad_norm": 0.4205836057662964,
+      "learning_rate": 1.0607839424351449e-05,
+      "loss": 0.1051,
+      "step": 24800
+    },
+    {
+      "epoch": 2.3489869342927476,
+      "grad_norm": 0.37778088450431824,
+      "learning_rate": 1.060405226282901e-05,
+      "loss": 0.1046,
+      "step": 24810
+    },
+    {
+      "epoch": 2.3499337246733574,
+      "grad_norm": 0.49821335077285767,
+      "learning_rate": 1.060026510130657e-05,
+      "loss": 0.0951,
+      "step": 24820
+    },
+    {
+      "epoch": 2.350880515053967,
+      "grad_norm": 0.4278693199157715,
+      "learning_rate": 1.0596477939784134e-05,
+      "loss": 0.1021,
+      "step": 24830
+    },
+    {
+      "epoch": 2.351827305434577,
+      "grad_norm": 0.33827054500579834,
+      "learning_rate": 1.0592690778261694e-05,
+      "loss": 0.0946,
+      "step": 24840
+    },
+    {
+      "epoch": 2.3527740958151866,
+      "grad_norm": 0.38886088132858276,
+      "learning_rate": 1.0588903616739254e-05,
+      "loss": 0.1012,
+      "step": 24850
+    },
+    {
+      "epoch": 2.3537208861957963,
+      "grad_norm": 0.40593603253364563,
+      "learning_rate": 1.0585116455216816e-05,
+      "loss": 0.0967,
+      "step": 24860
+    },
+    {
+      "epoch": 2.354667676576406,
+      "grad_norm": 0.45751744508743286,
+      "learning_rate": 1.0581329293694377e-05,
+      "loss": 0.1053,
+      "step": 24870
+    },
+    {
+      "epoch": 2.355614466957016,
+      "grad_norm": 0.5506783127784729,
+      "learning_rate": 1.0577542132171939e-05,
+      "loss": 0.1005,
+      "step": 24880
+    },
+    {
+      "epoch": 2.3565612573376256,
+      "grad_norm": 0.39473041892051697,
+      "learning_rate": 1.05737549706495e-05,
+      "loss": 0.1158,
+      "step": 24890
+    },
+    {
+      "epoch": 2.3575080477182353,
+      "grad_norm": 0.4578403830528259,
+      "learning_rate": 1.056996780912706e-05,
+      "loss": 0.1007,
+      "step": 24900
+    },
+    {
+      "epoch": 2.358454838098845,
+      "grad_norm": 0.4987157881259918,
+      "learning_rate": 1.0566180647604621e-05,
+      "loss": 0.106,
+      "step": 24910
+    },
+    {
+      "epoch": 2.359401628479455,
+      "grad_norm": 0.4185522794723511,
+      "learning_rate": 1.0562393486082183e-05,
+      "loss": 0.1079,
+      "step": 24920
+    },
+    {
+      "epoch": 2.3603484188600645,
+      "grad_norm": 0.5091202259063721,
+      "learning_rate": 1.0558606324559744e-05,
+      "loss": 0.0975,
+      "step": 24930
+    },
+    {
+      "epoch": 2.3612952092406743,
+      "grad_norm": 0.3753291666507721,
+      "learning_rate": 1.0554819163037304e-05,
+      "loss": 0.0955,
+      "step": 24940
+    },
+    {
+      "epoch": 2.362241999621284,
+      "grad_norm": 0.5717144012451172,
+      "learning_rate": 1.0551032001514865e-05,
+      "loss": 0.1101,
+      "step": 24950
+    },
+    {
+      "epoch": 2.3631887900018937,
+      "grad_norm": 0.3344045877456665,
+      "learning_rate": 1.0547244839992425e-05,
+      "loss": 0.0962,
+      "step": 24960
+    },
+    {
+      "epoch": 2.3641355803825035,
+      "grad_norm": 0.41045841574668884,
+      "learning_rate": 1.0543457678469988e-05,
+      "loss": 0.0969,
+      "step": 24970
+    },
+    {
+      "epoch": 2.3650823707631132,
+      "grad_norm": 0.37063470482826233,
+      "learning_rate": 1.0539670516947548e-05,
+      "loss": 0.1095,
+      "step": 24980
+    },
+    {
+      "epoch": 2.366029161143723,
+      "grad_norm": 0.43232864141464233,
+      "learning_rate": 1.053588335542511e-05,
+      "loss": 0.1046,
+      "step": 24990
+    },
+    {
+      "epoch": 2.3669759515243323,
+      "grad_norm": 0.3335595726966858,
+      "learning_rate": 1.053209619390267e-05,
+      "loss": 0.0973,
+      "step": 25000
+    },
+    {
+      "epoch": 2.367922741904942,
+      "grad_norm": 0.3094400465488434,
+      "learning_rate": 1.0528309032380233e-05,
+      "loss": 0.103,
+      "step": 25010
+    },
+    {
+      "epoch": 2.3688695322855517,
+      "grad_norm": 0.4552101790904999,
+      "learning_rate": 1.0524521870857793e-05,
+      "loss": 0.1036,
+      "step": 25020
+    },
+    {
+      "epoch": 2.3698163226661615,
+      "grad_norm": 0.44783899188041687,
+      "learning_rate": 1.0520734709335353e-05,
+      "loss": 0.0968,
+      "step": 25030
+    },
+    {
+      "epoch": 2.3707631130467712,
+      "grad_norm": 0.39976587891578674,
+      "learning_rate": 1.0516947547812915e-05,
+      "loss": 0.1006,
+      "step": 25040
+    },
+    {
+      "epoch": 2.371709903427381,
+      "grad_norm": 0.4066604673862457,
+      "learning_rate": 1.0513160386290475e-05,
+      "loss": 0.1012,
+      "step": 25050
+    },
+    {
+      "epoch": 2.3726566938079907,
+      "grad_norm": 0.3684154152870178,
+      "learning_rate": 1.0509373224768039e-05,
+      "loss": 0.1036,
+      "step": 25060
+    },
+    {
+      "epoch": 2.3736034841886005,
+      "grad_norm": 0.44917765259742737,
+      "learning_rate": 1.0505586063245599e-05,
+      "loss": 0.0948,
+      "step": 25070
+    },
+    {
+      "epoch": 2.37455027456921,
+      "grad_norm": 0.41735780239105225,
+      "learning_rate": 1.0501798901723159e-05,
+      "loss": 0.0976,
+      "step": 25080
+    },
+    {
+      "epoch": 2.37549706494982,
+      "grad_norm": 0.4906679093837738,
+      "learning_rate": 1.049801174020072e-05,
+      "loss": 0.1085,
+      "step": 25090
+    },
+    {
+      "epoch": 2.3764438553304297,
+      "grad_norm": 0.36485058069229126,
+      "learning_rate": 1.0494224578678282e-05,
+      "loss": 0.1058,
+      "step": 25100
+    },
+    {
+      "epoch": 2.3773906457110394,
+      "grad_norm": 0.41198959946632385,
+      "learning_rate": 1.0490437417155844e-05,
+      "loss": 0.1033,
+      "step": 25110
+    },
+    {
+      "epoch": 2.378337436091649,
+      "grad_norm": 0.3697530925273895,
+      "learning_rate": 1.0486650255633404e-05,
+      "loss": 0.1002,
+      "step": 25120
+    },
+    {
+      "epoch": 2.379284226472259,
+      "grad_norm": 0.3377944529056549,
+      "learning_rate": 1.0482863094110964e-05,
+      "loss": 0.1087,
+      "step": 25130
+    },
+    {
+      "epoch": 2.3802310168528686,
+      "grad_norm": 0.3125544786453247,
+      "learning_rate": 1.0479075932588524e-05,
+      "loss": 0.0987,
+      "step": 25140
+    },
+    {
+      "epoch": 2.3811778072334784,
+      "grad_norm": 0.5143665075302124,
+      "learning_rate": 1.0475288771066088e-05,
+      "loss": 0.1159,
+      "step": 25150
+    },
+    {
+      "epoch": 2.382124597614088,
+      "grad_norm": 0.6163843274116516,
+      "learning_rate": 1.0471501609543648e-05,
+      "loss": 0.1035,
+      "step": 25160
+    },
+    {
+      "epoch": 2.383071387994698,
+      "grad_norm": 0.37617814540863037,
+      "learning_rate": 1.046771444802121e-05,
+      "loss": 0.1069,
+      "step": 25170
+    },
+    {
+      "epoch": 2.3840181783753076,
+      "grad_norm": 0.33066341280937195,
+      "learning_rate": 1.046392728649877e-05,
+      "loss": 0.0948,
+      "step": 25180
+    },
+    {
+      "epoch": 2.3849649687559173,
+      "grad_norm": 0.4607396125793457,
+      "learning_rate": 1.046014012497633e-05,
+      "loss": 0.1102,
+      "step": 25190
+    },
+    {
+      "epoch": 2.385911759136527,
+      "grad_norm": 0.513787031173706,
+      "learning_rate": 1.0456352963453893e-05,
+      "loss": 0.1104,
+      "step": 25200
+    },
+    {
+      "epoch": 2.386858549517137,
+      "grad_norm": 0.4283960461616516,
+      "learning_rate": 1.0452565801931453e-05,
+      "loss": 0.1061,
+      "step": 25210
+    },
+    {
+      "epoch": 2.3878053398977466,
+      "grad_norm": 0.4610239267349243,
+      "learning_rate": 1.0448778640409015e-05,
+      "loss": 0.1001,
+      "step": 25220
+    },
+    {
+      "epoch": 2.3887521302783563,
+      "grad_norm": 0.44825395941734314,
+      "learning_rate": 1.0444991478886575e-05,
+      "loss": 0.103,
+      "step": 25230
+    },
+    {
+      "epoch": 2.389698920658966,
+      "grad_norm": 0.5586211681365967,
+      "learning_rate": 1.0441204317364138e-05,
+      "loss": 0.1103,
+      "step": 25240
+    },
+    {
+      "epoch": 2.390645711039576,
+      "grad_norm": 0.49608463048934937,
+      "learning_rate": 1.0437417155841698e-05,
+      "loss": 0.108,
+      "step": 25250
+    },
+    {
+      "epoch": 2.3915925014201855,
+      "grad_norm": 0.44712910056114197,
+      "learning_rate": 1.0433629994319258e-05,
+      "loss": 0.1046,
+      "step": 25260
+    },
+    {
+      "epoch": 2.3925392918007953,
+      "grad_norm": 0.39927998185157776,
+      "learning_rate": 1.042984283279682e-05,
+      "loss": 0.0994,
+      "step": 25270
+    },
+    {
+      "epoch": 2.393486082181405,
+      "grad_norm": 0.3669414222240448,
+      "learning_rate": 1.042605567127438e-05,
+      "loss": 0.1054,
+      "step": 25280
+    },
+    {
+      "epoch": 2.3944328725620148,
+      "grad_norm": 0.369989812374115,
+      "learning_rate": 1.0422268509751943e-05,
+      "loss": 0.0967,
+      "step": 25290
+    },
+    {
+      "epoch": 2.3953796629426245,
+      "grad_norm": 0.41249990463256836,
+      "learning_rate": 1.0418481348229504e-05,
+      "loss": 0.1009,
+      "step": 25300
+    },
+    {
+      "epoch": 2.3963264533232342,
+      "grad_norm": 0.4730875492095947,
+      "learning_rate": 1.0414694186707064e-05,
+      "loss": 0.1078,
+      "step": 25310
+    },
+    {
+      "epoch": 2.397273243703844,
+      "grad_norm": 0.3919539451599121,
+      "learning_rate": 1.0410907025184624e-05,
+      "loss": 0.101,
+      "step": 25320
+    },
+    {
+      "epoch": 2.3982200340844537,
+      "grad_norm": 0.3141326308250427,
+      "learning_rate": 1.0407119863662185e-05,
+      "loss": 0.0976,
+      "step": 25330
+    },
+    {
+      "epoch": 2.3991668244650635,
+      "grad_norm": 0.5199115872383118,
+      "learning_rate": 1.0403332702139747e-05,
+      "loss": 0.0982,
+      "step": 25340
+    },
+    {
+      "epoch": 2.400113614845673,
+      "grad_norm": 0.4684145152568817,
+      "learning_rate": 1.0399545540617309e-05,
+      "loss": 0.1113,
+      "step": 25350
+    },
+    {
+      "epoch": 2.401060405226283,
+      "grad_norm": 0.3334536552429199,
+      "learning_rate": 1.0395758379094869e-05,
+      "loss": 0.1024,
+      "step": 25360
+    },
+    {
+      "epoch": 2.4020071956068927,
+      "grad_norm": 0.4707101583480835,
+      "learning_rate": 1.0391971217572429e-05,
+      "loss": 0.0954,
+      "step": 25370
+    },
+    {
+      "epoch": 2.4029539859875024,
+      "grad_norm": 0.40972524881362915,
+      "learning_rate": 1.0388184056049992e-05,
+      "loss": 0.0996,
+      "step": 25380
+    },
+    {
+      "epoch": 2.403900776368112,
+      "grad_norm": 0.41831696033477783,
+      "learning_rate": 1.0384396894527552e-05,
+      "loss": 0.0975,
+      "step": 25390
+    },
+    {
+      "epoch": 2.404847566748722,
+      "grad_norm": 0.3924408555030823,
+      "learning_rate": 1.0380609733005114e-05,
+      "loss": 0.0992,
+      "step": 25400
+    },
+    {
+      "epoch": 2.4057943571293317,
+      "grad_norm": 0.30749720335006714,
+      "learning_rate": 1.0376822571482674e-05,
+      "loss": 0.0926,
+      "step": 25410
+    },
+    {
+      "epoch": 2.4067411475099414,
+      "grad_norm": 0.5540740489959717,
+      "learning_rate": 1.0373035409960234e-05,
+      "loss": 0.1105,
+      "step": 25420
+    },
+    {
+      "epoch": 2.407687937890551,
+      "grad_norm": 0.3289966285228729,
+      "learning_rate": 1.0369248248437798e-05,
+      "loss": 0.0948,
+      "step": 25430
+    },
+    {
+      "epoch": 2.408634728271161,
+      "grad_norm": 0.4328863322734833,
+      "learning_rate": 1.0365461086915358e-05,
+      "loss": 0.1089,
+      "step": 25440
+    },
+    {
+      "epoch": 2.4095815186517706,
+      "grad_norm": 0.40920546650886536,
+      "learning_rate": 1.036167392539292e-05,
+      "loss": 0.1039,
+      "step": 25450
+    },
+    {
+      "epoch": 2.4105283090323804,
+      "grad_norm": 0.3753580152988434,
+      "learning_rate": 1.035788676387048e-05,
+      "loss": 0.1008,
+      "step": 25460
+    },
+    {
+      "epoch": 2.41147509941299,
+      "grad_norm": 0.39004644751548767,
+      "learning_rate": 1.035409960234804e-05,
+      "loss": 0.0979,
+      "step": 25470
+    },
+    {
+      "epoch": 2.4124218897936,
+      "grad_norm": 0.4152458608150482,
+      "learning_rate": 1.0350312440825603e-05,
+      "loss": 0.0988,
+      "step": 25480
+    },
+    {
+      "epoch": 2.4133686801742096,
+      "grad_norm": 0.3045837879180908,
+      "learning_rate": 1.0346525279303163e-05,
+      "loss": 0.0986,
+      "step": 25490
+    },
+    {
+      "epoch": 2.4143154705548193,
+      "grad_norm": 0.4492361545562744,
+      "learning_rate": 1.0342738117780723e-05,
+      "loss": 0.1012,
+      "step": 25500
+    },
+    {
+      "epoch": 2.415262260935429,
+      "grad_norm": 0.35907822847366333,
+      "learning_rate": 1.0338950956258285e-05,
+      "loss": 0.1057,
+      "step": 25510
+    },
+    {
+      "epoch": 2.416209051316039,
+      "grad_norm": 0.39662420749664307,
+      "learning_rate": 1.0335163794735847e-05,
+      "loss": 0.097,
+      "step": 25520
+    },
+    {
+      "epoch": 2.4171558416966485,
+      "grad_norm": 0.3447544276714325,
+      "learning_rate": 1.0331376633213408e-05,
+      "loss": 0.0984,
+      "step": 25530
+    },
+    {
+      "epoch": 2.4181026320772583,
+      "grad_norm": 0.3469262719154358,
+      "learning_rate": 1.0327589471690968e-05,
+      "loss": 0.0924,
+      "step": 25540
+    },
+    {
+      "epoch": 2.419049422457868,
+      "grad_norm": 0.4851413369178772,
+      "learning_rate": 1.0323802310168528e-05,
+      "loss": 0.0913,
+      "step": 25550
+    },
+    {
+      "epoch": 2.4199962128384778,
+      "grad_norm": 0.4492618143558502,
+      "learning_rate": 1.032001514864609e-05,
+      "loss": 0.102,
+      "step": 25560
+    },
+    {
+      "epoch": 2.4209430032190875,
+      "grad_norm": 0.35623621940612793,
+      "learning_rate": 1.0316227987123652e-05,
+      "loss": 0.102,
+      "step": 25570
+    },
+    {
+      "epoch": 2.4218897935996972,
+      "grad_norm": 0.3780003488063812,
+      "learning_rate": 1.0312440825601214e-05,
+      "loss": 0.1039,
+      "step": 25580
+    },
+    {
+      "epoch": 2.422836583980307,
+      "grad_norm": 0.32068032026290894,
+      "learning_rate": 1.0308653664078774e-05,
+      "loss": 0.1024,
+      "step": 25590
+    },
+    {
+      "epoch": 2.4237833743609167,
+      "grad_norm": 0.49494946002960205,
+      "learning_rate": 1.0304866502556334e-05,
+      "loss": 0.1009,
+      "step": 25600
+    },
+    {
+      "epoch": 2.4247301647415265,
+      "grad_norm": 0.38599666953086853,
+      "learning_rate": 1.0301079341033896e-05,
+      "loss": 0.0999,
+      "step": 25610
+    },
+    {
+      "epoch": 2.4256769551221358,
+      "grad_norm": 0.4353082776069641,
+      "learning_rate": 1.0297292179511457e-05,
+      "loss": 0.0963,
+      "step": 25620
+    },
+    {
+      "epoch": 2.4266237455027455,
+      "grad_norm": 0.46552932262420654,
+      "learning_rate": 1.0293505017989019e-05,
+      "loss": 0.1018,
+      "step": 25630
+    },
+    {
+      "epoch": 2.4275705358833553,
+      "grad_norm": 0.45082664489746094,
+      "learning_rate": 1.0289717856466579e-05,
+      "loss": 0.0984,
+      "step": 25640
+    },
+    {
+      "epoch": 2.428517326263965,
+      "grad_norm": 0.4540878236293793,
+      "learning_rate": 1.028593069494414e-05,
+      "loss": 0.0968,
+      "step": 25650
+    },
+    {
+      "epoch": 2.4294641166445747,
+      "grad_norm": 0.43615978956222534,
+      "learning_rate": 1.0282143533421703e-05,
+      "loss": 0.1084,
+      "step": 25660
+    },
+    {
+      "epoch": 2.4304109070251845,
+      "grad_norm": 0.3706541359424591,
+      "learning_rate": 1.0278356371899263e-05,
+      "loss": 0.1014,
+      "step": 25670
+    },
+    {
+      "epoch": 2.431357697405794,
+      "grad_norm": 0.37163105607032776,
+      "learning_rate": 1.0274569210376823e-05,
+      "loss": 0.1007,
+      "step": 25680
+    },
+    {
+      "epoch": 2.432304487786404,
+      "grad_norm": 0.40251484513282776,
+      "learning_rate": 1.0270782048854384e-05,
+      "loss": 0.0975,
+      "step": 25690
+    },
+    {
+      "epoch": 2.4332512781670137,
+      "grad_norm": 0.49345412850379944,
+      "learning_rate": 1.0266994887331944e-05,
+      "loss": 0.1049,
+      "step": 25700
+    },
+    {
+      "epoch": 2.4341980685476234,
+      "grad_norm": 0.4760383665561676,
+      "learning_rate": 1.0263207725809508e-05,
+      "loss": 0.1036,
+      "step": 25710
+    },
+    {
+      "epoch": 2.435144858928233,
+      "grad_norm": 0.3557586669921875,
+      "learning_rate": 1.0259420564287068e-05,
+      "loss": 0.1005,
+      "step": 25720
+    },
+    {
+      "epoch": 2.436091649308843,
+      "grad_norm": 0.40292495489120483,
+      "learning_rate": 1.0255633402764628e-05,
+      "loss": 0.1133,
+      "step": 25730
+    },
+    {
+      "epoch": 2.4370384396894527,
+      "grad_norm": 0.47368377447128296,
+      "learning_rate": 1.025184624124219e-05,
+      "loss": 0.1011,
+      "step": 25740
+    },
+    {
+      "epoch": 2.4379852300700624,
+      "grad_norm": 0.36609113216400146,
+      "learning_rate": 1.024805907971975e-05,
+      "loss": 0.1008,
+      "step": 25750
+    },
+    {
+      "epoch": 2.438932020450672,
+      "grad_norm": 0.46925249695777893,
+      "learning_rate": 1.0244271918197313e-05,
+      "loss": 0.1067,
+      "step": 25760
+    },
+    {
+      "epoch": 2.439878810831282,
+      "grad_norm": 0.4214388132095337,
+      "learning_rate": 1.0240484756674873e-05,
+      "loss": 0.098,
+      "step": 25770
+    },
+    {
+      "epoch": 2.4408256012118916,
+      "grad_norm": 0.40375715494155884,
+      "learning_rate": 1.0236697595152433e-05,
+      "loss": 0.1016,
+      "step": 25780
+    },
+    {
+      "epoch": 2.4417723915925014,
+      "grad_norm": 0.4981019198894501,
+      "learning_rate": 1.0232910433629995e-05,
+      "loss": 0.0995,
+      "step": 25790
+    },
+    {
+      "epoch": 2.442719181973111,
+      "grad_norm": 0.363440603017807,
+      "learning_rate": 1.0229123272107557e-05,
+      "loss": 0.1016,
+      "step": 25800
+    },
+    {
+      "epoch": 2.443665972353721,
+      "grad_norm": 0.3500191569328308,
+      "learning_rate": 1.0225336110585119e-05,
+      "loss": 0.0983,
+      "step": 25810
+    },
+    {
+      "epoch": 2.4446127627343306,
+      "grad_norm": 0.4750962555408478,
+      "learning_rate": 1.0221548949062679e-05,
+      "loss": 0.1119,
+      "step": 25820
+    },
+    {
+      "epoch": 2.4455595531149403,
+      "grad_norm": 0.4848047196865082,
+      "learning_rate": 1.0217761787540239e-05,
+      "loss": 0.1047,
+      "step": 25830
+    },
+    {
+      "epoch": 2.44650634349555,
+      "grad_norm": 0.370795875787735,
+      "learning_rate": 1.02139746260178e-05,
+      "loss": 0.102,
+      "step": 25840
+    },
+    {
+      "epoch": 2.44745313387616,
+      "grad_norm": 0.32790902256965637,
+      "learning_rate": 1.0210187464495362e-05,
+      "loss": 0.1046,
+      "step": 25850
+    },
+    {
+      "epoch": 2.4483999242567696,
+      "grad_norm": 0.45114025473594666,
+      "learning_rate": 1.0206400302972922e-05,
+      "loss": 0.1025,
+      "step": 25860
+    },
+    {
+      "epoch": 2.4493467146373793,
+      "grad_norm": 0.45215851068496704,
+      "learning_rate": 1.0202613141450484e-05,
+      "loss": 0.0962,
+      "step": 25870
+    },
+    {
+      "epoch": 2.450293505017989,
+      "grad_norm": 0.5089224576950073,
+      "learning_rate": 1.0198825979928044e-05,
+      "loss": 0.0939,
+      "step": 25880
+    },
+    {
+      "epoch": 2.451240295398599,
+      "grad_norm": 0.4181041121482849,
+      "learning_rate": 1.0195038818405604e-05,
+      "loss": 0.0912,
+      "step": 25890
+    },
+    {
+      "epoch": 2.4521870857792085,
+      "grad_norm": 0.39679673314094543,
+      "learning_rate": 1.0191251656883167e-05,
+      "loss": 0.1028,
+      "step": 25900
+    },
+    {
+      "epoch": 2.4531338761598183,
+      "grad_norm": 0.434073269367218,
+      "learning_rate": 1.0187464495360728e-05,
+      "loss": 0.1004,
+      "step": 25910
+    },
+    {
+      "epoch": 2.454080666540428,
+      "grad_norm": 0.38664939999580383,
+      "learning_rate": 1.018367733383829e-05,
+      "loss": 0.1054,
+      "step": 25920
+    },
+    {
+      "epoch": 2.4550274569210377,
+      "grad_norm": 0.3743431568145752,
+      "learning_rate": 1.017989017231585e-05,
+      "loss": 0.1084,
+      "step": 25930
+    },
+    {
+      "epoch": 2.4559742473016475,
+      "grad_norm": 0.33121728897094727,
+      "learning_rate": 1.0176103010793413e-05,
+      "loss": 0.1053,
+      "step": 25940
+    },
+    {
+      "epoch": 2.4569210376822572,
+      "grad_norm": 0.40912964940071106,
+      "learning_rate": 1.0172315849270973e-05,
+      "loss": 0.1018,
+      "step": 25950
+    },
+    {
+      "epoch": 2.457867828062867,
+      "grad_norm": 0.36607125401496887,
+      "learning_rate": 1.0168528687748533e-05,
+      "loss": 0.0844,
+      "step": 25960
+    },
+    {
+      "epoch": 2.4588146184434767,
+      "grad_norm": 0.3174314796924591,
+      "learning_rate": 1.0164741526226095e-05,
+      "loss": 0.0992,
+      "step": 25970
+    },
+    {
+      "epoch": 2.4597614088240864,
+      "grad_norm": 0.47616884112358093,
+      "learning_rate": 1.0160954364703655e-05,
+      "loss": 0.0967,
+      "step": 25980
+    },
+    {
+      "epoch": 2.460708199204696,
+      "grad_norm": 0.44378870725631714,
+      "learning_rate": 1.0157167203181218e-05,
+      "loss": 0.1127,
+      "step": 25990
+    },
+    {
+      "epoch": 2.461654989585306,
+      "grad_norm": 0.37386101484298706,
+      "learning_rate": 1.0153380041658778e-05,
+      "loss": 0.103,
+      "step": 26000
+    },
+    {
+      "epoch": 2.4626017799659157,
+      "grad_norm": 0.3897585868835449,
+      "learning_rate": 1.0149592880136338e-05,
+      "loss": 0.1054,
+      "step": 26010
+    },
+    {
+      "epoch": 2.4635485703465254,
+      "grad_norm": 0.39490073919296265,
+      "learning_rate": 1.01458057186139e-05,
+      "loss": 0.0868,
+      "step": 26020
+    },
+    {
+      "epoch": 2.464495360727135,
+      "grad_norm": 0.3911207616329193,
+      "learning_rate": 1.014201855709146e-05,
+      "loss": 0.0994,
+      "step": 26030
+    },
+    {
+      "epoch": 2.465442151107745,
+      "grad_norm": 0.4419346749782562,
+      "learning_rate": 1.0138231395569022e-05,
+      "loss": 0.1015,
+      "step": 26040
+    },
+    {
+      "epoch": 2.4663889414883546,
+      "grad_norm": 0.38570520281791687,
+      "learning_rate": 1.0134444234046583e-05,
+      "loss": 0.1021,
+      "step": 26050
+    },
+    {
+      "epoch": 2.4673357318689644,
+      "grad_norm": 0.477276474237442,
+      "learning_rate": 1.0130657072524144e-05,
+      "loss": 0.115,
+      "step": 26060
+    },
+    {
+      "epoch": 2.468282522249574,
+      "grad_norm": 0.5218283534049988,
+      "learning_rate": 1.0126869911001704e-05,
+      "loss": 0.1097,
+      "step": 26070
+    },
+    {
+      "epoch": 2.469229312630184,
+      "grad_norm": 0.4555433690547943,
+      "learning_rate": 1.0123082749479267e-05,
+      "loss": 0.0965,
+      "step": 26080
+    },
+    {
+      "epoch": 2.4701761030107936,
+      "grad_norm": 0.44638416171073914,
+      "learning_rate": 1.0119295587956827e-05,
+      "loss": 0.1144,
+      "step": 26090
+    },
+    {
+      "epoch": 2.471122893391403,
+      "grad_norm": 0.4732908010482788,
+      "learning_rate": 1.0115508426434389e-05,
+      "loss": 0.1105,
+      "step": 26100
+    },
+    {
+      "epoch": 2.4720696837720126,
+      "grad_norm": 0.359528511762619,
+      "learning_rate": 1.0111721264911949e-05,
+      "loss": 0.1026,
+      "step": 26110
+    },
+    {
+      "epoch": 2.4730164741526224,
+      "grad_norm": 0.37714725732803345,
+      "learning_rate": 1.0107934103389509e-05,
+      "loss": 0.1029,
+      "step": 26120
+    },
+    {
+      "epoch": 2.473963264533232,
+      "grad_norm": 0.40320953726768494,
+      "learning_rate": 1.0104146941867072e-05,
+      "loss": 0.1089,
+      "step": 26130
+    },
+    {
+      "epoch": 2.474910054913842,
+      "grad_norm": 0.44699448347091675,
+      "learning_rate": 1.0100359780344632e-05,
+      "loss": 0.0999,
+      "step": 26140
+    },
+    {
+      "epoch": 2.4758568452944516,
+      "grad_norm": 0.4459933638572693,
+      "learning_rate": 1.0096572618822194e-05,
+      "loss": 0.1066,
+      "step": 26150
+    },
+    {
+      "epoch": 2.4768036356750613,
+      "grad_norm": 0.4070841670036316,
+      "learning_rate": 1.0092785457299754e-05,
+      "loss": 0.0888,
+      "step": 26160
+    },
+    {
+      "epoch": 2.477750426055671,
+      "grad_norm": 0.3742724359035492,
+      "learning_rate": 1.0088998295777314e-05,
+      "loss": 0.1042,
+      "step": 26170
+    },
+    {
+      "epoch": 2.478697216436281,
+      "grad_norm": 0.392292857170105,
+      "learning_rate": 1.0085211134254878e-05,
+      "loss": 0.1116,
+      "step": 26180
+    },
+    {
+      "epoch": 2.4796440068168906,
+      "grad_norm": 0.471701979637146,
+      "learning_rate": 1.0081423972732438e-05,
+      "loss": 0.1021,
+      "step": 26190
+    },
+    {
+      "epoch": 2.4805907971975003,
+      "grad_norm": 0.4960010349750519,
+      "learning_rate": 1.007763681121e-05,
+      "loss": 0.1079,
+      "step": 26200
+    },
+    {
+      "epoch": 2.48153758757811,
+      "grad_norm": 0.39826032519340515,
+      "learning_rate": 1.007384964968756e-05,
+      "loss": 0.1085,
+      "step": 26210
+    },
+    {
+      "epoch": 2.48248437795872,
+      "grad_norm": 0.6029541492462158,
+      "learning_rate": 1.0070062488165121e-05,
+      "loss": 0.1003,
+      "step": 26220
+    },
+    {
+      "epoch": 2.4834311683393295,
+      "grad_norm": 0.3961453437805176,
+      "learning_rate": 1.0066275326642683e-05,
+      "loss": 0.104,
+      "step": 26230
+    },
+    {
+      "epoch": 2.4843779587199393,
+      "grad_norm": 0.3852611482143402,
+      "learning_rate": 1.0062488165120243e-05,
+      "loss": 0.0978,
+      "step": 26240
+    },
+    {
+      "epoch": 2.485324749100549,
+      "grad_norm": 0.38596782088279724,
+      "learning_rate": 1.0058701003597803e-05,
+      "loss": 0.1016,
+      "step": 26250
+    },
+    {
+      "epoch": 2.4862715394811588,
+      "grad_norm": 0.4030351936817169,
+      "learning_rate": 1.0054913842075365e-05,
+      "loss": 0.0972,
+      "step": 26260
+    },
+    {
+      "epoch": 2.4872183298617685,
+      "grad_norm": 0.3803381621837616,
+      "learning_rate": 1.0051126680552927e-05,
+      "loss": 0.1077,
+      "step": 26270
+    },
+    {
+      "epoch": 2.4881651202423782,
+      "grad_norm": 0.36874496936798096,
+      "learning_rate": 1.0047339519030488e-05,
+      "loss": 0.0997,
+      "step": 26280
+    },
+    {
+      "epoch": 2.489111910622988,
+      "grad_norm": 0.3608757257461548,
+      "learning_rate": 1.0043552357508048e-05,
+      "loss": 0.1122,
+      "step": 26290
+    },
+    {
+      "epoch": 2.4900587010035977,
+      "grad_norm": 0.4197673499584198,
+      "learning_rate": 1.0039765195985608e-05,
+      "loss": 0.106,
+      "step": 26300
+    },
+    {
+      "epoch": 2.4910054913842075,
+      "grad_norm": 0.2702406644821167,
+      "learning_rate": 1.003597803446317e-05,
+      "loss": 0.0992,
+      "step": 26310
+    },
+    {
+      "epoch": 2.491952281764817,
+      "grad_norm": 0.41988605260849,
+      "learning_rate": 1.0032190872940732e-05,
+      "loss": 0.1146,
+      "step": 26320
+    },
+    {
+      "epoch": 2.492899072145427,
+      "grad_norm": 0.36301714181900024,
+      "learning_rate": 1.0028403711418294e-05,
+      "loss": 0.1033,
+      "step": 26330
+    },
+    {
+      "epoch": 2.4938458625260367,
+      "grad_norm": 0.3015279471874237,
+      "learning_rate": 1.0024616549895854e-05,
+      "loss": 0.1009,
+      "step": 26340
+    },
+    {
+      "epoch": 2.4947926529066464,
+      "grad_norm": 0.35299044847488403,
+      "learning_rate": 1.0020829388373414e-05,
+      "loss": 0.1005,
+      "step": 26350
+    },
+    {
+      "epoch": 2.495739443287256,
+      "grad_norm": 0.4299270212650299,
+      "learning_rate": 1.0017042226850977e-05,
+      "loss": 0.101,
+      "step": 26360
+    },
+    {
+      "epoch": 2.496686233667866,
+      "grad_norm": 0.37819668650627136,
+      "learning_rate": 1.0013255065328537e-05,
+      "loss": 0.1016,
+      "step": 26370
+    },
+    {
+      "epoch": 2.4976330240484756,
+      "grad_norm": 0.3589572012424469,
+      "learning_rate": 1.0009467903806099e-05,
+      "loss": 0.1035,
+      "step": 26380
+    },
+    {
+      "epoch": 2.4985798144290854,
+      "grad_norm": 0.46611204743385315,
+      "learning_rate": 1.0005680742283659e-05,
+      "loss": 0.0983,
+      "step": 26390
+    },
+    {
+      "epoch": 2.499526604809695,
+      "grad_norm": 0.3120664358139038,
+      "learning_rate": 1.0001893580761219e-05,
+      "loss": 0.0976,
+      "step": 26400
+    },
+    {
+      "epoch": 2.500473395190305,
+      "grad_norm": 0.5002902746200562,
+      "learning_rate": 9.99810641923878e-06,
+      "loss": 0.1132,
+      "step": 26410
+    },
+    {
+      "epoch": 2.5014201855709146,
+      "grad_norm": 0.4584709405899048,
+      "learning_rate": 9.994319257716343e-06,
+      "loss": 0.1046,
+      "step": 26420
+    },
+    {
+      "epoch": 2.5023669759515244,
+      "grad_norm": 0.3668809235095978,
+      "learning_rate": 9.990532096193903e-06,
+      "loss": 0.1001,
+      "step": 26430
+    },
+    {
+      "epoch": 2.503313766332134,
+      "grad_norm": 0.3679928183555603,
+      "learning_rate": 9.986744934671464e-06,
+      "loss": 0.1007,
+      "step": 26440
+    },
+    {
+      "epoch": 2.504260556712744,
+      "grad_norm": 0.5092624425888062,
+      "learning_rate": 9.982957773149026e-06,
+      "loss": 0.0932,
+      "step": 26450
+    },
+    {
+      "epoch": 2.5052073470933536,
+      "grad_norm": 0.441669762134552,
+      "learning_rate": 9.979170611626586e-06,
+      "loss": 0.1069,
+      "step": 26460
+    },
+    {
+      "epoch": 2.5061541374739633,
+      "grad_norm": 0.3531561493873596,
+      "learning_rate": 9.975383450104148e-06,
+      "loss": 0.099,
+      "step": 26470
+    },
+    {
+      "epoch": 2.507100927854573,
+      "grad_norm": 0.37312039732933044,
+      "learning_rate": 9.971596288581708e-06,
+      "loss": 0.0952,
+      "step": 26480
+    },
+    {
+      "epoch": 2.508047718235183,
+      "grad_norm": 0.34380894899368286,
+      "learning_rate": 9.96780912705927e-06,
+      "loss": 0.0982,
+      "step": 26490
+    },
+    {
+      "epoch": 2.5089945086157925,
+      "grad_norm": 0.4413946568965912,
+      "learning_rate": 9.964021965536831e-06,
+      "loss": 0.0978,
+      "step": 26500
+    },
+    {
+      "epoch": 2.5099412989964023,
+      "grad_norm": 0.31311383843421936,
+      "learning_rate": 9.960234804014393e-06,
+      "loss": 0.0958,
+      "step": 26510
+    },
+    {
+      "epoch": 2.510888089377012,
+      "grad_norm": 0.6745802164077759,
+      "learning_rate": 9.956447642491953e-06,
+      "loss": 0.1069,
+      "step": 26520
+    },
+    {
+      "epoch": 2.5118348797576218,
+      "grad_norm": 0.4960954189300537,
+      "learning_rate": 9.952660480969513e-06,
+      "loss": 0.1153,
+      "step": 26530
+    },
+    {
+      "epoch": 2.5127816701382315,
+      "grad_norm": 0.41089656949043274,
+      "learning_rate": 9.948873319447075e-06,
+      "loss": 0.104,
+      "step": 26540
+    },
+    {
+      "epoch": 2.5137284605188412,
+      "grad_norm": 0.4213186502456665,
+      "learning_rate": 9.945086157924637e-06,
+      "loss": 0.097,
+      "step": 26550
+    },
+    {
+      "epoch": 2.514675250899451,
+      "grad_norm": 0.5175119638442993,
+      "learning_rate": 9.941298996402199e-06,
+      "loss": 0.101,
+      "step": 26560
+    },
+    {
+      "epoch": 2.5156220412800607,
+      "grad_norm": 0.4612017273902893,
+      "learning_rate": 9.937511834879759e-06,
+      "loss": 0.1023,
+      "step": 26570
+    },
+    {
+      "epoch": 2.5165688316606705,
+      "grad_norm": 0.3659135699272156,
+      "learning_rate": 9.93372467335732e-06,
+      "loss": 0.0975,
+      "step": 26580
+    },
+    {
+      "epoch": 2.51751562204128,
+      "grad_norm": 0.38637617230415344,
+      "learning_rate": 9.92993751183488e-06,
+      "loss": 0.0971,
+      "step": 26590
+    },
+    {
+      "epoch": 2.51846241242189,
+      "grad_norm": 0.493749737739563,
+      "learning_rate": 9.92615035031244e-06,
+      "loss": 0.1004,
+      "step": 26600
+    },
+    {
+      "epoch": 2.5194092028024997,
+      "grad_norm": 0.36117368936538696,
+      "learning_rate": 9.922363188790002e-06,
+      "loss": 0.1025,
+      "step": 26610
+    },
+    {
+      "epoch": 2.5203559931831094,
+      "grad_norm": 0.36743780970573425,
+      "learning_rate": 9.918576027267564e-06,
+      "loss": 0.0948,
+      "step": 26620
+    },
+    {
+      "epoch": 2.521302783563719,
+      "grad_norm": 0.40864408016204834,
+      "learning_rate": 9.914788865745126e-06,
+      "loss": 0.1033,
+      "step": 26630
+    },
+    {
+      "epoch": 2.522249573944329,
+      "grad_norm": 0.41886240243911743,
+      "learning_rate": 9.911001704222686e-06,
+      "loss": 0.0943,
+      "step": 26640
+    },
+    {
+      "epoch": 2.5231963643249387,
+      "grad_norm": 0.42005595564842224,
+      "learning_rate": 9.907214542700247e-06,
+      "loss": 0.1059,
+      "step": 26650
+    },
+    {
+      "epoch": 2.5241431547055484,
+      "grad_norm": 0.4357232451438904,
+      "learning_rate": 9.903427381177807e-06,
+      "loss": 0.1176,
+      "step": 26660
+    },
+    {
+      "epoch": 2.525089945086158,
+      "grad_norm": 0.3606035113334656,
+      "learning_rate": 9.89964021965537e-06,
+      "loss": 0.0933,
+      "step": 26670
+    },
+    {
+      "epoch": 2.526036735466768,
+      "grad_norm": 0.40714558959007263,
+      "learning_rate": 9.895853058132931e-06,
+      "loss": 0.1028,
+      "step": 26680
+    },
+    {
+      "epoch": 2.5269835258473776,
+      "grad_norm": 0.4298784136772156,
+      "learning_rate": 9.892065896610491e-06,
+      "loss": 0.0944,
+      "step": 26690
+    },
+    {
+      "epoch": 2.5279303162279874,
+      "grad_norm": 0.46234458684921265,
+      "learning_rate": 9.888278735088053e-06,
+      "loss": 0.1018,
+      "step": 26700
+    },
+    {
+      "epoch": 2.528877106608597,
+      "grad_norm": 0.3541281819343567,
+      "learning_rate": 9.884491573565613e-06,
+      "loss": 0.0992,
+      "step": 26710
+    },
+    {
+      "epoch": 2.529823896989207,
+      "grad_norm": 0.4032933712005615,
+      "learning_rate": 9.880704412043175e-06,
+      "loss": 0.1043,
+      "step": 26720
+    },
+    {
+      "epoch": 2.5307706873698166,
+      "grad_norm": 0.39109474420547485,
+      "learning_rate": 9.876917250520736e-06,
+      "loss": 0.1141,
+      "step": 26730
+    },
+    {
+      "epoch": 2.5317174777504263,
+      "grad_norm": 0.36314740777015686,
+      "learning_rate": 9.873130088998296e-06,
+      "loss": 0.1048,
+      "step": 26740
+    },
+    {
+      "epoch": 2.5326642681310356,
+      "grad_norm": 0.42757153511047363,
+      "learning_rate": 9.869342927475858e-06,
+      "loss": 0.0944,
+      "step": 26750
+    },
+    {
+      "epoch": 2.5336110585116454,
+      "grad_norm": 0.43337103724479675,
+      "learning_rate": 9.865555765953418e-06,
+      "loss": 0.1059,
+      "step": 26760
+    },
+    {
+      "epoch": 2.534557848892255,
+      "grad_norm": 0.615561842918396,
+      "learning_rate": 9.86176860443098e-06,
+      "loss": 0.1011,
+      "step": 26770
+    },
+    {
+      "epoch": 2.535504639272865,
+      "grad_norm": 0.3941485583782196,
+      "learning_rate": 9.85798144290854e-06,
+      "loss": 0.1097,
+      "step": 26780
+    },
+    {
+      "epoch": 2.5364514296534746,
+      "grad_norm": 0.5565608739852905,
+      "learning_rate": 9.854194281386102e-06,
+      "loss": 0.1034,
+      "step": 26790
+    },
+    {
+      "epoch": 2.5373982200340843,
+      "grad_norm": 0.3893771469593048,
+      "learning_rate": 9.850407119863663e-06,
+      "loss": 0.105,
+      "step": 26800
+    },
+    {
+      "epoch": 2.538345010414694,
+      "grad_norm": 0.36800989508628845,
+      "learning_rate": 9.846619958341223e-06,
+      "loss": 0.1047,
+      "step": 26810
+    },
+    {
+      "epoch": 2.539291800795304,
+      "grad_norm": 0.3624418377876282,
+      "learning_rate": 9.842832796818785e-06,
+      "loss": 0.0904,
+      "step": 26820
+    },
+    {
+      "epoch": 2.5402385911759136,
+      "grad_norm": 0.4615015387535095,
+      "learning_rate": 9.839045635296345e-06,
+      "loss": 0.1125,
+      "step": 26830
+    },
+    {
+      "epoch": 2.5411853815565233,
+      "grad_norm": 0.4052465558052063,
+      "learning_rate": 9.835258473773907e-06,
+      "loss": 0.102,
+      "step": 26840
+    },
+    {
+      "epoch": 2.542132171937133,
+      "grad_norm": 0.3502803444862366,
+      "learning_rate": 9.831471312251469e-06,
+      "loss": 0.1073,
+      "step": 26850
+    },
+    {
+      "epoch": 2.5430789623177428,
+      "grad_norm": 0.3207018971443176,
+      "learning_rate": 9.82768415072903e-06,
+      "loss": 0.0975,
+      "step": 26860
+    },
+    {
+      "epoch": 2.5440257526983525,
+      "grad_norm": 0.7084227204322815,
+      "learning_rate": 9.82389698920659e-06,
+      "loss": 0.103,
+      "step": 26870
+    },
+    {
+      "epoch": 2.5449725430789623,
+      "grad_norm": 0.4235849976539612,
+      "learning_rate": 9.82010982768415e-06,
+      "loss": 0.101,
+      "step": 26880
+    },
+    {
+      "epoch": 2.545919333459572,
+      "grad_norm": 0.43773093819618225,
+      "learning_rate": 9.816322666161712e-06,
+      "loss": 0.0937,
+      "step": 26890
+    },
+    {
+      "epoch": 2.5468661238401817,
+      "grad_norm": 0.42589303851127625,
+      "learning_rate": 9.812535504639274e-06,
+      "loss": 0.0968,
+      "step": 26900
+    },
+    {
+      "epoch": 2.5478129142207915,
+      "grad_norm": 0.41181427240371704,
+      "learning_rate": 9.808748343116836e-06,
+      "loss": 0.1082,
+      "step": 26910
+    },
+    {
+      "epoch": 2.548759704601401,
+      "grad_norm": 0.4209132492542267,
+      "learning_rate": 9.804961181594396e-06,
+      "loss": 0.1005,
+      "step": 26920
+    },
+    {
+      "epoch": 2.549706494982011,
+      "grad_norm": 0.445416659116745,
+      "learning_rate": 9.801174020071958e-06,
+      "loss": 0.0934,
+      "step": 26930
+    },
+    {
+      "epoch": 2.5506532853626207,
+      "grad_norm": 0.392095148563385,
+      "learning_rate": 9.797386858549518e-06,
+      "loss": 0.1022,
+      "step": 26940
+    },
+    {
+      "epoch": 2.5516000757432304,
+      "grad_norm": 0.46347376704216003,
+      "learning_rate": 9.79359969702708e-06,
+      "loss": 0.1007,
+      "step": 26950
+    },
+    {
+      "epoch": 2.55254686612384,
+      "grad_norm": 0.37809625267982483,
+      "learning_rate": 9.78981253550464e-06,
+      "loss": 0.1021,
+      "step": 26960
+    },
+    {
+      "epoch": 2.55349365650445,
+      "grad_norm": 0.3950808346271515,
+      "learning_rate": 9.786025373982201e-06,
+      "loss": 0.1119,
+      "step": 26970
+    },
+    {
+      "epoch": 2.5544404468850597,
+      "grad_norm": 0.4668138325214386,
+      "learning_rate": 9.782238212459763e-06,
+      "loss": 0.1073,
+      "step": 26980
+    },
+    {
+      "epoch": 2.5553872372656694,
+      "grad_norm": 0.5842813849449158,
+      "learning_rate": 9.778451050937323e-06,
+      "loss": 0.1027,
+      "step": 26990
+    },
+    {
+      "epoch": 2.556334027646279,
+      "grad_norm": 0.4209204316139221,
+      "learning_rate": 9.774663889414885e-06,
+      "loss": 0.1093,
+      "step": 27000
+    },
+    {
+      "epoch": 2.557280818026889,
+      "grad_norm": 0.3692864775657654,
+      "learning_rate": 9.770876727892445e-06,
+      "loss": 0.099,
+      "step": 27010
+    },
+    {
+      "epoch": 2.5582276084074986,
+      "grad_norm": 0.4216177761554718,
+      "learning_rate": 9.767089566370007e-06,
+      "loss": 0.0859,
+      "step": 27020
+    },
+    {
+      "epoch": 2.5591743987881084,
+      "grad_norm": 0.3516777753829956,
+      "learning_rate": 9.763302404847568e-06,
+      "loss": 0.0966,
+      "step": 27030
+    },
+    {
+      "epoch": 2.560121189168718,
+      "grad_norm": 0.43485236167907715,
+      "learning_rate": 9.759515243325128e-06,
+      "loss": 0.1052,
+      "step": 27040
+    },
+    {
+      "epoch": 2.561067979549328,
+      "grad_norm": 0.3804272711277008,
+      "learning_rate": 9.75572808180269e-06,
+      "loss": 0.1069,
+      "step": 27050
+    },
+    {
+      "epoch": 2.5620147699299376,
+      "grad_norm": 0.7018653750419617,
+      "learning_rate": 9.75194092028025e-06,
+      "loss": 0.1099,
+      "step": 27060
+    },
+    {
+      "epoch": 2.5629615603105473,
+      "grad_norm": 0.37473297119140625,
+      "learning_rate": 9.748153758757812e-06,
+      "loss": 0.1122,
+      "step": 27070
+    },
+    {
+      "epoch": 2.563908350691157,
+      "grad_norm": 0.4419376850128174,
+      "learning_rate": 9.744366597235374e-06,
+      "loss": 0.1087,
+      "step": 27080
+    },
+    {
+      "epoch": 2.564855141071767,
+      "grad_norm": 0.4710046648979187,
+      "learning_rate": 9.740579435712934e-06,
+      "loss": 0.0974,
+      "step": 27090
+    },
+    {
+      "epoch": 2.5658019314523766,
+      "grad_norm": 0.32833173871040344,
+      "learning_rate": 9.736792274190495e-06,
+      "loss": 0.1037,
+      "step": 27100
+    },
+    {
+      "epoch": 2.5667487218329863,
+      "grad_norm": 0.44157248735427856,
+      "learning_rate": 9.733005112668055e-06,
+      "loss": 0.1031,
+      "step": 27110
+    },
+    {
+      "epoch": 2.567695512213596,
+      "grad_norm": 0.370108038187027,
+      "learning_rate": 9.729217951145617e-06,
+      "loss": 0.1024,
+      "step": 27120
+    },
+    {
+      "epoch": 2.568642302594206,
+      "grad_norm": 0.4278753697872162,
+      "learning_rate": 9.725430789623179e-06,
+      "loss": 0.107,
+      "step": 27130
+    },
+    {
+      "epoch": 2.5695890929748155,
+      "grad_norm": 0.3966275751590729,
+      "learning_rate": 9.721643628100739e-06,
+      "loss": 0.1007,
+      "step": 27140
+    },
+    {
+      "epoch": 2.5705358833554253,
+      "grad_norm": 0.4091242253780365,
+      "learning_rate": 9.7178564665783e-06,
+      "loss": 0.092,
+      "step": 27150
+    },
+    {
+      "epoch": 2.571482673736035,
+      "grad_norm": 0.47113844752311707,
+      "learning_rate": 9.71406930505586e-06,
+      "loss": 0.1154,
+      "step": 27160
+    },
+    {
+      "epoch": 2.5724294641166443,
+      "grad_norm": 0.4091624915599823,
+      "learning_rate": 9.710282143533423e-06,
+      "loss": 0.0941,
+      "step": 27170
+    },
+    {
+      "epoch": 2.573376254497254,
+      "grad_norm": 0.3549819886684418,
+      "learning_rate": 9.706494982010983e-06,
+      "loss": 0.1018,
+      "step": 27180
+    },
+    {
+      "epoch": 2.574323044877864,
+      "grad_norm": 0.4599911868572235,
+      "learning_rate": 9.702707820488544e-06,
+      "loss": 0.0995,
+      "step": 27190
+    },
+    {
+      "epoch": 2.5752698352584735,
+      "grad_norm": 0.39354264736175537,
+      "learning_rate": 9.698920658966106e-06,
+      "loss": 0.1001,
+      "step": 27200
+    },
+    {
+      "epoch": 2.5762166256390833,
+      "grad_norm": 0.6075589060783386,
+      "learning_rate": 9.695133497443668e-06,
+      "loss": 0.0988,
+      "step": 27210
+    },
+    {
+      "epoch": 2.577163416019693,
+      "grad_norm": 0.48551851511001587,
+      "learning_rate": 9.691346335921228e-06,
+      "loss": 0.1041,
+      "step": 27220
+    },
+    {
+      "epoch": 2.5781102064003028,
+      "grad_norm": 0.4040965139865875,
+      "learning_rate": 9.687559174398788e-06,
+      "loss": 0.1081,
+      "step": 27230
+    },
+    {
+      "epoch": 2.5790569967809125,
+      "grad_norm": 0.36837923526763916,
+      "learning_rate": 9.68377201287635e-06,
+      "loss": 0.0972,
+      "step": 27240
+    },
+    {
+      "epoch": 2.5800037871615222,
+      "grad_norm": 0.5304983258247375,
+      "learning_rate": 9.679984851353911e-06,
+      "loss": 0.0971,
+      "step": 27250
+    },
+    {
+      "epoch": 2.580950577542132,
+      "grad_norm": 0.4093058705329895,
+      "learning_rate": 9.676197689831473e-06,
+      "loss": 0.0941,
+      "step": 27260
+    },
+    {
+      "epoch": 2.5818973679227417,
+      "grad_norm": 0.4903060793876648,
+      "learning_rate": 9.672410528309033e-06,
+      "loss": 0.1164,
+      "step": 27270
+    },
+    {
+      "epoch": 2.5828441583033515,
+      "grad_norm": 0.41253364086151123,
+      "learning_rate": 9.668623366786595e-06,
+      "loss": 0.1075,
+      "step": 27280
+    },
+    {
+      "epoch": 2.583790948683961,
+      "grad_norm": 0.4047181010246277,
+      "learning_rate": 9.664836205264155e-06,
+      "loss": 0.0968,
+      "step": 27290
+    },
+    {
+      "epoch": 2.584737739064571,
+      "grad_norm": 0.5090596079826355,
+      "learning_rate": 9.661049043741717e-06,
+      "loss": 0.1035,
+      "step": 27300
+    },
+    {
+      "epoch": 2.5856845294451807,
+      "grad_norm": 0.378775030374527,
+      "learning_rate": 9.657261882219278e-06,
+      "loss": 0.1019,
+      "step": 27310
+    },
+    {
+      "epoch": 2.5866313198257904,
+      "grad_norm": 0.45318761467933655,
+      "learning_rate": 9.653474720696839e-06,
+      "loss": 0.107,
+      "step": 27320
+    },
+    {
+      "epoch": 2.5875781102064,
+      "grad_norm": 0.35085242986679077,
+      "learning_rate": 9.6496875591744e-06,
+      "loss": 0.1012,
+      "step": 27330
+    },
+    {
+      "epoch": 2.58852490058701,
+      "grad_norm": 0.31217142939567566,
+      "learning_rate": 9.64590039765196e-06,
+      "loss": 0.1014,
+      "step": 27340
+    },
+    {
+      "epoch": 2.5894716909676196,
+      "grad_norm": 0.3551445007324219,
+      "learning_rate": 9.642113236129522e-06,
+      "loss": 0.0934,
+      "step": 27350
+    },
+    {
+      "epoch": 2.5904184813482294,
+      "grad_norm": 0.37578850984573364,
+      "learning_rate": 9.638326074607082e-06,
+      "loss": 0.1089,
+      "step": 27360
+    },
+    {
+      "epoch": 2.591365271728839,
+      "grad_norm": 0.3457893133163452,
+      "learning_rate": 9.634538913084644e-06,
+      "loss": 0.1089,
+      "step": 27370
+    },
+    {
+      "epoch": 2.592312062109449,
+      "grad_norm": 0.5415358543395996,
+      "learning_rate": 9.630751751562206e-06,
+      "loss": 0.1047,
+      "step": 27380
+    },
+    {
+      "epoch": 2.5932588524900586,
+      "grad_norm": 0.4887268543243408,
+      "learning_rate": 9.626964590039766e-06,
+      "loss": 0.1154,
+      "step": 27390
+    },
+    {
+      "epoch": 2.5942056428706683,
+      "grad_norm": 0.348284512758255,
+      "learning_rate": 9.623177428517327e-06,
+      "loss": 0.0981,
+      "step": 27400
+    },
+    {
+      "epoch": 2.595152433251278,
+      "grad_norm": 0.3595815896987915,
+      "learning_rate": 9.619390266994887e-06,
+      "loss": 0.1144,
+      "step": 27410
+    },
+    {
+      "epoch": 2.596099223631888,
+      "grad_norm": 0.4330320656299591,
+      "learning_rate": 9.61560310547245e-06,
+      "loss": 0.1115,
+      "step": 27420
+    },
+    {
+      "epoch": 2.5970460140124976,
+      "grad_norm": 0.40695780515670776,
+      "learning_rate": 9.611815943950011e-06,
+      "loss": 0.1044,
+      "step": 27430
+    },
+    {
+      "epoch": 2.5979928043931073,
+      "grad_norm": 0.3428425192832947,
+      "learning_rate": 9.608028782427571e-06,
+      "loss": 0.0991,
+      "step": 27440
+    },
+    {
+      "epoch": 2.598939594773717,
+      "grad_norm": 0.4021109938621521,
+      "learning_rate": 9.604241620905133e-06,
+      "loss": 0.1029,
+      "step": 27450
+    },
+    {
+      "epoch": 2.599886385154327,
+      "grad_norm": 0.4028293490409851,
+      "learning_rate": 9.600454459382693e-06,
+      "loss": 0.106,
+      "step": 27460
+    },
+    {
+      "epoch": 2.6008331755349365,
+      "grad_norm": 0.4052664041519165,
+      "learning_rate": 9.596667297860255e-06,
+      "loss": 0.0996,
+      "step": 27470
+    },
+    {
+      "epoch": 2.6017799659155463,
+      "grad_norm": 0.4120866060256958,
+      "learning_rate": 9.592880136337816e-06,
+      "loss": 0.0961,
+      "step": 27480
+    },
+    {
+      "epoch": 2.602726756296156,
+      "grad_norm": 0.455219566822052,
+      "learning_rate": 9.589092974815378e-06,
+      "loss": 0.1044,
+      "step": 27490
+    },
+    {
+      "epoch": 2.6036735466767658,
+      "grad_norm": 0.45610177516937256,
+      "learning_rate": 9.585305813292938e-06,
+      "loss": 0.1042,
+      "step": 27500
+    },
+    {
+      "epoch": 2.6046203370573755,
+      "grad_norm": 0.42534270882606506,
+      "learning_rate": 9.581518651770498e-06,
+      "loss": 0.0997,
+      "step": 27510
+    },
+    {
+      "epoch": 2.6055671274379852,
+      "grad_norm": 0.45785534381866455,
+      "learning_rate": 9.57773149024806e-06,
+      "loss": 0.0963,
+      "step": 27520
+    },
+    {
+      "epoch": 2.606513917818595,
+      "grad_norm": 0.3305295705795288,
+      "learning_rate": 9.57394432872562e-06,
+      "loss": 0.1055,
+      "step": 27530
+    },
+    {
+      "epoch": 2.6074607081992047,
+      "grad_norm": 0.3611030578613281,
+      "learning_rate": 9.570157167203182e-06,
+      "loss": 0.1057,
+      "step": 27540
+    },
+    {
+      "epoch": 2.6084074985798145,
+      "grad_norm": 0.4136884808540344,
+      "learning_rate": 9.566370005680743e-06,
+      "loss": 0.1059,
+      "step": 27550
+    },
+    {
+      "epoch": 2.609354288960424,
+      "grad_norm": 0.5084044933319092,
+      "learning_rate": 9.562582844158305e-06,
+      "loss": 0.1014,
+      "step": 27560
+    },
+    {
+      "epoch": 2.610301079341034,
+      "grad_norm": 0.4754091799259186,
+      "learning_rate": 9.558795682635865e-06,
+      "loss": 0.1052,
+      "step": 27570
+    },
+    {
+      "epoch": 2.6112478697216437,
+      "grad_norm": 0.47309383749961853,
+      "learning_rate": 9.555008521113425e-06,
+      "loss": 0.1032,
+      "step": 27580
+    },
+    {
+      "epoch": 2.6121946601022534,
+      "grad_norm": 0.40714767575263977,
+      "learning_rate": 9.551221359590987e-06,
+      "loss": 0.1038,
+      "step": 27590
+    },
+    {
+      "epoch": 2.613141450482863,
+      "grad_norm": 0.5299632549285889,
+      "learning_rate": 9.547434198068549e-06,
+      "loss": 0.1031,
+      "step": 27600
+    },
+    {
+      "epoch": 2.614088240863473,
+      "grad_norm": 0.4505506157875061,
+      "learning_rate": 9.54364703654611e-06,
+      "loss": 0.1007,
+      "step": 27610
+    },
+    {
+      "epoch": 2.6150350312440827,
+      "grad_norm": 0.40479910373687744,
+      "learning_rate": 9.53985987502367e-06,
+      "loss": 0.0999,
+      "step": 27620
+    },
+    {
+      "epoch": 2.6159818216246924,
+      "grad_norm": 0.4297819137573242,
+      "learning_rate": 9.536072713501232e-06,
+      "loss": 0.0984,
+      "step": 27630
+    },
+    {
+      "epoch": 2.616928612005302,
+      "grad_norm": 0.48748689889907837,
+      "learning_rate": 9.532285551978792e-06,
+      "loss": 0.1042,
+      "step": 27640
+    },
+    {
+      "epoch": 2.617875402385912,
+      "grad_norm": 0.3475923240184784,
+      "learning_rate": 9.528498390456354e-06,
+      "loss": 0.1109,
+      "step": 27650
+    },
+    {
+      "epoch": 2.6188221927665216,
+      "grad_norm": 0.4447574019432068,
+      "learning_rate": 9.524711228933916e-06,
+      "loss": 0.0934,
+      "step": 27660
+    },
+    {
+      "epoch": 2.6197689831471314,
+      "grad_norm": 0.44557541608810425,
+      "learning_rate": 9.520924067411476e-06,
+      "loss": 0.1007,
+      "step": 27670
+    },
+    {
+      "epoch": 2.620715773527741,
+      "grad_norm": 0.32018914818763733,
+      "learning_rate": 9.517136905889038e-06,
+      "loss": 0.0991,
+      "step": 27680
+    },
+    {
+      "epoch": 2.621662563908351,
+      "grad_norm": 0.5814476609230042,
+      "learning_rate": 9.513349744366598e-06,
+      "loss": 0.1069,
+      "step": 27690
+    },
+    {
+      "epoch": 2.6226093542889606,
+      "grad_norm": 0.3966127634048462,
+      "learning_rate": 9.50956258284416e-06,
+      "loss": 0.0957,
+      "step": 27700
+    },
+    {
+      "epoch": 2.6235561446695703,
+      "grad_norm": 0.420902281999588,
+      "learning_rate": 9.50577542132172e-06,
+      "loss": 0.1155,
+      "step": 27710
+    },
+    {
+      "epoch": 2.62450293505018,
+      "grad_norm": 0.4212897717952728,
+      "learning_rate": 9.501988259799281e-06,
+      "loss": 0.101,
+      "step": 27720
+    },
+    {
+      "epoch": 2.62544972543079,
+      "grad_norm": 0.40939393639564514,
+      "learning_rate": 9.498201098276843e-06,
+      "loss": 0.0981,
+      "step": 27730
+    },
+    {
+      "epoch": 2.6263965158113995,
+      "grad_norm": 0.5749149918556213,
+      "learning_rate": 9.494413936754403e-06,
+      "loss": 0.1014,
+      "step": 27740
+    },
+    {
+      "epoch": 2.6273433061920093,
+      "grad_norm": 0.5041487812995911,
+      "learning_rate": 9.490626775231965e-06,
+      "loss": 0.099,
+      "step": 27750
+    },
+    {
+      "epoch": 2.628290096572619,
+      "grad_norm": 0.41404733061790466,
+      "learning_rate": 9.486839613709525e-06,
+      "loss": 0.0995,
+      "step": 27760
+    },
+    {
+      "epoch": 2.6292368869532288,
+      "grad_norm": 0.41370484232902527,
+      "learning_rate": 9.483052452187086e-06,
+      "loss": 0.0932,
+      "step": 27770
+    },
+    {
+      "epoch": 2.6301836773338385,
+      "grad_norm": 0.3428075909614563,
+      "learning_rate": 9.479265290664648e-06,
+      "loss": 0.0948,
+      "step": 27780
+    },
+    {
+      "epoch": 2.6311304677144483,
+      "grad_norm": 0.4095135033130646,
+      "learning_rate": 9.475478129142208e-06,
+      "loss": 0.1154,
+      "step": 27790
+    },
+    {
+      "epoch": 2.632077258095058,
+      "grad_norm": 0.40821513533592224,
+      "learning_rate": 9.47169096761977e-06,
+      "loss": 0.0991,
+      "step": 27800
+    },
+    {
+      "epoch": 2.6330240484756677,
+      "grad_norm": 0.46019446849823,
+      "learning_rate": 9.46790380609733e-06,
+      "loss": 0.1024,
+      "step": 27810
+    },
+    {
+      "epoch": 2.6339708388562775,
+      "grad_norm": 0.4173412024974823,
+      "learning_rate": 9.464116644574892e-06,
+      "loss": 0.1104,
+      "step": 27820
+    },
+    {
+      "epoch": 2.634917629236887,
+      "grad_norm": 0.39484941959381104,
+      "learning_rate": 9.460329483052454e-06,
+      "loss": 0.1056,
+      "step": 27830
+    },
+    {
+      "epoch": 2.635864419617497,
+      "grad_norm": 0.41165807843208313,
+      "learning_rate": 9.456542321530015e-06,
+      "loss": 0.1044,
+      "step": 27840
+    },
+    {
+      "epoch": 2.6368112099981063,
+      "grad_norm": 0.38451337814331055,
+      "learning_rate": 9.452755160007575e-06,
+      "loss": 0.0996,
+      "step": 27850
+    },
+    {
+      "epoch": 2.637758000378716,
+      "grad_norm": 0.382441908121109,
+      "learning_rate": 9.448967998485135e-06,
+      "loss": 0.1104,
+      "step": 27860
+    },
+    {
+      "epoch": 2.6387047907593257,
+      "grad_norm": 0.32897433638572693,
+      "learning_rate": 9.445180836962697e-06,
+      "loss": 0.0944,
+      "step": 27870
+    },
+    {
+      "epoch": 2.6396515811399355,
+      "grad_norm": 0.4390844404697418,
+      "learning_rate": 9.441393675440257e-06,
+      "loss": 0.1113,
+      "step": 27880
+    },
+    {
+      "epoch": 2.640598371520545,
+      "grad_norm": 0.38016265630722046,
+      "learning_rate": 9.437606513917819e-06,
+      "loss": 0.1019,
+      "step": 27890
+    },
+    {
+      "epoch": 2.641545161901155,
+      "grad_norm": 0.4154101610183716,
+      "learning_rate": 9.43381935239538e-06,
+      "loss": 0.1052,
+      "step": 27900
+    },
+    {
+      "epoch": 2.6424919522817647,
+      "grad_norm": 0.4230835437774658,
+      "learning_rate": 9.430032190872942e-06,
+      "loss": 0.108,
+      "step": 27910
+    },
+    {
+      "epoch": 2.6434387426623744,
+      "grad_norm": 0.41765904426574707,
+      "learning_rate": 9.426245029350502e-06,
+      "loss": 0.1081,
+      "step": 27920
+    },
+    {
+      "epoch": 2.644385533042984,
+      "grad_norm": 0.3967103362083435,
+      "learning_rate": 9.422457867828063e-06,
+      "loss": 0.0963,
+      "step": 27930
+    },
+    {
+      "epoch": 2.645332323423594,
+      "grad_norm": 0.39287203550338745,
+      "learning_rate": 9.418670706305624e-06,
+      "loss": 0.1028,
+      "step": 27940
+    },
+    {
+      "epoch": 2.6462791138042037,
+      "grad_norm": 0.38825488090515137,
+      "learning_rate": 9.414883544783186e-06,
+      "loss": 0.1122,
+      "step": 27950
+    },
+    {
+      "epoch": 2.6472259041848134,
+      "grad_norm": 0.3866090178489685,
+      "learning_rate": 9.411096383260748e-06,
+      "loss": 0.1035,
+      "step": 27960
+    },
+    {
+      "epoch": 2.648172694565423,
+      "grad_norm": 0.5346670150756836,
+      "learning_rate": 9.407309221738308e-06,
+      "loss": 0.1051,
+      "step": 27970
+    },
+    {
+      "epoch": 2.649119484946033,
+      "grad_norm": 0.3844970464706421,
+      "learning_rate": 9.40352206021587e-06,
+      "loss": 0.1082,
+      "step": 27980
+    },
+    {
+      "epoch": 2.6500662753266426,
+      "grad_norm": 0.39131975173950195,
+      "learning_rate": 9.39973489869343e-06,
+      "loss": 0.1047,
+      "step": 27990
+    },
+    {
+      "epoch": 2.6510130657072524,
+      "grad_norm": 0.3815428912639618,
+      "learning_rate": 9.395947737170991e-06,
+      "loss": 0.1039,
+      "step": 28000
+    },
+    {
+      "epoch": 2.651959856087862,
+      "grad_norm": 0.42498138546943665,
+      "learning_rate": 9.392160575648553e-06,
+      "loss": 0.1035,
+      "step": 28010
+    },
+    {
+      "epoch": 2.652906646468472,
+      "grad_norm": 0.32439425587654114,
+      "learning_rate": 9.388373414126113e-06,
+      "loss": 0.1077,
+      "step": 28020
+    },
+    {
+      "epoch": 2.6538534368490816,
+      "grad_norm": 0.40164288878440857,
+      "learning_rate": 9.384586252603675e-06,
+      "loss": 0.1024,
+      "step": 28030
+    },
+    {
+      "epoch": 2.6548002272296913,
+      "grad_norm": 0.4523821771144867,
+      "learning_rate": 9.380799091081235e-06,
+      "loss": 0.1006,
+      "step": 28040
+    },
+    {
+      "epoch": 2.655747017610301,
+      "grad_norm": 0.4752205014228821,
+      "learning_rate": 9.377011929558797e-06,
+      "loss": 0.0968,
+      "step": 28050
+    },
+    {
+      "epoch": 2.656693807990911,
+      "grad_norm": 0.3431170880794525,
+      "learning_rate": 9.373224768036357e-06,
+      "loss": 0.1011,
+      "step": 28060
+    },
+    {
+      "epoch": 2.6576405983715206,
+      "grad_norm": 0.37541231513023376,
+      "learning_rate": 9.369437606513918e-06,
+      "loss": 0.1063,
+      "step": 28070
+    },
+    {
+      "epoch": 2.6585873887521303,
+      "grad_norm": 0.3426431715488434,
+      "learning_rate": 9.36565044499148e-06,
+      "loss": 0.1008,
+      "step": 28080
+    },
+    {
+      "epoch": 2.65953417913274,
+      "grad_norm": 0.2883263826370239,
+      "learning_rate": 9.36186328346904e-06,
+      "loss": 0.0901,
+      "step": 28090
+    },
+    {
+      "epoch": 2.66048096951335,
+      "grad_norm": 0.5053014755249023,
+      "learning_rate": 9.358076121946602e-06,
+      "loss": 0.1126,
+      "step": 28100
+    },
+    {
+      "epoch": 2.6614277598939595,
+      "grad_norm": 0.5336264967918396,
+      "learning_rate": 9.354288960424162e-06,
+      "loss": 0.1027,
+      "step": 28110
+    },
+    {
+      "epoch": 2.6623745502745693,
+      "grad_norm": 0.3703152537345886,
+      "learning_rate": 9.350501798901724e-06,
+      "loss": 0.0943,
+      "step": 28120
+    },
+    {
+      "epoch": 2.663321340655179,
+      "grad_norm": 0.35078686475753784,
+      "learning_rate": 9.346714637379286e-06,
+      "loss": 0.1003,
+      "step": 28130
+    },
+    {
+      "epoch": 2.6642681310357887,
+      "grad_norm": 0.6297541260719299,
+      "learning_rate": 9.342927475856846e-06,
+      "loss": 0.1065,
+      "step": 28140
+    },
+    {
+      "epoch": 2.6652149214163985,
+      "grad_norm": 0.4121268093585968,
+      "learning_rate": 9.339140314334407e-06,
+      "loss": 0.0953,
+      "step": 28150
+    },
+    {
+      "epoch": 2.6661617117970082,
+      "grad_norm": 0.3645578622817993,
+      "learning_rate": 9.335353152811967e-06,
+      "loss": 0.095,
+      "step": 28160
+    },
+    {
+      "epoch": 2.667108502177618,
+      "grad_norm": 0.36650314927101135,
+      "learning_rate": 9.331565991289529e-06,
+      "loss": 0.0971,
+      "step": 28170
+    },
+    {
+      "epoch": 2.6680552925582277,
+      "grad_norm": 0.41033312678337097,
+      "learning_rate": 9.327778829767091e-06,
+      "loss": 0.1084,
+      "step": 28180
+    },
+    {
+      "epoch": 2.6690020829388375,
+      "grad_norm": 0.3698018193244934,
+      "learning_rate": 9.323991668244653e-06,
+      "loss": 0.0964,
+      "step": 28190
+    },
+    {
+      "epoch": 2.669948873319447,
+      "grad_norm": 0.4807288944721222,
+      "learning_rate": 9.320204506722213e-06,
+      "loss": 0.0984,
+      "step": 28200
+    },
+    {
+      "epoch": 2.670895663700057,
+      "grad_norm": 0.41571494936943054,
+      "learning_rate": 9.316417345199773e-06,
+      "loss": 0.1018,
+      "step": 28210
+    },
+    {
+      "epoch": 2.6718424540806667,
+      "grad_norm": 0.48666346073150635,
+      "learning_rate": 9.312630183677334e-06,
+      "loss": 0.0965,
+      "step": 28220
+    },
+    {
+      "epoch": 2.6727892444612764,
+      "grad_norm": 0.39642250537872314,
+      "learning_rate": 9.308843022154895e-06,
+      "loss": 0.0924,
+      "step": 28230
+    },
+    {
+      "epoch": 2.673736034841886,
+      "grad_norm": 0.5108928680419922,
+      "learning_rate": 9.305055860632456e-06,
+      "loss": 0.0957,
+      "step": 28240
+    },
+    {
+      "epoch": 2.674682825222496,
+      "grad_norm": 0.36331531405448914,
+      "learning_rate": 9.301268699110018e-06,
+      "loss": 0.1091,
+      "step": 28250
+    },
+    {
+      "epoch": 2.6756296156031056,
+      "grad_norm": 0.42903733253479004,
+      "learning_rate": 9.29748153758758e-06,
+      "loss": 0.0933,
+      "step": 28260
+    },
+    {
+      "epoch": 2.676576405983715,
+      "grad_norm": 0.4590227007865906,
+      "learning_rate": 9.29369437606514e-06,
+      "loss": 0.1076,
+      "step": 28270
+    },
+    {
+      "epoch": 2.6775231963643247,
+      "grad_norm": 0.3357694447040558,
+      "learning_rate": 9.2899072145427e-06,
+      "loss": 0.0952,
+      "step": 28280
+    },
+    {
+      "epoch": 2.6784699867449344,
+      "grad_norm": 0.441007137298584,
+      "learning_rate": 9.286120053020262e-06,
+      "loss": 0.096,
+      "step": 28290
+    },
+    {
+      "epoch": 2.679416777125544,
+      "grad_norm": 0.47577551007270813,
+      "learning_rate": 9.282332891497823e-06,
+      "loss": 0.1044,
+      "step": 28300
+    },
+    {
+      "epoch": 2.680363567506154,
+      "grad_norm": 0.3424263298511505,
+      "learning_rate": 9.278545729975385e-06,
+      "loss": 0.1007,
+      "step": 28310
+    },
+    {
+      "epoch": 2.6813103578867636,
+      "grad_norm": 0.46387240290641785,
+      "learning_rate": 9.274758568452945e-06,
+      "loss": 0.1009,
+      "step": 28320
+    },
+    {
+      "epoch": 2.6822571482673734,
+      "grad_norm": 0.3691031336784363,
+      "learning_rate": 9.270971406930507e-06,
+      "loss": 0.0984,
+      "step": 28330
+    },
+    {
+      "epoch": 2.683203938647983,
+      "grad_norm": 0.42378664016723633,
+      "learning_rate": 9.267184245408067e-06,
+      "loss": 0.1095,
+      "step": 28340
+    },
+    {
+      "epoch": 2.684150729028593,
+      "grad_norm": 0.4089430570602417,
+      "learning_rate": 9.263397083885629e-06,
+      "loss": 0.1112,
+      "step": 28350
+    },
+    {
+      "epoch": 2.6850975194092026,
+      "grad_norm": 0.41893911361694336,
+      "learning_rate": 9.25960992236319e-06,
+      "loss": 0.1007,
+      "step": 28360
+    },
+    {
+      "epoch": 2.6860443097898123,
+      "grad_norm": 0.40327945351600647,
+      "learning_rate": 9.25582276084075e-06,
+      "loss": 0.113,
+      "step": 28370
+    },
+    {
+      "epoch": 2.686991100170422,
+      "grad_norm": 0.39842647314071655,
+      "learning_rate": 9.252035599318312e-06,
+      "loss": 0.1074,
+      "step": 28380
+    },
+    {
+      "epoch": 2.687937890551032,
+      "grad_norm": 0.36703211069107056,
+      "learning_rate": 9.248248437795872e-06,
+      "loss": 0.1043,
+      "step": 28390
+    },
+    {
+      "epoch": 2.6888846809316416,
+      "grad_norm": 0.44750773906707764,
+      "learning_rate": 9.244461276273434e-06,
+      "loss": 0.1092,
+      "step": 28400
+    },
+    {
+      "epoch": 2.6898314713122513,
+      "grad_norm": 0.42375195026397705,
+      "learning_rate": 9.240674114750994e-06,
+      "loss": 0.1054,
+      "step": 28410
+    },
+    {
+      "epoch": 2.690778261692861,
+      "grad_norm": 0.4005579650402069,
+      "learning_rate": 9.236886953228556e-06,
+      "loss": 0.1114,
+      "step": 28420
+    },
+    {
+      "epoch": 2.691725052073471,
+      "grad_norm": 0.3783735930919647,
+      "learning_rate": 9.233099791706118e-06,
+      "loss": 0.0922,
+      "step": 28430
+    },
+    {
+      "epoch": 2.6926718424540805,
+      "grad_norm": 0.3596361577510834,
+      "learning_rate": 9.229312630183678e-06,
+      "loss": 0.1046,
+      "step": 28440
+    },
+    {
+      "epoch": 2.6936186328346903,
+      "grad_norm": 0.3943907916545868,
+      "learning_rate": 9.22552546866124e-06,
+      "loss": 0.1097,
+      "step": 28450
+    },
+    {
+      "epoch": 2.6945654232153,
+      "grad_norm": 0.3750338852405548,
+      "learning_rate": 9.2217383071388e-06,
+      "loss": 0.1066,
+      "step": 28460
+    },
+    {
+      "epoch": 2.6955122135959098,
+      "grad_norm": 0.34630802273750305,
+      "learning_rate": 9.217951145616361e-06,
+      "loss": 0.0975,
+      "step": 28470
+    },
+    {
+      "epoch": 2.6964590039765195,
+      "grad_norm": 0.3319561779499054,
+      "learning_rate": 9.214163984093923e-06,
+      "loss": 0.0873,
+      "step": 28480
+    },
+    {
+      "epoch": 2.6974057943571292,
+      "grad_norm": 0.4126971364021301,
+      "learning_rate": 9.210376822571483e-06,
+      "loss": 0.1056,
+      "step": 28490
+    },
+    {
+      "epoch": 2.698352584737739,
+      "grad_norm": 0.35686981678009033,
+      "learning_rate": 9.206589661049045e-06,
+      "loss": 0.0919,
+      "step": 28500
+    },
+    {
+      "epoch": 2.6992993751183487,
+      "grad_norm": 0.3125929832458496,
+      "learning_rate": 9.202802499526605e-06,
+      "loss": 0.1018,
+      "step": 28510
+    },
+    {
+      "epoch": 2.7002461654989585,
+      "grad_norm": 0.47137147188186646,
+      "learning_rate": 9.199015338004166e-06,
+      "loss": 0.0992,
+      "step": 28520
+    },
+    {
+      "epoch": 2.701192955879568,
+      "grad_norm": 0.34128260612487793,
+      "learning_rate": 9.195228176481728e-06,
+      "loss": 0.1002,
+      "step": 28530
+    },
+    {
+      "epoch": 2.702139746260178,
+      "grad_norm": 0.3971148133277893,
+      "learning_rate": 9.19144101495929e-06,
+      "loss": 0.0989,
+      "step": 28540
+    },
+    {
+      "epoch": 2.7030865366407877,
+      "grad_norm": 0.5022692680358887,
+      "learning_rate": 9.18765385343685e-06,
+      "loss": 0.1152,
+      "step": 28550
+    },
+    {
+      "epoch": 2.7040333270213974,
+      "grad_norm": 0.4363287687301636,
+      "learning_rate": 9.18386669191441e-06,
+      "loss": 0.1099,
+      "step": 28560
+    },
+    {
+      "epoch": 2.704980117402007,
+      "grad_norm": 0.46551355719566345,
+      "learning_rate": 9.180079530391972e-06,
+      "loss": 0.1023,
+      "step": 28570
+    },
+    {
+      "epoch": 2.705926907782617,
+      "grad_norm": 0.32491248846054077,
+      "learning_rate": 9.176292368869534e-06,
+      "loss": 0.0966,
+      "step": 28580
+    },
+    {
+      "epoch": 2.7068736981632266,
+      "grad_norm": 0.46695515513420105,
+      "learning_rate": 9.172505207347094e-06,
+      "loss": 0.1006,
+      "step": 28590
+    },
+    {
+      "epoch": 2.7078204885438364,
+      "grad_norm": 0.38846078515052795,
+      "learning_rate": 9.168718045824655e-06,
+      "loss": 0.11,
+      "step": 28600
+    },
+    {
+      "epoch": 2.708767278924446,
+      "grad_norm": 0.4434100091457367,
+      "learning_rate": 9.164930884302217e-06,
+      "loss": 0.1016,
+      "step": 28610
+    },
+    {
+      "epoch": 2.709714069305056,
+      "grad_norm": 0.5759299993515015,
+      "learning_rate": 9.161143722779777e-06,
+      "loss": 0.101,
+      "step": 28620
+    },
+    {
+      "epoch": 2.7106608596856656,
+      "grad_norm": 0.4500638246536255,
+      "learning_rate": 9.157356561257337e-06,
+      "loss": 0.1084,
+      "step": 28630
+    },
+    {
+      "epoch": 2.7116076500662754,
+      "grad_norm": 0.4416857659816742,
+      "learning_rate": 9.153569399734899e-06,
+      "loss": 0.1003,
+      "step": 28640
+    },
+    {
+      "epoch": 2.712554440446885,
+      "grad_norm": 0.35741469264030457,
+      "learning_rate": 9.14978223821246e-06,
+      "loss": 0.1061,
+      "step": 28650
+    },
+    {
+      "epoch": 2.713501230827495,
+      "grad_norm": 0.39139777421951294,
+      "learning_rate": 9.145995076690022e-06,
+      "loss": 0.0985,
+      "step": 28660
+    },
+    {
+      "epoch": 2.7144480212081046,
+      "grad_norm": 0.778073787689209,
+      "learning_rate": 9.142207915167582e-06,
+      "loss": 0.1092,
+      "step": 28670
+    },
+    {
+      "epoch": 2.7153948115887143,
+      "grad_norm": 0.4162473976612091,
+      "learning_rate": 9.138420753645144e-06,
+      "loss": 0.1075,
+      "step": 28680
+    },
+    {
+      "epoch": 2.716341601969324,
+      "grad_norm": 0.4190244972705841,
+      "learning_rate": 9.134633592122704e-06,
+      "loss": 0.1046,
+      "step": 28690
+    },
+    {
+      "epoch": 2.717288392349934,
+      "grad_norm": 0.4655373990535736,
+      "learning_rate": 9.130846430600266e-06,
+      "loss": 0.1061,
+      "step": 28700
+    },
+    {
+      "epoch": 2.7182351827305435,
+      "grad_norm": 0.43157628178596497,
+      "learning_rate": 9.127059269077828e-06,
+      "loss": 0.104,
+      "step": 28710
+    },
+    {
+      "epoch": 2.7191819731111533,
+      "grad_norm": 0.35439619421958923,
+      "learning_rate": 9.123272107555388e-06,
+      "loss": 0.0998,
+      "step": 28720
+    },
+    {
+      "epoch": 2.720128763491763,
+      "grad_norm": 0.4340245723724365,
+      "learning_rate": 9.11948494603295e-06,
+      "loss": 0.1002,
+      "step": 28730
+    },
+    {
+      "epoch": 2.7210755538723728,
+      "grad_norm": 0.40614044666290283,
+      "learning_rate": 9.11569778451051e-06,
+      "loss": 0.0958,
+      "step": 28740
+    },
+    {
+      "epoch": 2.7220223442529825,
+      "grad_norm": 0.4191066324710846,
+      "learning_rate": 9.111910622988071e-06,
+      "loss": 0.0966,
+      "step": 28750
+    },
+    {
+      "epoch": 2.7229691346335922,
+      "grad_norm": 0.4126628637313843,
+      "learning_rate": 9.108123461465633e-06,
+      "loss": 0.1123,
+      "step": 28760
+    },
+    {
+      "epoch": 2.723915925014202,
+      "grad_norm": 0.515906035900116,
+      "learning_rate": 9.104336299943193e-06,
+      "loss": 0.1094,
+      "step": 28770
+    },
+    {
+      "epoch": 2.7248627153948117,
+      "grad_norm": 0.3944433331489563,
+      "learning_rate": 9.100549138420755e-06,
+      "loss": 0.1047,
+      "step": 28780
+    },
+    {
+      "epoch": 2.7258095057754215,
+      "grad_norm": 0.5118011832237244,
+      "learning_rate": 9.096761976898315e-06,
+      "loss": 0.1007,
+      "step": 28790
+    },
+    {
+      "epoch": 2.726756296156031,
+      "grad_norm": 0.4848870038986206,
+      "learning_rate": 9.092974815375877e-06,
+      "loss": 0.1019,
+      "step": 28800
+    },
+    {
+      "epoch": 2.727703086536641,
+      "grad_norm": 0.5556014776229858,
+      "learning_rate": 9.089187653853437e-06,
+      "loss": 0.1018,
+      "step": 28810
+    },
+    {
+      "epoch": 2.7286498769172507,
+      "grad_norm": 0.4127148687839508,
+      "learning_rate": 9.085400492330998e-06,
+      "loss": 0.1007,
+      "step": 28820
+    },
+    {
+      "epoch": 2.7295966672978604,
+      "grad_norm": 0.4346880614757538,
+      "learning_rate": 9.08161333080856e-06,
+      "loss": 0.094,
+      "step": 28830
+    },
+    {
+      "epoch": 2.73054345767847,
+      "grad_norm": 0.41865330934524536,
+      "learning_rate": 9.07782616928612e-06,
+      "loss": 0.0938,
+      "step": 28840
+    },
+    {
+      "epoch": 2.73149024805908,
+      "grad_norm": 0.3504531681537628,
+      "learning_rate": 9.074039007763682e-06,
+      "loss": 0.096,
+      "step": 28850
+    },
+    {
+      "epoch": 2.7324370384396897,
+      "grad_norm": 0.43708714842796326,
+      "learning_rate": 9.070251846241242e-06,
+      "loss": 0.1066,
+      "step": 28860
+    },
+    {
+      "epoch": 2.7333838288202994,
+      "grad_norm": 0.3688562512397766,
+      "learning_rate": 9.066464684718804e-06,
+      "loss": 0.0983,
+      "step": 28870
+    },
+    {
+      "epoch": 2.734330619200909,
+      "grad_norm": 0.41222694516181946,
+      "learning_rate": 9.062677523196365e-06,
+      "loss": 0.1142,
+      "step": 28880
+    },
+    {
+      "epoch": 2.735277409581519,
+      "grad_norm": 0.329771488904953,
+      "learning_rate": 9.058890361673927e-06,
+      "loss": 0.092,
+      "step": 28890
+    },
+    {
+      "epoch": 2.7362241999621286,
+      "grad_norm": 0.3156600594520569,
+      "learning_rate": 9.055103200151487e-06,
+      "loss": 0.1007,
+      "step": 28900
+    },
+    {
+      "epoch": 2.7371709903427384,
+      "grad_norm": 0.48856040835380554,
+      "learning_rate": 9.051316038629047e-06,
+      "loss": 0.1107,
+      "step": 28910
+    },
+    {
+      "epoch": 2.738117780723348,
+      "grad_norm": 0.39420148730278015,
+      "learning_rate": 9.047528877106609e-06,
+      "loss": 0.097,
+      "step": 28920
+    },
+    {
+      "epoch": 2.739064571103958,
+      "grad_norm": 0.4240681827068329,
+      "learning_rate": 9.04374171558417e-06,
+      "loss": 0.1077,
+      "step": 28930
+    },
+    {
+      "epoch": 2.7400113614845676,
+      "grad_norm": 0.43309491872787476,
+      "learning_rate": 9.039954554061733e-06,
+      "loss": 0.1095,
+      "step": 28940
+    },
+    {
+      "epoch": 2.740958151865177,
+      "grad_norm": 0.48016998171806335,
+      "learning_rate": 9.036167392539293e-06,
+      "loss": 0.096,
+      "step": 28950
+    },
+    {
+      "epoch": 2.7419049422457866,
+      "grad_norm": 0.34501051902770996,
+      "learning_rate": 9.032380231016854e-06,
+      "loss": 0.1063,
+      "step": 28960
+    },
+    {
+      "epoch": 2.7428517326263964,
+      "grad_norm": 0.42231935262680054,
+      "learning_rate": 9.028593069494414e-06,
+      "loss": 0.1087,
+      "step": 28970
+    },
+    {
+      "epoch": 2.743798523007006,
+      "grad_norm": 0.3782620429992676,
+      "learning_rate": 9.024805907971974e-06,
+      "loss": 0.099,
+      "step": 28980
+    },
+    {
+      "epoch": 2.744745313387616,
+      "grad_norm": 0.4469917118549347,
+      "learning_rate": 9.021018746449536e-06,
+      "loss": 0.096,
+      "step": 28990
+    },
+    {
+      "epoch": 2.7456921037682256,
+      "grad_norm": 0.3554760217666626,
+      "learning_rate": 9.017231584927098e-06,
+      "loss": 0.0918,
+      "step": 29000
+    },
+    {
+      "epoch": 2.7466388941488353,
+      "grad_norm": 0.3817310929298401,
+      "learning_rate": 9.01344442340466e-06,
+      "loss": 0.1054,
+      "step": 29010
+    },
+    {
+      "epoch": 2.747585684529445,
+      "grad_norm": 0.37116482853889465,
+      "learning_rate": 9.00965726188222e-06,
+      "loss": 0.1005,
+      "step": 29020
+    },
+    {
+      "epoch": 2.748532474910055,
+      "grad_norm": 0.41311806440353394,
+      "learning_rate": 9.005870100359781e-06,
+      "loss": 0.1086,
+      "step": 29030
+    },
+    {
+      "epoch": 2.7494792652906646,
+      "grad_norm": 0.648277223110199,
+      "learning_rate": 9.002082938837342e-06,
+      "loss": 0.1081,
+      "step": 29040
+    },
+    {
+      "epoch": 2.7504260556712743,
+      "grad_norm": 0.39448094367980957,
+      "learning_rate": 8.998295777314903e-06,
+      "loss": 0.105,
+      "step": 29050
+    },
+    {
+      "epoch": 2.751372846051884,
+      "grad_norm": 0.4399425983428955,
+      "learning_rate": 8.994508615792465e-06,
+      "loss": 0.1081,
+      "step": 29060
+    },
+    {
+      "epoch": 2.7523196364324938,
+      "grad_norm": 0.4444071650505066,
+      "learning_rate": 8.990721454270025e-06,
+      "loss": 0.0978,
+      "step": 29070
+    },
+    {
+      "epoch": 2.7532664268131035,
+      "grad_norm": 0.5003328919410706,
+      "learning_rate": 8.986934292747587e-06,
+      "loss": 0.1057,
+      "step": 29080
+    },
+    {
+      "epoch": 2.7542132171937133,
+      "grad_norm": 0.42164021730422974,
+      "learning_rate": 8.983147131225147e-06,
+      "loss": 0.1144,
+      "step": 29090
+    },
+    {
+      "epoch": 2.755160007574323,
+      "grad_norm": 0.43083667755126953,
+      "learning_rate": 8.979359969702709e-06,
+      "loss": 0.0995,
+      "step": 29100
+    },
+    {
+      "epoch": 2.7561067979549327,
+      "grad_norm": 0.4107886552810669,
+      "learning_rate": 8.97557280818027e-06,
+      "loss": 0.104,
+      "step": 29110
+    },
+    {
+      "epoch": 2.7570535883355425,
+      "grad_norm": 0.3801655173301697,
+      "learning_rate": 8.97178564665783e-06,
+      "loss": 0.0987,
+      "step": 29120
+    },
+    {
+      "epoch": 2.7580003787161522,
+      "grad_norm": 0.38127654790878296,
+      "learning_rate": 8.967998485135392e-06,
+      "loss": 0.1015,
+      "step": 29130
+    },
+    {
+      "epoch": 2.758947169096762,
+      "grad_norm": 0.3987577259540558,
+      "learning_rate": 8.964211323612952e-06,
+      "loss": 0.11,
+      "step": 29140
+    },
+    {
+      "epoch": 2.7598939594773717,
+      "grad_norm": 0.4438229203224182,
+      "learning_rate": 8.960424162090514e-06,
+      "loss": 0.1074,
+      "step": 29150
+    },
+    {
+      "epoch": 2.7608407498579814,
+      "grad_norm": 0.4837617874145508,
+      "learning_rate": 8.956637000568074e-06,
+      "loss": 0.0993,
+      "step": 29160
+    },
+    {
+      "epoch": 2.761787540238591,
+      "grad_norm": 0.3521498739719391,
+      "learning_rate": 8.952849839045636e-06,
+      "loss": 0.1008,
+      "step": 29170
+    },
+    {
+      "epoch": 2.762734330619201,
+      "grad_norm": 0.40667611360549927,
+      "learning_rate": 8.949062677523197e-06,
+      "loss": 0.1001,
+      "step": 29180
+    },
+    {
+      "epoch": 2.7636811209998107,
+      "grad_norm": 0.420484334230423,
+      "learning_rate": 8.94527551600076e-06,
+      "loss": 0.1032,
+      "step": 29190
+    },
+    {
+      "epoch": 2.7646279113804204,
+      "grad_norm": 0.49111509323120117,
+      "learning_rate": 8.94148835447832e-06,
+      "loss": 0.1025,
+      "step": 29200
+    },
+    {
+      "epoch": 2.76557470176103,
+      "grad_norm": 0.4495706260204315,
+      "learning_rate": 8.93770119295588e-06,
+      "loss": 0.1005,
+      "step": 29210
+    },
+    {
+      "epoch": 2.76652149214164,
+      "grad_norm": 0.45926108956336975,
+      "learning_rate": 8.933914031433441e-06,
+      "loss": 0.0989,
+      "step": 29220
+    },
+    {
+      "epoch": 2.7674682825222496,
+      "grad_norm": 0.4548151195049286,
+      "learning_rate": 8.930126869911003e-06,
+      "loss": 0.0988,
+      "step": 29230
+    },
+    {
+      "epoch": 2.7684150729028594,
+      "grad_norm": 0.47408318519592285,
+      "learning_rate": 8.926339708388565e-06,
+      "loss": 0.0912,
+      "step": 29240
+    },
+    {
+      "epoch": 2.769361863283469,
+      "grad_norm": 0.41850709915161133,
+      "learning_rate": 8.922552546866125e-06,
+      "loss": 0.0955,
+      "step": 29250
+    },
+    {
+      "epoch": 2.770308653664079,
+      "grad_norm": 0.4139980673789978,
+      "learning_rate": 8.918765385343686e-06,
+      "loss": 0.0994,
+      "step": 29260
+    },
+    {
+      "epoch": 2.7712554440446886,
+      "grad_norm": 0.4438539743423462,
+      "learning_rate": 8.914978223821246e-06,
+      "loss": 0.1026,
+      "step": 29270
+    },
+    {
+      "epoch": 2.7722022344252983,
+      "grad_norm": 0.34790053963661194,
+      "learning_rate": 8.911191062298808e-06,
+      "loss": 0.103,
+      "step": 29280
+    },
+    {
+      "epoch": 2.773149024805908,
+      "grad_norm": 0.48705199360847473,
+      "learning_rate": 8.90740390077637e-06,
+      "loss": 0.1025,
+      "step": 29290
+    },
+    {
+      "epoch": 2.774095815186518,
+      "grad_norm": 0.46644020080566406,
+      "learning_rate": 8.90361673925393e-06,
+      "loss": 0.1061,
+      "step": 29300
+    },
+    {
+      "epoch": 2.7750426055671276,
+      "grad_norm": 0.28609853982925415,
+      "learning_rate": 8.899829577731492e-06,
+      "loss": 0.105,
+      "step": 29310
+    },
+    {
+      "epoch": 2.7759893959477373,
+      "grad_norm": 0.47828933596611023,
+      "learning_rate": 8.896042416209052e-06,
+      "loss": 0.1074,
+      "step": 29320
+    },
+    {
+      "epoch": 2.776936186328347,
+      "grad_norm": 0.530474841594696,
+      "learning_rate": 8.892255254686613e-06,
+      "loss": 0.0944,
+      "step": 29330
+    },
+    {
+      "epoch": 2.777882976708957,
+      "grad_norm": 0.46874600648880005,
+      "learning_rate": 8.888468093164174e-06,
+      "loss": 0.0999,
+      "step": 29340
+    },
+    {
+      "epoch": 2.7788297670895665,
+      "grad_norm": 0.4388161301612854,
+      "learning_rate": 8.884680931641735e-06,
+      "loss": 0.092,
+      "step": 29350
+    },
+    {
+      "epoch": 2.7797765574701763,
+      "grad_norm": 0.35012832283973694,
+      "learning_rate": 8.880893770119297e-06,
+      "loss": 0.1041,
+      "step": 29360
+    },
+    {
+      "epoch": 2.7807233478507856,
+      "grad_norm": 0.4594600796699524,
+      "learning_rate": 8.877106608596857e-06,
+      "loss": 0.096,
+      "step": 29370
+    },
+    {
+      "epoch": 2.7816701382313953,
+      "grad_norm": 0.3499717116355896,
+      "learning_rate": 8.873319447074419e-06,
+      "loss": 0.0911,
+      "step": 29380
+    },
+    {
+      "epoch": 2.782616928612005,
+      "grad_norm": 0.431905597448349,
+      "learning_rate": 8.869532285551979e-06,
+      "loss": 0.0965,
+      "step": 29390
+    },
+    {
+      "epoch": 2.783563718992615,
+      "grad_norm": 0.3726125955581665,
+      "learning_rate": 8.86574512402954e-06,
+      "loss": 0.0958,
+      "step": 29400
+    },
+    {
+      "epoch": 2.7845105093732245,
+      "grad_norm": 0.38136884570121765,
+      "learning_rate": 8.861957962507102e-06,
+      "loss": 0.1041,
+      "step": 29410
+    },
+    {
+      "epoch": 2.7854572997538343,
+      "grad_norm": 0.398968368768692,
+      "learning_rate": 8.858170800984662e-06,
+      "loss": 0.0988,
+      "step": 29420
+    },
+    {
+      "epoch": 2.786404090134444,
+      "grad_norm": 0.4601745307445526,
+      "learning_rate": 8.854383639462224e-06,
+      "loss": 0.1027,
+      "step": 29430
+    },
+    {
+      "epoch": 2.7873508805150538,
+      "grad_norm": 0.34036746621131897,
+      "learning_rate": 8.850596477939784e-06,
+      "loss": 0.1155,
+      "step": 29440
+    },
+    {
+      "epoch": 2.7882976708956635,
+      "grad_norm": 0.3742331266403198,
+      "learning_rate": 8.846809316417346e-06,
+      "loss": 0.1044,
+      "step": 29450
+    },
+    {
+      "epoch": 2.7892444612762732,
+      "grad_norm": 0.40490207076072693,
+      "learning_rate": 8.843022154894908e-06,
+      "loss": 0.1057,
+      "step": 29460
+    },
+    {
+      "epoch": 2.790191251656883,
+      "grad_norm": 0.37119242548942566,
+      "learning_rate": 8.83923499337247e-06,
+      "loss": 0.099,
+      "step": 29470
+    },
+    {
+      "epoch": 2.7911380420374927,
+      "grad_norm": 0.36027467250823975,
+      "learning_rate": 8.83544783185003e-06,
+      "loss": 0.0929,
+      "step": 29480
+    },
+    {
+      "epoch": 2.7920848324181025,
+      "grad_norm": 0.4557902216911316,
+      "learning_rate": 8.83166067032759e-06,
+      "loss": 0.0963,
+      "step": 29490
+    },
+    {
+      "epoch": 2.793031622798712,
+      "grad_norm": 0.4901174306869507,
+      "learning_rate": 8.827873508805151e-06,
+      "loss": 0.1129,
+      "step": 29500
+    },
+    {
+      "epoch": 2.793978413179322,
+      "grad_norm": 0.5360562205314636,
+      "learning_rate": 8.824086347282711e-06,
+      "loss": 0.1027,
+      "step": 29510
+    },
+    {
+      "epoch": 2.7949252035599317,
+      "grad_norm": 0.36238235235214233,
+      "learning_rate": 8.820299185760273e-06,
+      "loss": 0.0943,
+      "step": 29520
+    },
+    {
+      "epoch": 2.7958719939405414,
+      "grad_norm": 0.39753592014312744,
+      "learning_rate": 8.816512024237835e-06,
+      "loss": 0.1038,
+      "step": 29530
+    },
+    {
+      "epoch": 2.796818784321151,
+      "grad_norm": 0.4467945098876953,
+      "learning_rate": 8.812724862715397e-06,
+      "loss": 0.098,
+      "step": 29540
+    },
+    {
+      "epoch": 2.797765574701761,
+      "grad_norm": 0.3594754636287689,
+      "learning_rate": 8.808937701192957e-06,
+      "loss": 0.1085,
+      "step": 29550
+    },
+    {
+      "epoch": 2.7987123650823706,
+      "grad_norm": 0.4449863135814667,
+      "learning_rate": 8.805150539670517e-06,
+      "loss": 0.1032,
+      "step": 29560
+    },
+    {
+      "epoch": 2.7996591554629804,
+      "grad_norm": 0.45748332142829895,
+      "learning_rate": 8.801363378148078e-06,
+      "loss": 0.0975,
+      "step": 29570
+    },
+    {
+      "epoch": 2.80060594584359,
+      "grad_norm": 0.42559105157852173,
+      "learning_rate": 8.79757621662564e-06,
+      "loss": 0.101,
+      "step": 29580
+    },
+    {
+      "epoch": 2.8015527362242,
+      "grad_norm": 0.3052559494972229,
+      "learning_rate": 8.793789055103202e-06,
+      "loss": 0.0929,
+      "step": 29590
+    },
+    {
+      "epoch": 2.8024995266048096,
+      "grad_norm": 0.30140575766563416,
+      "learning_rate": 8.790001893580762e-06,
+      "loss": 0.0913,
+      "step": 29600
+    },
+    {
+      "epoch": 2.8034463169854194,
+      "grad_norm": 0.5412850975990295,
+      "learning_rate": 8.786214732058324e-06,
+      "loss": 0.1077,
+      "step": 29610
+    },
+    {
+      "epoch": 2.804393107366029,
+      "grad_norm": 0.3537708520889282,
+      "learning_rate": 8.782427570535884e-06,
+      "loss": 0.0994,
+      "step": 29620
+    },
+    {
+      "epoch": 2.805339897746639,
+      "grad_norm": 0.38238513469696045,
+      "learning_rate": 8.778640409013445e-06,
+      "loss": 0.1039,
+      "step": 29630
+    },
+    {
+      "epoch": 2.8062866881272486,
+      "grad_norm": 0.40344172716140747,
+      "learning_rate": 8.774853247491007e-06,
+      "loss": 0.0976,
+      "step": 29640
+    },
+    {
+      "epoch": 2.8072334785078583,
+      "grad_norm": 0.33826422691345215,
+      "learning_rate": 8.771066085968567e-06,
+      "loss": 0.0959,
+      "step": 29650
+    },
+    {
+      "epoch": 2.808180268888468,
+      "grad_norm": 0.42276862263679504,
+      "learning_rate": 8.767278924446129e-06,
+      "loss": 0.1051,
+      "step": 29660
+    },
+    {
+      "epoch": 2.809127059269078,
+      "grad_norm": 0.48476433753967285,
+      "learning_rate": 8.763491762923689e-06,
+      "loss": 0.1041,
+      "step": 29670
+    },
+    {
+      "epoch": 2.8100738496496875,
+      "grad_norm": 0.37548232078552246,
+      "learning_rate": 8.75970460140125e-06,
+      "loss": 0.0978,
+      "step": 29680
+    },
+    {
+      "epoch": 2.8110206400302973,
+      "grad_norm": 0.4841581881046295,
+      "learning_rate": 8.75591743987881e-06,
+      "loss": 0.0984,
+      "step": 29690
+    },
+    {
+      "epoch": 2.811967430410907,
+      "grad_norm": 0.8658079504966736,
+      "learning_rate": 8.752130278356373e-06,
+      "loss": 0.1072,
+      "step": 29700
+    },
+    {
+      "epoch": 2.8129142207915168,
+      "grad_norm": 0.4591039717197418,
+      "learning_rate": 8.748343116833934e-06,
+      "loss": 0.1086,
+      "step": 29710
+    },
+    {
+      "epoch": 2.8138610111721265,
+      "grad_norm": 0.4760480225086212,
+      "learning_rate": 8.744555955311494e-06,
+      "loss": 0.097,
+      "step": 29720
+    },
+    {
+      "epoch": 2.8148078015527362,
+      "grad_norm": 0.3787853717803955,
+      "learning_rate": 8.740768793789056e-06,
+      "loss": 0.1036,
+      "step": 29730
+    },
+    {
+      "epoch": 2.815754591933346,
+      "grad_norm": 0.46122756600379944,
+      "learning_rate": 8.736981632266616e-06,
+      "loss": 0.0953,
+      "step": 29740
+    },
+    {
+      "epoch": 2.8167013823139557,
+      "grad_norm": 0.37929031252861023,
+      "learning_rate": 8.733194470744178e-06,
+      "loss": 0.0986,
+      "step": 29750
+    },
+    {
+      "epoch": 2.8176481726945655,
+      "grad_norm": 0.3670518696308136,
+      "learning_rate": 8.72940730922174e-06,
+      "loss": 0.1008,
+      "step": 29760
+    },
+    {
+      "epoch": 2.818594963075175,
+      "grad_norm": 0.35238727927207947,
+      "learning_rate": 8.7256201476993e-06,
+      "loss": 0.1015,
+      "step": 29770
+    },
+    {
+      "epoch": 2.819541753455785,
+      "grad_norm": 0.42419156432151794,
+      "learning_rate": 8.721832986176861e-06,
+      "loss": 0.094,
+      "step": 29780
+    },
+    {
+      "epoch": 2.8204885438363947,
+      "grad_norm": 0.3655274510383606,
+      "learning_rate": 8.718045824654421e-06,
+      "loss": 0.1012,
+      "step": 29790
+    },
+    {
+      "epoch": 2.8214353342170044,
+      "grad_norm": 0.42500266432762146,
+      "learning_rate": 8.714258663131983e-06,
+      "loss": 0.107,
+      "step": 29800
+    },
+    {
+      "epoch": 2.822382124597614,
+      "grad_norm": 0.43933090567588806,
+      "learning_rate": 8.710471501609545e-06,
+      "loss": 0.1035,
+      "step": 29810
+    },
+    {
+      "epoch": 2.823328914978224,
+      "grad_norm": 0.4164799153804779,
+      "learning_rate": 8.706684340087107e-06,
+      "loss": 0.1074,
+      "step": 29820
+    },
+    {
+      "epoch": 2.8242757053588337,
+      "grad_norm": 0.46152105927467346,
+      "learning_rate": 8.702897178564667e-06,
+      "loss": 0.1,
+      "step": 29830
+    },
+    {
+      "epoch": 2.8252224957394434,
+      "grad_norm": 0.4320620000362396,
+      "learning_rate": 8.699110017042227e-06,
+      "loss": 0.1056,
+      "step": 29840
+    },
+    {
+      "epoch": 2.826169286120053,
+      "grad_norm": 0.42175158858299255,
+      "learning_rate": 8.695322855519789e-06,
+      "loss": 0.1018,
+      "step": 29850
+    },
+    {
+      "epoch": 2.827116076500663,
+      "grad_norm": 0.43998804688453674,
+      "learning_rate": 8.691535693997349e-06,
+      "loss": 0.0971,
+      "step": 29860
+    },
+    {
+      "epoch": 2.8280628668812726,
+      "grad_norm": 0.4695793688297272,
+      "learning_rate": 8.68774853247491e-06,
+      "loss": 0.1024,
+      "step": 29870
+    },
+    {
+      "epoch": 2.8290096572618824,
+      "grad_norm": 0.3865763545036316,
+      "learning_rate": 8.683961370952472e-06,
+      "loss": 0.1038,
+      "step": 29880
+    },
+    {
+      "epoch": 2.829956447642492,
+      "grad_norm": 0.3920170068740845,
+      "learning_rate": 8.680174209430034e-06,
+      "loss": 0.1063,
+      "step": 29890
+    },
+    {
+      "epoch": 2.830903238023102,
+      "grad_norm": 0.3742087781429291,
+      "learning_rate": 8.676387047907594e-06,
+      "loss": 0.1004,
+      "step": 29900
+    },
+    {
+      "epoch": 2.8318500284037116,
+      "grad_norm": 0.40059271454811096,
+      "learning_rate": 8.672599886385154e-06,
+      "loss": 0.0996,
+      "step": 29910
+    },
+    {
+      "epoch": 2.8327968187843213,
+      "grad_norm": 0.3583625257015228,
+      "learning_rate": 8.668812724862716e-06,
+      "loss": 0.0944,
+      "step": 29920
+    },
+    {
+      "epoch": 2.833743609164931,
+      "grad_norm": 0.4899672865867615,
+      "learning_rate": 8.665025563340277e-06,
+      "loss": 0.1033,
+      "step": 29930
+    },
+    {
+      "epoch": 2.834690399545541,
+      "grad_norm": 0.3681374788284302,
+      "learning_rate": 8.661238401817839e-06,
+      "loss": 0.0964,
+      "step": 29940
+    },
+    {
+      "epoch": 2.8356371899261505,
+      "grad_norm": 0.4126521050930023,
+      "learning_rate": 8.6574512402954e-06,
+      "loss": 0.086,
+      "step": 29950
+    },
+    {
+      "epoch": 2.8365839803067603,
+      "grad_norm": 0.5693396925926208,
+      "learning_rate": 8.653664078772961e-06,
+      "loss": 0.11,
+      "step": 29960
+    },
+    {
+      "epoch": 2.83753077068737,
+      "grad_norm": 0.5090595483779907,
+      "learning_rate": 8.649876917250521e-06,
+      "loss": 0.1099,
+      "step": 29970
+    },
+    {
+      "epoch": 2.8384775610679798,
+      "grad_norm": 0.49548888206481934,
+      "learning_rate": 8.646089755728083e-06,
+      "loss": 0.1036,
+      "step": 29980
+    },
+    {
+      "epoch": 2.8394243514485895,
+      "grad_norm": 0.3540317416191101,
+      "learning_rate": 8.642302594205644e-06,
+      "loss": 0.101,
+      "step": 29990
+    },
+    {
+      "epoch": 2.8403711418291993,
+      "grad_norm": 0.3879123032093048,
+      "learning_rate": 8.638515432683205e-06,
+      "loss": 0.094,
+      "step": 30000
+    },
+    {
+      "epoch": 2.841317932209809,
+      "grad_norm": 0.39417970180511475,
+      "learning_rate": 8.634728271160766e-06,
+      "loss": 0.0965,
+      "step": 30010
+    },
+    {
+      "epoch": 2.8422647225904187,
+      "grad_norm": 0.38096165657043457,
+      "learning_rate": 8.630941109638326e-06,
+      "loss": 0.1065,
+      "step": 30020
+    },
+    {
+      "epoch": 2.8432115129710285,
+      "grad_norm": 0.3729986548423767,
+      "learning_rate": 8.627153948115888e-06,
+      "loss": 0.1002,
+      "step": 30030
+    },
+    {
+      "epoch": 2.844158303351638,
+      "grad_norm": 0.37190115451812744,
+      "learning_rate": 8.623366786593448e-06,
+      "loss": 0.1023,
+      "step": 30040
+    },
+    {
+      "epoch": 2.8451050937322475,
+      "grad_norm": 0.2926923334598541,
+      "learning_rate": 8.61957962507101e-06,
+      "loss": 0.1041,
+      "step": 30050
+    },
+    {
+      "epoch": 2.8460518841128573,
+      "grad_norm": 0.4673272371292114,
+      "learning_rate": 8.615792463548572e-06,
+      "loss": 0.1016,
+      "step": 30060
+    },
+    {
+      "epoch": 2.846998674493467,
+      "grad_norm": 0.4333568215370178,
+      "learning_rate": 8.612005302026132e-06,
+      "loss": 0.0999,
+      "step": 30070
+    },
+    {
+      "epoch": 2.8479454648740767,
+      "grad_norm": 0.385219931602478,
+      "learning_rate": 8.608218140503693e-06,
+      "loss": 0.0973,
+      "step": 30080
+    },
+    {
+      "epoch": 2.8488922552546865,
+      "grad_norm": 0.3480224311351776,
+      "learning_rate": 8.604430978981253e-06,
+      "loss": 0.1044,
+      "step": 30090
+    },
+    {
+      "epoch": 2.849839045635296,
+      "grad_norm": 0.3795575499534607,
+      "learning_rate": 8.600643817458815e-06,
+      "loss": 0.0946,
+      "step": 30100
+    },
+    {
+      "epoch": 2.850785836015906,
+      "grad_norm": 0.3406654894351959,
+      "learning_rate": 8.596856655936377e-06,
+      "loss": 0.1109,
+      "step": 30110
+    },
+    {
+      "epoch": 2.8517326263965157,
+      "grad_norm": 0.4390551745891571,
+      "learning_rate": 8.593069494413937e-06,
+      "loss": 0.105,
+      "step": 30120
+    },
+    {
+      "epoch": 2.8526794167771254,
+      "grad_norm": 0.3838212490081787,
+      "learning_rate": 8.589282332891499e-06,
+      "loss": 0.1058,
+      "step": 30130
+    },
+    {
+      "epoch": 2.853626207157735,
+      "grad_norm": 0.5091719627380371,
+      "learning_rate": 8.585495171369059e-06,
+      "loss": 0.1062,
+      "step": 30140
+    },
+    {
+      "epoch": 2.854572997538345,
+      "grad_norm": 0.4112211763858795,
+      "learning_rate": 8.58170800984662e-06,
+      "loss": 0.1028,
+      "step": 30150
+    },
+    {
+      "epoch": 2.8555197879189547,
+      "grad_norm": 0.3643788695335388,
+      "learning_rate": 8.577920848324182e-06,
+      "loss": 0.0972,
+      "step": 30160
+    },
+    {
+      "epoch": 2.8564665782995644,
+      "grad_norm": 0.38067877292633057,
+      "learning_rate": 8.574133686801744e-06,
+      "loss": 0.1092,
+      "step": 30170
+    },
+    {
+      "epoch": 2.857413368680174,
+      "grad_norm": 0.38998544216156006,
+      "learning_rate": 8.570346525279304e-06,
+      "loss": 0.1089,
+      "step": 30180
+    },
+    {
+      "epoch": 2.858360159060784,
+      "grad_norm": 0.4681199789047241,
+      "learning_rate": 8.566559363756864e-06,
+      "loss": 0.0997,
+      "step": 30190
+    },
+    {
+      "epoch": 2.8593069494413936,
+      "grad_norm": 0.4034975469112396,
+      "learning_rate": 8.562772202234426e-06,
+      "loss": 0.1003,
+      "step": 30200
+    },
+    {
+      "epoch": 2.8602537398220034,
+      "grad_norm": 0.42309436202049255,
+      "learning_rate": 8.558985040711988e-06,
+      "loss": 0.1071,
+      "step": 30210
+    },
+    {
+      "epoch": 2.861200530202613,
+      "grad_norm": 0.42179518938064575,
+      "learning_rate": 8.555197879189548e-06,
+      "loss": 0.1047,
+      "step": 30220
+    },
+    {
+      "epoch": 2.862147320583223,
+      "grad_norm": 0.4579227566719055,
+      "learning_rate": 8.55141071766711e-06,
+      "loss": 0.1007,
+      "step": 30230
+    },
+    {
+      "epoch": 2.8630941109638326,
+      "grad_norm": 0.4164641499519348,
+      "learning_rate": 8.547623556144671e-06,
+      "loss": 0.1086,
+      "step": 30240
+    },
+    {
+      "epoch": 2.8640409013444423,
+      "grad_norm": 0.3335796296596527,
+      "learning_rate": 8.543836394622231e-06,
+      "loss": 0.0975,
+      "step": 30250
+    },
+    {
+      "epoch": 2.864987691725052,
+      "grad_norm": 0.4051174819469452,
+      "learning_rate": 8.540049233099791e-06,
+      "loss": 0.112,
+      "step": 30260
+    },
+    {
+      "epoch": 2.865934482105662,
+      "grad_norm": 0.3602439761161804,
+      "learning_rate": 8.536262071577353e-06,
+      "loss": 0.0976,
+      "step": 30270
+    },
+    {
+      "epoch": 2.8668812724862716,
+      "grad_norm": 0.48511287569999695,
+      "learning_rate": 8.532474910054915e-06,
+      "loss": 0.0992,
+      "step": 30280
+    },
+    {
+      "epoch": 2.8678280628668813,
+      "grad_norm": 0.5073041915893555,
+      "learning_rate": 8.528687748532476e-06,
+      "loss": 0.0991,
+      "step": 30290
+    },
+    {
+      "epoch": 2.868774853247491,
+      "grad_norm": 1.1230666637420654,
+      "learning_rate": 8.524900587010037e-06,
+      "loss": 0.1025,
+      "step": 30300
+    },
+    {
+      "epoch": 2.869721643628101,
+      "grad_norm": 0.4232262670993805,
+      "learning_rate": 8.521113425487598e-06,
+      "loss": 0.1106,
+      "step": 30310
+    },
+    {
+      "epoch": 2.8706684340087105,
+      "grad_norm": 0.4858565032482147,
+      "learning_rate": 8.517326263965158e-06,
+      "loss": 0.1069,
+      "step": 30320
+    },
+    {
+      "epoch": 2.8716152243893203,
+      "grad_norm": 0.4056932330131531,
+      "learning_rate": 8.51353910244272e-06,
+      "loss": 0.1081,
+      "step": 30330
+    },
+    {
+      "epoch": 2.87256201476993,
+      "grad_norm": 0.3665787875652313,
+      "learning_rate": 8.509751940920282e-06,
+      "loss": 0.0968,
+      "step": 30340
+    },
+    {
+      "epoch": 2.8735088051505397,
+      "grad_norm": 0.5439797639846802,
+      "learning_rate": 8.505964779397842e-06,
+      "loss": 0.0961,
+      "step": 30350
+    },
+    {
+      "epoch": 2.8744555955311495,
+      "grad_norm": 0.37947776913642883,
+      "learning_rate": 8.502177617875404e-06,
+      "loss": 0.11,
+      "step": 30360
+    },
+    {
+      "epoch": 2.8754023859117592,
+      "grad_norm": 0.5098512768745422,
+      "learning_rate": 8.498390456352964e-06,
+      "loss": 0.1019,
+      "step": 30370
+    },
+    {
+      "epoch": 2.876349176292369,
+      "grad_norm": 0.43195387721061707,
+      "learning_rate": 8.494603294830525e-06,
+      "loss": 0.1083,
+      "step": 30380
+    },
+    {
+      "epoch": 2.8772959666729787,
+      "grad_norm": 0.534127414226532,
+      "learning_rate": 8.490816133308087e-06,
+      "loss": 0.1054,
+      "step": 30390
+    },
+    {
+      "epoch": 2.8782427570535885,
+      "grad_norm": 0.3190009295940399,
+      "learning_rate": 8.487028971785647e-06,
+      "loss": 0.1028,
+      "step": 30400
+    },
+    {
+      "epoch": 2.879189547434198,
+      "grad_norm": 0.31885769963264465,
+      "learning_rate": 8.483241810263209e-06,
+      "loss": 0.0895,
+      "step": 30410
+    },
+    {
+      "epoch": 2.880136337814808,
+      "grad_norm": 0.37694260478019714,
+      "learning_rate": 8.479454648740769e-06,
+      "loss": 0.1054,
+      "step": 30420
+    },
+    {
+      "epoch": 2.8810831281954177,
+      "grad_norm": 0.38295096158981323,
+      "learning_rate": 8.47566748721833e-06,
+      "loss": 0.1009,
+      "step": 30430
+    },
+    {
+      "epoch": 2.8820299185760274,
+      "grad_norm": 0.4697856903076172,
+      "learning_rate": 8.47188032569589e-06,
+      "loss": 0.1041,
+      "step": 30440
+    },
+    {
+      "epoch": 2.882976708956637,
+      "grad_norm": 0.4540785551071167,
+      "learning_rate": 8.468093164173452e-06,
+      "loss": 0.1118,
+      "step": 30450
+    },
+    {
+      "epoch": 2.883923499337247,
+      "grad_norm": 0.3804856538772583,
+      "learning_rate": 8.464306002651014e-06,
+      "loss": 0.0938,
+      "step": 30460
+    },
+    {
+      "epoch": 2.884870289717856,
+      "grad_norm": 0.46269491314888,
+      "learning_rate": 8.460518841128574e-06,
+      "loss": 0.104,
+      "step": 30470
+    },
+    {
+      "epoch": 2.885817080098466,
+      "grad_norm": 0.4231495261192322,
+      "learning_rate": 8.456731679606136e-06,
+      "loss": 0.104,
+      "step": 30480
+    },
+    {
+      "epoch": 2.8867638704790757,
+      "grad_norm": 0.37176263332366943,
+      "learning_rate": 8.452944518083696e-06,
+      "loss": 0.1049,
+      "step": 30490
+    },
+    {
+      "epoch": 2.8877106608596854,
+      "grad_norm": 0.39734789729118347,
+      "learning_rate": 8.449157356561258e-06,
+      "loss": 0.1014,
+      "step": 30500
+    },
+    {
+      "epoch": 2.888657451240295,
+      "grad_norm": 0.49146899580955505,
+      "learning_rate": 8.44537019503882e-06,
+      "loss": 0.1101,
+      "step": 30510
+    },
+    {
+      "epoch": 2.889604241620905,
+      "grad_norm": 0.4609476625919342,
+      "learning_rate": 8.441583033516381e-06,
+      "loss": 0.0973,
+      "step": 30520
+    },
+    {
+      "epoch": 2.8905510320015146,
+      "grad_norm": 0.44791117310523987,
+      "learning_rate": 8.437795871993941e-06,
+      "loss": 0.1032,
+      "step": 30530
+    },
+    {
+      "epoch": 2.8914978223821244,
+      "grad_norm": 0.3995889127254486,
+      "learning_rate": 8.434008710471501e-06,
+      "loss": 0.0989,
+      "step": 30540
+    },
+    {
+      "epoch": 2.892444612762734,
+      "grad_norm": 0.4357839822769165,
+      "learning_rate": 8.430221548949063e-06,
+      "loss": 0.0979,
+      "step": 30550
+    },
+    {
+      "epoch": 2.893391403143344,
+      "grad_norm": 0.5497676730155945,
+      "learning_rate": 8.426434387426625e-06,
+      "loss": 0.1051,
+      "step": 30560
+    },
+    {
+      "epoch": 2.8943381935239536,
+      "grad_norm": 0.43459513783454895,
+      "learning_rate": 8.422647225904187e-06,
+      "loss": 0.1036,
+      "step": 30570
+    },
+    {
+      "epoch": 2.8952849839045633,
+      "grad_norm": 0.3618446886539459,
+      "learning_rate": 8.418860064381747e-06,
+      "loss": 0.0987,
+      "step": 30580
+    },
+    {
+      "epoch": 2.896231774285173,
+      "grad_norm": 0.43975988030433655,
+      "learning_rate": 8.415072902859308e-06,
+      "loss": 0.1078,
+      "step": 30590
+    },
+    {
+      "epoch": 2.897178564665783,
+      "grad_norm": 0.460006445646286,
+      "learning_rate": 8.411285741336868e-06,
+      "loss": 0.0949,
+      "step": 30600
+    },
+    {
+      "epoch": 2.8981253550463926,
+      "grad_norm": 0.6009049415588379,
+      "learning_rate": 8.407498579814429e-06,
+      "loss": 0.1095,
+      "step": 30610
+    },
+    {
+      "epoch": 2.8990721454270023,
+      "grad_norm": 0.4352884888648987,
+      "learning_rate": 8.40371141829199e-06,
+      "loss": 0.1125,
+      "step": 30620
+    },
+    {
+      "epoch": 2.900018935807612,
+      "grad_norm": 0.46387022733688354,
+      "learning_rate": 8.399924256769552e-06,
+      "loss": 0.1033,
+      "step": 30630
+    },
+    {
+      "epoch": 2.900965726188222,
+      "grad_norm": 0.3929131031036377,
+      "learning_rate": 8.396137095247114e-06,
+      "loss": 0.1045,
+      "step": 30640
+    },
+    {
+      "epoch": 2.9019125165688315,
+      "grad_norm": 0.29404574632644653,
+      "learning_rate": 8.392349933724674e-06,
+      "loss": 0.0872,
+      "step": 30650
+    },
+    {
+      "epoch": 2.9028593069494413,
+      "grad_norm": 0.5976027250289917,
+      "learning_rate": 8.388562772202236e-06,
+      "loss": 0.1156,
+      "step": 30660
+    },
+    {
+      "epoch": 2.903806097330051,
+      "grad_norm": 0.2850901782512665,
+      "learning_rate": 8.384775610679796e-06,
+      "loss": 0.0923,
+      "step": 30670
+    },
+    {
+      "epoch": 2.9047528877106608,
+      "grad_norm": 0.4848046898841858,
+      "learning_rate": 8.380988449157357e-06,
+      "loss": 0.1097,
+      "step": 30680
+    },
+    {
+      "epoch": 2.9056996780912705,
+      "grad_norm": 0.4893626272678375,
+      "learning_rate": 8.377201287634919e-06,
+      "loss": 0.1069,
+      "step": 30690
+    },
+    {
+      "epoch": 2.9066464684718802,
+      "grad_norm": 0.44268444180488586,
+      "learning_rate": 8.373414126112479e-06,
+      "loss": 0.0992,
+      "step": 30700
+    },
+    {
+      "epoch": 2.90759325885249,
+      "grad_norm": 0.44177380204200745,
+      "learning_rate": 8.369626964590041e-06,
+      "loss": 0.1033,
+      "step": 30710
+    },
+    {
+      "epoch": 2.9085400492330997,
+      "grad_norm": 0.37980490922927856,
+      "learning_rate": 8.365839803067601e-06,
+      "loss": 0.1114,
+      "step": 30720
+    },
+    {
+      "epoch": 2.9094868396137095,
+      "grad_norm": 0.32962408661842346,
+      "learning_rate": 8.362052641545163e-06,
+      "loss": 0.1095,
+      "step": 30730
+    },
+    {
+      "epoch": 2.910433629994319,
+      "grad_norm": 0.39001163840293884,
+      "learning_rate": 8.358265480022724e-06,
+      "loss": 0.1004,
+      "step": 30740
+    },
+    {
+      "epoch": 2.911380420374929,
+      "grad_norm": 0.4204920828342438,
+      "learning_rate": 8.354478318500284e-06,
+      "loss": 0.0961,
+      "step": 30750
+    },
+    {
+      "epoch": 2.9123272107555387,
+      "grad_norm": 0.397558331489563,
+      "learning_rate": 8.350691156977846e-06,
+      "loss": 0.0963,
+      "step": 30760
+    },
+    {
+      "epoch": 2.9132740011361484,
+      "grad_norm": 0.4277651906013489,
+      "learning_rate": 8.346903995455406e-06,
+      "loss": 0.1019,
+      "step": 30770
+    },
+    {
+      "epoch": 2.914220791516758,
+      "grad_norm": 0.4084283411502838,
+      "learning_rate": 8.343116833932968e-06,
+      "loss": 0.1075,
+      "step": 30780
+    },
+    {
+      "epoch": 2.915167581897368,
+      "grad_norm": 0.41574668884277344,
+      "learning_rate": 8.339329672410528e-06,
+      "loss": 0.0916,
+      "step": 30790
+    },
+    {
+      "epoch": 2.9161143722779777,
+      "grad_norm": 0.419190376996994,
+      "learning_rate": 8.33554251088809e-06,
+      "loss": 0.0954,
+      "step": 30800
+    },
+    {
+      "epoch": 2.9170611626585874,
+      "grad_norm": 0.5709701180458069,
+      "learning_rate": 8.331755349365652e-06,
+      "loss": 0.1003,
+      "step": 30810
+    },
+    {
+      "epoch": 2.918007953039197,
+      "grad_norm": 0.2899015545845032,
+      "learning_rate": 8.327968187843212e-06,
+      "loss": 0.0999,
+      "step": 30820
+    },
+    {
+      "epoch": 2.918954743419807,
+      "grad_norm": 0.40506407618522644,
+      "learning_rate": 8.324181026320773e-06,
+      "loss": 0.1095,
+      "step": 30830
+    },
+    {
+      "epoch": 2.9199015338004166,
+      "grad_norm": 0.4087386429309845,
+      "learning_rate": 8.320393864798333e-06,
+      "loss": 0.1019,
+      "step": 30840
+    },
+    {
+      "epoch": 2.9208483241810264,
+      "grad_norm": 0.45319652557373047,
+      "learning_rate": 8.316606703275895e-06,
+      "loss": 0.1031,
+      "step": 30850
+    },
+    {
+      "epoch": 2.921795114561636,
+      "grad_norm": 0.44078201055526733,
+      "learning_rate": 8.312819541753457e-06,
+      "loss": 0.1081,
+      "step": 30860
+    },
+    {
+      "epoch": 2.922741904942246,
+      "grad_norm": 0.34192386269569397,
+      "learning_rate": 8.309032380231019e-06,
+      "loss": 0.0939,
+      "step": 30870
+    },
+    {
+      "epoch": 2.9236886953228556,
+      "grad_norm": 0.38007766008377075,
+      "learning_rate": 8.305245218708579e-06,
+      "loss": 0.09,
+      "step": 30880
+    },
+    {
+      "epoch": 2.9246354857034653,
+      "grad_norm": 0.45848166942596436,
+      "learning_rate": 8.301458057186139e-06,
+      "loss": 0.0951,
+      "step": 30890
+    },
+    {
+      "epoch": 2.925582276084075,
+      "grad_norm": 0.3923320472240448,
+      "learning_rate": 8.2976708956637e-06,
+      "loss": 0.098,
+      "step": 30900
+    },
+    {
+      "epoch": 2.926529066464685,
+      "grad_norm": 0.4490647614002228,
+      "learning_rate": 8.293883734141262e-06,
+      "loss": 0.1143,
+      "step": 30910
+    },
+    {
+      "epoch": 2.9274758568452945,
+      "grad_norm": 0.4360518455505371,
+      "learning_rate": 8.290096572618824e-06,
+      "loss": 0.109,
+      "step": 30920
+    },
+    {
+      "epoch": 2.9284226472259043,
+      "grad_norm": 0.41733619570732117,
+      "learning_rate": 8.286309411096384e-06,
+      "loss": 0.1024,
+      "step": 30930
+    },
+    {
+      "epoch": 2.929369437606514,
+      "grad_norm": 0.4506927728652954,
+      "learning_rate": 8.282522249573946e-06,
+      "loss": 0.0936,
+      "step": 30940
+    },
+    {
+      "epoch": 2.9303162279871238,
+      "grad_norm": 0.46986424922943115,
+      "learning_rate": 8.278735088051506e-06,
+      "loss": 0.1047,
+      "step": 30950
+    },
+    {
+      "epoch": 2.9312630183677335,
+      "grad_norm": 0.4629628360271454,
+      "learning_rate": 8.274947926529066e-06,
+      "loss": 0.1027,
+      "step": 30960
+    },
+    {
+      "epoch": 2.9322098087483432,
+      "grad_norm": 0.48981261253356934,
+      "learning_rate": 8.271160765006628e-06,
+      "loss": 0.1122,
+      "step": 30970
+    },
+    {
+      "epoch": 2.933156599128953,
+      "grad_norm": 0.4340340793132782,
+      "learning_rate": 8.26737360348419e-06,
+      "loss": 0.1002,
+      "step": 30980
+    },
+    {
+      "epoch": 2.9341033895095627,
+      "grad_norm": 0.36658594012260437,
+      "learning_rate": 8.263586441961751e-06,
+      "loss": 0.1019,
+      "step": 30990
+    },
+    {
+      "epoch": 2.9350501798901725,
+      "grad_norm": 0.36050164699554443,
+      "learning_rate": 8.259799280439311e-06,
+      "loss": 0.1025,
+      "step": 31000
+    },
+    {
+      "epoch": 2.935996970270782,
+      "grad_norm": 0.4739544093608856,
+      "learning_rate": 8.256012118916873e-06,
+      "loss": 0.106,
+      "step": 31010
+    },
+    {
+      "epoch": 2.936943760651392,
+      "grad_norm": 0.4080391228199005,
+      "learning_rate": 8.252224957394433e-06,
+      "loss": 0.1066,
+      "step": 31020
+    },
+    {
+      "epoch": 2.9378905510320017,
+      "grad_norm": 0.46320927143096924,
+      "learning_rate": 8.248437795871995e-06,
+      "loss": 0.1067,
+      "step": 31030
+    },
+    {
+      "epoch": 2.9388373414126114,
+      "grad_norm": 0.4874909222126007,
+      "learning_rate": 8.244650634349556e-06,
+      "loss": 0.1084,
+      "step": 31040
+    },
+    {
+      "epoch": 2.939784131793221,
+      "grad_norm": 0.4553222060203552,
+      "learning_rate": 8.240863472827116e-06,
+      "loss": 0.1105,
+      "step": 31050
+    },
+    {
+      "epoch": 2.940730922173831,
+      "grad_norm": 0.37739482522010803,
+      "learning_rate": 8.237076311304678e-06,
+      "loss": 0.1,
+      "step": 31060
+    },
+    {
+      "epoch": 2.9416777125544407,
+      "grad_norm": 0.4049690067768097,
+      "learning_rate": 8.233289149782238e-06,
+      "loss": 0.1007,
+      "step": 31070
+    },
+    {
+      "epoch": 2.9426245029350504,
+      "grad_norm": 0.4319703280925751,
+      "learning_rate": 8.2295019882598e-06,
+      "loss": 0.1003,
+      "step": 31080
+    },
+    {
+      "epoch": 2.94357129331566,
+      "grad_norm": 0.410161554813385,
+      "learning_rate": 8.225714826737362e-06,
+      "loss": 0.0976,
+      "step": 31090
+    },
+    {
+      "epoch": 2.94451808369627,
+      "grad_norm": 0.4579450190067291,
+      "learning_rate": 8.221927665214922e-06,
+      "loss": 0.1117,
+      "step": 31100
+    },
+    {
+      "epoch": 2.9454648740768796,
+      "grad_norm": 0.4039101302623749,
+      "learning_rate": 8.218140503692484e-06,
+      "loss": 0.0941,
+      "step": 31110
+    },
+    {
+      "epoch": 2.9464116644574894,
+      "grad_norm": 0.3221927583217621,
+      "learning_rate": 8.214353342170044e-06,
+      "loss": 0.0926,
+      "step": 31120
+    },
+    {
+      "epoch": 2.947358454838099,
+      "grad_norm": 0.44777631759643555,
+      "learning_rate": 8.210566180647605e-06,
+      "loss": 0.091,
+      "step": 31130
+    },
+    {
+      "epoch": 2.948305245218709,
+      "grad_norm": 0.4191173315048218,
+      "learning_rate": 8.206779019125165e-06,
+      "loss": 0.1019,
+      "step": 31140
+    },
+    {
+      "epoch": 2.949252035599318,
+      "grad_norm": 0.3635985255241394,
+      "learning_rate": 8.202991857602727e-06,
+      "loss": 0.0973,
+      "step": 31150
+    },
+    {
+      "epoch": 2.950198825979928,
+      "grad_norm": 0.4057210087776184,
+      "learning_rate": 8.199204696080289e-06,
+      "loss": 0.1028,
+      "step": 31160
+    },
+    {
+      "epoch": 2.9511456163605376,
+      "grad_norm": 0.41892239451408386,
+      "learning_rate": 8.195417534557849e-06,
+      "loss": 0.0964,
+      "step": 31170
+    },
+    {
+      "epoch": 2.9520924067411474,
+      "grad_norm": 0.4387721121311188,
+      "learning_rate": 8.19163037303541e-06,
+      "loss": 0.111,
+      "step": 31180
+    },
+    {
+      "epoch": 2.953039197121757,
+      "grad_norm": 0.3993951082229614,
+      "learning_rate": 8.18784321151297e-06,
+      "loss": 0.106,
+      "step": 31190
+    },
+    {
+      "epoch": 2.953985987502367,
+      "grad_norm": 0.39773261547088623,
+      "learning_rate": 8.184056049990532e-06,
+      "loss": 0.1029,
+      "step": 31200
+    },
+    {
+      "epoch": 2.9549327778829766,
+      "grad_norm": 0.4858965575695038,
+      "learning_rate": 8.180268888468094e-06,
+      "loss": 0.1036,
+      "step": 31210
+    },
+    {
+      "epoch": 2.9558795682635863,
+      "grad_norm": 0.4160551130771637,
+      "learning_rate": 8.176481726945656e-06,
+      "loss": 0.1045,
+      "step": 31220
+    },
+    {
+      "epoch": 2.956826358644196,
+      "grad_norm": 0.34373006224632263,
+      "learning_rate": 8.172694565423216e-06,
+      "loss": 0.0965,
+      "step": 31230
+    },
+    {
+      "epoch": 2.957773149024806,
+      "grad_norm": 0.416771799325943,
+      "learning_rate": 8.168907403900776e-06,
+      "loss": 0.1053,
+      "step": 31240
+    },
+    {
+      "epoch": 2.9587199394054156,
+      "grad_norm": 0.31747400760650635,
+      "learning_rate": 8.165120242378338e-06,
+      "loss": 0.0999,
+      "step": 31250
+    },
+    {
+      "epoch": 2.9596667297860253,
+      "grad_norm": 0.3813968598842621,
+      "learning_rate": 8.1613330808559e-06,
+      "loss": 0.1069,
+      "step": 31260
+    },
+    {
+      "epoch": 2.960613520166635,
+      "grad_norm": 0.48185887932777405,
+      "learning_rate": 8.157545919333461e-06,
+      "loss": 0.1049,
+      "step": 31270
+    },
+    {
+      "epoch": 2.961560310547245,
+      "grad_norm": 0.4670296311378479,
+      "learning_rate": 8.153758757811021e-06,
+      "loss": 0.0993,
+      "step": 31280
+    },
+    {
+      "epoch": 2.9625071009278545,
+      "grad_norm": 0.41335439682006836,
+      "learning_rate": 8.149971596288583e-06,
+      "loss": 0.1024,
+      "step": 31290
+    },
+    {
+      "epoch": 2.9634538913084643,
+      "grad_norm": 0.5939239263534546,
+      "learning_rate": 8.146184434766143e-06,
+      "loss": 0.1149,
+      "step": 31300
+    },
+    {
+      "epoch": 2.964400681689074,
+      "grad_norm": 0.4662041962146759,
+      "learning_rate": 8.142397273243703e-06,
+      "loss": 0.0962,
+      "step": 31310
+    },
+    {
+      "epoch": 2.9653474720696837,
+      "grad_norm": 0.44135046005249023,
+      "learning_rate": 8.138610111721265e-06,
+      "loss": 0.1022,
+      "step": 31320
+    },
+    {
+      "epoch": 2.9662942624502935,
+      "grad_norm": 0.37937572598457336,
+      "learning_rate": 8.134822950198827e-06,
+      "loss": 0.0962,
+      "step": 31330
+    },
+    {
+      "epoch": 2.9672410528309032,
+      "grad_norm": 0.5369215607643127,
+      "learning_rate": 8.131035788676388e-06,
+      "loss": 0.1068,
+      "step": 31340
+    },
+    {
+      "epoch": 2.968187843211513,
+      "grad_norm": 0.40610820055007935,
+      "learning_rate": 8.127248627153948e-06,
+      "loss": 0.1103,
+      "step": 31350
+    },
+    {
+      "epoch": 2.9691346335921227,
+      "grad_norm": 0.37899497151374817,
+      "learning_rate": 8.12346146563151e-06,
+      "loss": 0.1,
+      "step": 31360
+    },
+    {
+      "epoch": 2.9700814239727324,
+      "grad_norm": 0.387104332447052,
+      "learning_rate": 8.11967430410907e-06,
+      "loss": 0.1107,
+      "step": 31370
+    },
+    {
+      "epoch": 2.971028214353342,
+      "grad_norm": 0.43317079544067383,
+      "learning_rate": 8.115887142586632e-06,
+      "loss": 0.0966,
+      "step": 31380
+    },
+    {
+      "epoch": 2.971975004733952,
+      "grad_norm": 0.4942995309829712,
+      "learning_rate": 8.112099981064194e-06,
+      "loss": 0.1081,
+      "step": 31390
+    },
+    {
+      "epoch": 2.9729217951145617,
+      "grad_norm": 0.43864768743515015,
+      "learning_rate": 8.108312819541754e-06,
+      "loss": 0.1045,
+      "step": 31400
+    },
+    {
+      "epoch": 2.9738685854951714,
+      "grad_norm": 0.4242566227912903,
+      "learning_rate": 8.104525658019316e-06,
+      "loss": 0.1189,
+      "step": 31410
+    },
+    {
+      "epoch": 2.974815375875781,
+      "grad_norm": 0.3926277756690979,
+      "learning_rate": 8.100738496496876e-06,
+      "loss": 0.1059,
+      "step": 31420
+    },
+    {
+      "epoch": 2.975762166256391,
+      "grad_norm": 0.4414288401603699,
+      "learning_rate": 8.096951334974437e-06,
+      "loss": 0.1075,
+      "step": 31430
+    },
+    {
+      "epoch": 2.9767089566370006,
+      "grad_norm": 0.39135444164276123,
+      "learning_rate": 8.093164173451999e-06,
+      "loss": 0.1038,
+      "step": 31440
+    },
+    {
+      "epoch": 2.9776557470176104,
+      "grad_norm": 0.3688848912715912,
+      "learning_rate": 8.089377011929559e-06,
+      "loss": 0.1032,
+      "step": 31450
+    },
+    {
+      "epoch": 2.97860253739822,
+      "grad_norm": 0.3285348117351532,
+      "learning_rate": 8.08558985040712e-06,
+      "loss": 0.0941,
+      "step": 31460
+    },
+    {
+      "epoch": 2.97954932777883,
+      "grad_norm": 0.4666126072406769,
+      "learning_rate": 8.081802688884681e-06,
+      "loss": 0.1117,
+      "step": 31470
+    },
+    {
+      "epoch": 2.9804961181594396,
+      "grad_norm": 0.3238498568534851,
+      "learning_rate": 8.078015527362243e-06,
+      "loss": 0.1034,
+      "step": 31480
+    },
+    {
+      "epoch": 2.9814429085400493,
+      "grad_norm": 0.3677513003349304,
+      "learning_rate": 8.074228365839803e-06,
+      "loss": 0.1045,
+      "step": 31490
+    },
+    {
+      "epoch": 2.982389698920659,
+      "grad_norm": 0.4916920065879822,
+      "learning_rate": 8.070441204317364e-06,
+      "loss": 0.1064,
+      "step": 31500
+    },
+    {
+      "epoch": 2.983336489301269,
+      "grad_norm": 0.47723665833473206,
+      "learning_rate": 8.066654042794926e-06,
+      "loss": 0.1071,
+      "step": 31510
+    },
+    {
+      "epoch": 2.9842832796818786,
+      "grad_norm": 0.43712523579597473,
+      "learning_rate": 8.062866881272486e-06,
+      "loss": 0.106,
+      "step": 31520
+    },
+    {
+      "epoch": 2.9852300700624883,
+      "grad_norm": 0.40085798501968384,
+      "learning_rate": 8.059079719750048e-06,
+      "loss": 0.1053,
+      "step": 31530
+    },
+    {
+      "epoch": 2.986176860443098,
+      "grad_norm": 0.39306896924972534,
+      "learning_rate": 8.055292558227608e-06,
+      "loss": 0.104,
+      "step": 31540
+    },
+    {
+      "epoch": 2.987123650823708,
+      "grad_norm": 0.38417014479637146,
+      "learning_rate": 8.05150539670517e-06,
+      "loss": 0.0958,
+      "step": 31550
+    },
+    {
+      "epoch": 2.9880704412043175,
+      "grad_norm": 0.5245347619056702,
+      "learning_rate": 8.047718235182731e-06,
+      "loss": 0.104,
+      "step": 31560
+    },
+    {
+      "epoch": 2.989017231584927,
+      "grad_norm": 0.5753111839294434,
+      "learning_rate": 8.043931073660293e-06,
+      "loss": 0.1133,
+      "step": 31570
+    },
+    {
+      "epoch": 2.9899640219655366,
+      "grad_norm": 0.4295615553855896,
+      "learning_rate": 8.040143912137853e-06,
+      "loss": 0.0972,
+      "step": 31580
+    },
+    {
+      "epoch": 2.9909108123461463,
+      "grad_norm": 0.37313562631607056,
+      "learning_rate": 8.036356750615413e-06,
+      "loss": 0.1035,
+      "step": 31590
+    },
+    {
+      "epoch": 2.991857602726756,
+      "grad_norm": 0.4777378439903259,
+      "learning_rate": 8.032569589092975e-06,
+      "loss": 0.1051,
+      "step": 31600
+    },
+    {
+      "epoch": 2.992804393107366,
+      "grad_norm": 0.5843793153762817,
+      "learning_rate": 8.028782427570537e-06,
+      "loss": 0.1132,
+      "step": 31610
+    },
+    {
+      "epoch": 2.9937511834879755,
+      "grad_norm": 0.38907530903816223,
+      "learning_rate": 8.024995266048099e-06,
+      "loss": 0.0922,
+      "step": 31620
+    },
+    {
+      "epoch": 2.9946979738685853,
+      "grad_norm": 0.45201992988586426,
+      "learning_rate": 8.021208104525659e-06,
+      "loss": 0.1044,
+      "step": 31630
+    },
+    {
+      "epoch": 2.995644764249195,
+      "grad_norm": 0.35846221446990967,
+      "learning_rate": 8.01742094300322e-06,
+      "loss": 0.1043,
+      "step": 31640
+    },
+    {
+      "epoch": 2.9965915546298048,
+      "grad_norm": 0.3565918803215027,
+      "learning_rate": 8.01363378148078e-06,
+      "loss": 0.0996,
+      "step": 31650
+    },
+    {
+      "epoch": 2.9975383450104145,
+      "grad_norm": 0.4547620117664337,
+      "learning_rate": 8.009846619958342e-06,
+      "loss": 0.103,
+      "step": 31660
+    },
+    {
+      "epoch": 2.9984851353910242,
+      "grad_norm": 0.41886186599731445,
+      "learning_rate": 8.006059458435902e-06,
+      "loss": 0.1025,
+      "step": 31670
+    },
+    {
+      "epoch": 2.999431925771634,
+      "grad_norm": 0.40914952754974365,
+      "learning_rate": 8.002272296913464e-06,
+      "loss": 0.105,
+      "step": 31680
+    },
+    {
+      "epoch": 3.0,
+      "eval_f1_micro": 0.3799037598890792,
+      "eval_loss": 0.11170154809951782,
+      "eval_precision": 0.58840372226199,
+      "eval_recall": 0.2805066543549391,
+      "eval_runtime": 333.7804,
+      "eval_samples_per_second": 126.565,
+      "eval_steps_per_second": 7.912,
+      "step": 31686
+    },
+    {
+      "epoch": 3.0003787161522437,
+      "grad_norm": 0.3885652720928192,
+      "learning_rate": 7.998485135391026e-06,
+      "loss": 0.1001,
+      "step": 31690
+    },
+    {
+      "epoch": 3.0013255065328535,
+      "grad_norm": 0.38781455159187317,
+      "learning_rate": 7.994697973868586e-06,
+      "loss": 0.0948,
+      "step": 31700
+    },
+    {
+      "epoch": 3.002272296913463,
+      "grad_norm": 0.3879997432231903,
+      "learning_rate": 7.990910812346147e-06,
+      "loss": 0.1031,
+      "step": 31710
+    },
+    {
+      "epoch": 3.003219087294073,
+      "grad_norm": 0.4263533055782318,
+      "learning_rate": 7.987123650823708e-06,
+      "loss": 0.0944,
+      "step": 31720
+    },
+    {
+      "epoch": 3.0041658776746827,
+      "grad_norm": 0.3836594521999359,
+      "learning_rate": 7.98333648930127e-06,
+      "loss": 0.096,
+      "step": 31730
+    },
+    {
+      "epoch": 3.0051126680552924,
+      "grad_norm": 0.4415520429611206,
+      "learning_rate": 7.979549327778831e-06,
+      "loss": 0.0971,
+      "step": 31740
+    },
+    {
+      "epoch": 3.006059458435902,
+      "grad_norm": 0.3368381857872009,
+      "learning_rate": 7.975762166256391e-06,
+      "loss": 0.0865,
+      "step": 31750
+    },
+    {
+      "epoch": 3.007006248816512,
+      "grad_norm": 0.47972026467323303,
+      "learning_rate": 7.971975004733953e-06,
+      "loss": 0.0996,
+      "step": 31760
+    },
+    {
+      "epoch": 3.0079530391971216,
+      "grad_norm": 0.44021525979042053,
+      "learning_rate": 7.968187843211513e-06,
+      "loss": 0.102,
+      "step": 31770
+    },
+    {
+      "epoch": 3.0088998295777314,
+      "grad_norm": 0.4182315170764923,
+      "learning_rate": 7.964400681689075e-06,
+      "loss": 0.0991,
+      "step": 31780
+    },
+    {
+      "epoch": 3.009846619958341,
+      "grad_norm": 0.39695531129837036,
+      "learning_rate": 7.960613520166636e-06,
+      "loss": 0.1008,
+      "step": 31790
+    },
+    {
+      "epoch": 3.010793410338951,
+      "grad_norm": 0.36578354239463806,
+      "learning_rate": 7.956826358644196e-06,
+      "loss": 0.089,
+      "step": 31800
+    },
+    {
+      "epoch": 3.0117402007195606,
+      "grad_norm": 0.5266786813735962,
+      "learning_rate": 7.953039197121758e-06,
+      "loss": 0.0971,
+      "step": 31810
+    },
+    {
+      "epoch": 3.0126869911001704,
+      "grad_norm": 0.37619754672050476,
+      "learning_rate": 7.949252035599318e-06,
+      "loss": 0.0951,
+      "step": 31820
+    },
+    {
+      "epoch": 3.01363378148078,
+      "grad_norm": 0.4449104368686676,
+      "learning_rate": 7.94546487407688e-06,
+      "loss": 0.1009,
+      "step": 31830
+    },
+    {
+      "epoch": 3.01458057186139,
+      "grad_norm": 0.44374340772628784,
+      "learning_rate": 7.941677712554442e-06,
+      "loss": 0.0898,
+      "step": 31840
+    },
+    {
+      "epoch": 3.0155273622419996,
+      "grad_norm": 0.4777348041534424,
+      "learning_rate": 7.937890551032002e-06,
+      "loss": 0.0909,
+      "step": 31850
+    },
+    {
+      "epoch": 3.0164741526226093,
+      "grad_norm": 0.4121871888637543,
+      "learning_rate": 7.934103389509563e-06,
+      "loss": 0.0946,
+      "step": 31860
+    },
+    {
+      "epoch": 3.017420943003219,
+      "grad_norm": 0.38763758540153503,
+      "learning_rate": 7.930316227987124e-06,
+      "loss": 0.0976,
+      "step": 31870
+    },
+    {
+      "epoch": 3.018367733383829,
+      "grad_norm": 0.35308313369750977,
+      "learning_rate": 7.926529066464685e-06,
+      "loss": 0.0887,
+      "step": 31880
+    },
+    {
+      "epoch": 3.0193145237644385,
+      "grad_norm": 0.47998329997062683,
+      "learning_rate": 7.922741904942245e-06,
+      "loss": 0.0919,
+      "step": 31890
+    },
+    {
+      "epoch": 3.0202613141450483,
+      "grad_norm": 0.4650914669036865,
+      "learning_rate": 7.918954743419807e-06,
+      "loss": 0.0915,
+      "step": 31900
+    },
+    {
+      "epoch": 3.021208104525658,
+      "grad_norm": 0.3876132369041443,
+      "learning_rate": 7.915167581897369e-06,
+      "loss": 0.0955,
+      "step": 31910
+    },
+    {
+      "epoch": 3.0221548949062678,
+      "grad_norm": 0.524383544921875,
+      "learning_rate": 7.91138042037493e-06,
+      "loss": 0.0975,
+      "step": 31920
+    },
+    {
+      "epoch": 3.0231016852868775,
+      "grad_norm": 0.42943963408470154,
+      "learning_rate": 7.90759325885249e-06,
+      "loss": 0.0814,
+      "step": 31930
+    },
+    {
+      "epoch": 3.0240484756674872,
+      "grad_norm": 0.4390105903148651,
+      "learning_rate": 7.90380609733005e-06,
+      "loss": 0.0999,
+      "step": 31940
+    },
+    {
+      "epoch": 3.024995266048097,
+      "grad_norm": 0.46827784180641174,
+      "learning_rate": 7.900018935807612e-06,
+      "loss": 0.0896,
+      "step": 31950
+    },
+    {
+      "epoch": 3.0259420564287067,
+      "grad_norm": 0.45552289485931396,
+      "learning_rate": 7.896231774285174e-06,
+      "loss": 0.1136,
+      "step": 31960
+    },
+    {
+      "epoch": 3.0268888468093165,
+      "grad_norm": 0.4800921082496643,
+      "learning_rate": 7.892444612762736e-06,
+      "loss": 0.0991,
+      "step": 31970
+    },
+    {
+      "epoch": 3.027835637189926,
+      "grad_norm": 0.484578400850296,
+      "learning_rate": 7.888657451240296e-06,
+      "loss": 0.0985,
+      "step": 31980
+    },
+    {
+      "epoch": 3.028782427570536,
+      "grad_norm": 0.4732307195663452,
+      "learning_rate": 7.884870289717858e-06,
+      "loss": 0.091,
+      "step": 31990
+    },
+    {
+      "epoch": 3.0297292179511457,
+      "grad_norm": 0.4047934114933014,
+      "learning_rate": 7.881083128195418e-06,
+      "loss": 0.0935,
+      "step": 32000
+    },
+    {
+      "epoch": 3.0306760083317554,
+      "grad_norm": 0.32263126969337463,
+      "learning_rate": 7.87729596667298e-06,
+      "loss": 0.0917,
+      "step": 32010
+    },
+    {
+      "epoch": 3.031622798712365,
+      "grad_norm": 0.40960872173309326,
+      "learning_rate": 7.873508805150541e-06,
+      "loss": 0.095,
+      "step": 32020
+    },
+    {
+      "epoch": 3.032569589092975,
+      "grad_norm": 0.30367690324783325,
+      "learning_rate": 7.869721643628101e-06,
+      "loss": 0.0926,
+      "step": 32030
+    },
+    {
+      "epoch": 3.0335163794735847,
+      "grad_norm": 0.348137766122818,
+      "learning_rate": 7.865934482105663e-06,
+      "loss": 0.0897,
+      "step": 32040
+    },
+    {
+      "epoch": 3.0344631698541944,
+      "grad_norm": 0.45152419805526733,
+      "learning_rate": 7.862147320583223e-06,
+      "loss": 0.098,
+      "step": 32050
+    },
+    {
+      "epoch": 3.035409960234804,
+      "grad_norm": 0.47513529658317566,
+      "learning_rate": 7.858360159060785e-06,
+      "loss": 0.1028,
+      "step": 32060
+    },
+    {
+      "epoch": 3.036356750615414,
+      "grad_norm": 0.4300587773323059,
+      "learning_rate": 7.854572997538345e-06,
+      "loss": 0.0911,
+      "step": 32070
+    },
+    {
+      "epoch": 3.0373035409960236,
+      "grad_norm": 0.47339463233947754,
+      "learning_rate": 7.850785836015907e-06,
+      "loss": 0.1049,
+      "step": 32080
+    },
+    {
+      "epoch": 3.0382503313766334,
+      "grad_norm": 0.42077648639678955,
+      "learning_rate": 7.846998674493468e-06,
+      "loss": 0.0898,
+      "step": 32090
+    },
+    {
+      "epoch": 3.039197121757243,
+      "grad_norm": 0.4841423034667969,
+      "learning_rate": 7.843211512971028e-06,
+      "loss": 0.0965,
+      "step": 32100
+    },
+    {
+      "epoch": 3.040143912137853,
+      "grad_norm": 0.6945616006851196,
+      "learning_rate": 7.83942435144859e-06,
+      "loss": 0.0957,
+      "step": 32110
+    },
+    {
+      "epoch": 3.0410907025184626,
+      "grad_norm": 0.573688805103302,
+      "learning_rate": 7.83563718992615e-06,
+      "loss": 0.0892,
+      "step": 32120
+    },
+    {
+      "epoch": 3.0420374928990723,
+      "grad_norm": 0.4494647979736328,
+      "learning_rate": 7.831850028403712e-06,
+      "loss": 0.0929,
+      "step": 32130
+    },
+    {
+      "epoch": 3.042984283279682,
+      "grad_norm": 0.42762213945388794,
+      "learning_rate": 7.828062866881274e-06,
+      "loss": 0.0937,
+      "step": 32140
+    },
+    {
+      "epoch": 3.043931073660292,
+      "grad_norm": 0.6574509143829346,
+      "learning_rate": 7.824275705358835e-06,
+      "loss": 0.1021,
+      "step": 32150
+    },
+    {
+      "epoch": 3.0448778640409015,
+      "grad_norm": 0.5261147022247314,
+      "learning_rate": 7.820488543836395e-06,
+      "loss": 0.0967,
+      "step": 32160
+    },
+    {
+      "epoch": 3.0458246544215113,
+      "grad_norm": 0.509415864944458,
+      "learning_rate": 7.816701382313956e-06,
+      "loss": 0.0968,
+      "step": 32170
+    },
+    {
+      "epoch": 3.046771444802121,
+      "grad_norm": 0.5804086327552795,
+      "learning_rate": 7.812914220791517e-06,
+      "loss": 0.0962,
+      "step": 32180
+    },
+    {
+      "epoch": 3.0477182351827303,
+      "grad_norm": 0.5601807236671448,
+      "learning_rate": 7.809127059269079e-06,
+      "loss": 0.0982,
+      "step": 32190
+    },
+    {
+      "epoch": 3.04866502556334,
+      "grad_norm": 0.42098745703697205,
+      "learning_rate": 7.80533989774664e-06,
+      "loss": 0.0931,
+      "step": 32200
+    },
+    {
+      "epoch": 3.04961181594395,
+      "grad_norm": 0.36883658170700073,
+      "learning_rate": 7.8015527362242e-06,
+      "loss": 0.1013,
+      "step": 32210
+    },
+    {
+      "epoch": 3.0505586063245596,
+      "grad_norm": 0.4977787137031555,
+      "learning_rate": 7.797765574701763e-06,
+      "loss": 0.0951,
+      "step": 32220
+    },
+    {
+      "epoch": 3.0515053967051693,
+      "grad_norm": 0.49785080552101135,
+      "learning_rate": 7.793978413179323e-06,
+      "loss": 0.0951,
+      "step": 32230
+    },
+    {
+      "epoch": 3.052452187085779,
+      "grad_norm": 0.40809834003448486,
+      "learning_rate": 7.790191251656883e-06,
+      "loss": 0.0918,
+      "step": 32240
+    },
+    {
+      "epoch": 3.0533989774663888,
+      "grad_norm": 0.39890798926353455,
+      "learning_rate": 7.786404090134444e-06,
+      "loss": 0.0907,
+      "step": 32250
+    },
+    {
+      "epoch": 3.0543457678469985,
+      "grad_norm": 0.40804323554039,
+      "learning_rate": 7.782616928612006e-06,
+      "loss": 0.0976,
+      "step": 32260
+    },
+    {
+      "epoch": 3.0552925582276083,
+      "grad_norm": 0.48888227343559265,
+      "learning_rate": 7.778829767089568e-06,
+      "loss": 0.0928,
+      "step": 32270
+    },
+    {
+      "epoch": 3.056239348608218,
+      "grad_norm": 0.5478982329368591,
+      "learning_rate": 7.775042605567128e-06,
+      "loss": 0.1114,
+      "step": 32280
+    },
+    {
+      "epoch": 3.0571861389888277,
+      "grad_norm": 0.4051608145236969,
+      "learning_rate": 7.77125544404469e-06,
+      "loss": 0.098,
+      "step": 32290
+    },
+    {
+      "epoch": 3.0581329293694375,
+      "grad_norm": 0.38355210423469543,
+      "learning_rate": 7.76746828252225e-06,
+      "loss": 0.0954,
+      "step": 32300
+    },
+    {
+      "epoch": 3.059079719750047,
+      "grad_norm": 0.46852338314056396,
+      "learning_rate": 7.763681120999811e-06,
+      "loss": 0.0961,
+      "step": 32310
+    },
+    {
+      "epoch": 3.060026510130657,
+      "grad_norm": 0.45753785967826843,
+      "learning_rate": 7.759893959477373e-06,
+      "loss": 0.0942,
+      "step": 32320
+    },
+    {
+      "epoch": 3.0609733005112667,
+      "grad_norm": 0.41343867778778076,
+      "learning_rate": 7.756106797954933e-06,
+      "loss": 0.1011,
+      "step": 32330
+    },
+    {
+      "epoch": 3.0619200908918764,
+      "grad_norm": 0.41205528378486633,
+      "learning_rate": 7.752319636432495e-06,
+      "loss": 0.0954,
+      "step": 32340
+    },
+    {
+      "epoch": 3.062866881272486,
+      "grad_norm": 0.45434126257896423,
+      "learning_rate": 7.748532474910055e-06,
+      "loss": 0.0873,
+      "step": 32350
+    },
+    {
+      "epoch": 3.063813671653096,
+      "grad_norm": 0.45336347818374634,
+      "learning_rate": 7.744745313387617e-06,
+      "loss": 0.0993,
+      "step": 32360
+    },
+    {
+      "epoch": 3.0647604620337057,
+      "grad_norm": 0.4804220199584961,
+      "learning_rate": 7.740958151865179e-06,
+      "loss": 0.0911,
+      "step": 32370
+    },
+    {
+      "epoch": 3.0657072524143154,
+      "grad_norm": 0.43861210346221924,
+      "learning_rate": 7.737170990342739e-06,
+      "loss": 0.0841,
+      "step": 32380
+    },
+    {
+      "epoch": 3.066654042794925,
+      "grad_norm": 0.4974420964717865,
+      "learning_rate": 7.7333838288203e-06,
+      "loss": 0.0982,
+      "step": 32390
+    },
+    {
+      "epoch": 3.067600833175535,
+      "grad_norm": 0.571367621421814,
+      "learning_rate": 7.72959666729786e-06,
+      "loss": 0.1027,
+      "step": 32400
+    },
+    {
+      "epoch": 3.0685476235561446,
+      "grad_norm": 0.4368377923965454,
+      "learning_rate": 7.725809505775422e-06,
+      "loss": 0.0971,
+      "step": 32410
+    },
+    {
+      "epoch": 3.0694944139367544,
+      "grad_norm": 0.5460110306739807,
+      "learning_rate": 7.722022344252982e-06,
+      "loss": 0.094,
+      "step": 32420
+    },
+    {
+      "epoch": 3.070441204317364,
+      "grad_norm": 0.4684341847896576,
+      "learning_rate": 7.718235182730544e-06,
+      "loss": 0.1011,
+      "step": 32430
+    },
+    {
+      "epoch": 3.071387994697974,
+      "grad_norm": 0.7317419648170471,
+      "learning_rate": 7.714448021208106e-06,
+      "loss": 0.0971,
+      "step": 32440
+    },
+    {
+      "epoch": 3.0723347850785836,
+      "grad_norm": 0.3783852159976959,
+      "learning_rate": 7.710660859685666e-06,
+      "loss": 0.0983,
+      "step": 32450
+    },
+    {
+      "epoch": 3.0732815754591933,
+      "grad_norm": 0.4761520326137543,
+      "learning_rate": 7.706873698163227e-06,
+      "loss": 0.0923,
+      "step": 32460
+    },
+    {
+      "epoch": 3.074228365839803,
+      "grad_norm": 0.6367383599281311,
+      "learning_rate": 7.703086536640787e-06,
+      "loss": 0.1035,
+      "step": 32470
+    },
+    {
+      "epoch": 3.075175156220413,
+      "grad_norm": 0.36390215158462524,
+      "learning_rate": 7.69929937511835e-06,
+      "loss": 0.0949,
+      "step": 32480
+    },
+    {
+      "epoch": 3.0761219466010226,
+      "grad_norm": 0.41366979479789734,
+      "learning_rate": 7.695512213595911e-06,
+      "loss": 0.0998,
+      "step": 32490
+    },
+    {
+      "epoch": 3.0770687369816323,
+      "grad_norm": 0.493579238653183,
+      "learning_rate": 7.691725052073473e-06,
+      "loss": 0.0885,
+      "step": 32500
+    },
+    {
+      "epoch": 3.078015527362242,
+      "grad_norm": 0.5382437109947205,
+      "learning_rate": 7.687937890551033e-06,
+      "loss": 0.1057,
+      "step": 32510
+    },
+    {
+      "epoch": 3.078962317742852,
+      "grad_norm": 0.4758876860141754,
+      "learning_rate": 7.684150729028593e-06,
+      "loss": 0.0985,
+      "step": 32520
+    },
+    {
+      "epoch": 3.0799091081234615,
+      "grad_norm": 0.3409093916416168,
+      "learning_rate": 7.680363567506155e-06,
+      "loss": 0.0913,
+      "step": 32530
+    },
+    {
+      "epoch": 3.0808558985040713,
+      "grad_norm": 0.3846043050289154,
+      "learning_rate": 7.676576405983716e-06,
+      "loss": 0.1007,
+      "step": 32540
+    },
+    {
+      "epoch": 3.081802688884681,
+      "grad_norm": 0.5030474662780762,
+      "learning_rate": 7.672789244461278e-06,
+      "loss": 0.0932,
+      "step": 32550
+    },
+    {
+      "epoch": 3.0827494792652907,
+      "grad_norm": 0.5268787145614624,
+      "learning_rate": 7.669002082938838e-06,
+      "loss": 0.1036,
+      "step": 32560
+    },
+    {
+      "epoch": 3.0836962696459005,
+      "grad_norm": 0.49069228768348694,
+      "learning_rate": 7.6652149214164e-06,
+      "loss": 0.1017,
+      "step": 32570
+    },
+    {
+      "epoch": 3.0846430600265102,
+      "grad_norm": 0.44971224665641785,
+      "learning_rate": 7.66142775989396e-06,
+      "loss": 0.1015,
+      "step": 32580
+    },
+    {
+      "epoch": 3.08558985040712,
+      "grad_norm": 0.3594251573085785,
+      "learning_rate": 7.65764059837152e-06,
+      "loss": 0.0984,
+      "step": 32590
+    },
+    {
+      "epoch": 3.0865366407877297,
+      "grad_norm": 0.49989503622055054,
+      "learning_rate": 7.653853436849082e-06,
+      "loss": 0.1014,
+      "step": 32600
+    },
+    {
+      "epoch": 3.0874834311683395,
+      "grad_norm": 0.4332210123538971,
+      "learning_rate": 7.650066275326643e-06,
+      "loss": 0.0901,
+      "step": 32610
+    },
+    {
+      "epoch": 3.088430221548949,
+      "grad_norm": 0.4731793999671936,
+      "learning_rate": 7.646279113804205e-06,
+      "loss": 0.0934,
+      "step": 32620
+    },
+    {
+      "epoch": 3.089377011929559,
+      "grad_norm": 0.5124056339263916,
+      "learning_rate": 7.642491952281765e-06,
+      "loss": 0.096,
+      "step": 32630
+    },
+    {
+      "epoch": 3.0903238023101687,
+      "grad_norm": 0.40301021933555603,
+      "learning_rate": 7.638704790759327e-06,
+      "loss": 0.0938,
+      "step": 32640
+    },
+    {
+      "epoch": 3.0912705926907784,
+      "grad_norm": 0.35446423292160034,
+      "learning_rate": 7.634917629236887e-06,
+      "loss": 0.0888,
+      "step": 32650
+    },
+    {
+      "epoch": 3.092217383071388,
+      "grad_norm": 0.43065381050109863,
+      "learning_rate": 7.631130467714449e-06,
+      "loss": 0.0855,
+      "step": 32660
+    },
+    {
+      "epoch": 3.093164173451998,
+      "grad_norm": 0.41086545586586,
+      "learning_rate": 7.62734330619201e-06,
+      "loss": 0.0959,
+      "step": 32670
+    },
+    {
+      "epoch": 3.0941109638326076,
+      "grad_norm": 0.5487180948257446,
+      "learning_rate": 7.6235561446695705e-06,
+      "loss": 0.0874,
+      "step": 32680
+    },
+    {
+      "epoch": 3.0950577542132174,
+      "grad_norm": 0.4843038320541382,
+      "learning_rate": 7.619768983147132e-06,
+      "loss": 0.0943,
+      "step": 32690
+    },
+    {
+      "epoch": 3.096004544593827,
+      "grad_norm": 0.49333345890045166,
+      "learning_rate": 7.615981821624693e-06,
+      "loss": 0.0942,
+      "step": 32700
+    },
+    {
+      "epoch": 3.096951334974437,
+      "grad_norm": 0.4573739767074585,
+      "learning_rate": 7.612194660102254e-06,
+      "loss": 0.0971,
+      "step": 32710
+    },
+    {
+      "epoch": 3.0978981253550466,
+      "grad_norm": 0.5025661587715149,
+      "learning_rate": 7.608407498579815e-06,
+      "loss": 0.0946,
+      "step": 32720
+    },
+    {
+      "epoch": 3.098844915735656,
+      "grad_norm": 0.3712051212787628,
+      "learning_rate": 7.604620337057376e-06,
+      "loss": 0.095,
+      "step": 32730
+    },
+    {
+      "epoch": 3.0997917061162656,
+      "grad_norm": 0.47726795077323914,
+      "learning_rate": 7.600833175534938e-06,
+      "loss": 0.0951,
+      "step": 32740
+    },
+    {
+      "epoch": 3.1007384964968754,
+      "grad_norm": 0.42070797085762024,
+      "learning_rate": 7.597046014012498e-06,
+      "loss": 0.0957,
+      "step": 32750
+    },
+    {
+      "epoch": 3.101685286877485,
+      "grad_norm": 0.4767979681491852,
+      "learning_rate": 7.593258852490059e-06,
+      "loss": 0.1072,
+      "step": 32760
+    },
+    {
+      "epoch": 3.102632077258095,
+      "grad_norm": 0.522358238697052,
+      "learning_rate": 7.58947169096762e-06,
+      "loss": 0.0874,
+      "step": 32770
+    },
+    {
+      "epoch": 3.1035788676387046,
+      "grad_norm": 0.36201608180999756,
+      "learning_rate": 7.585684529445182e-06,
+      "loss": 0.1002,
+      "step": 32780
+    },
+    {
+      "epoch": 3.1045256580193143,
+      "grad_norm": 0.4515857696533203,
+      "learning_rate": 7.581897367922743e-06,
+      "loss": 0.099,
+      "step": 32790
+    },
+    {
+      "epoch": 3.105472448399924,
+      "grad_norm": 0.3814016878604889,
+      "learning_rate": 7.578110206400303e-06,
+      "loss": 0.0921,
+      "step": 32800
+    },
+    {
+      "epoch": 3.106419238780534,
+      "grad_norm": 0.41360601782798767,
+      "learning_rate": 7.574323044877865e-06,
+      "loss": 0.0864,
+      "step": 32810
+    },
+    {
+      "epoch": 3.1073660291611436,
+      "grad_norm": 0.35619574785232544,
+      "learning_rate": 7.570535883355426e-06,
+      "loss": 0.0928,
+      "step": 32820
+    },
+    {
+      "epoch": 3.1083128195417533,
+      "grad_norm": 0.3933326005935669,
+      "learning_rate": 7.566748721832987e-06,
+      "loss": 0.0877,
+      "step": 32830
+    },
+    {
+      "epoch": 3.109259609922363,
+      "grad_norm": 0.45120447874069214,
+      "learning_rate": 7.5629615603105474e-06,
+      "loss": 0.0911,
+      "step": 32840
+    },
+    {
+      "epoch": 3.110206400302973,
+      "grad_norm": 0.5016258358955383,
+      "learning_rate": 7.559174398788109e-06,
+      "loss": 0.0954,
+      "step": 32850
+    },
+    {
+      "epoch": 3.1111531906835825,
+      "grad_norm": 0.45662152767181396,
+      "learning_rate": 7.55538723726567e-06,
+      "loss": 0.0894,
+      "step": 32860
+    },
+    {
+      "epoch": 3.1120999810641923,
+      "grad_norm": 0.4399343729019165,
+      "learning_rate": 7.551600075743231e-06,
+      "loss": 0.0987,
+      "step": 32870
+    },
+    {
+      "epoch": 3.113046771444802,
+      "grad_norm": 0.4458087980747223,
+      "learning_rate": 7.547812914220793e-06,
+      "loss": 0.0917,
+      "step": 32880
+    },
+    {
+      "epoch": 3.1139935618254118,
+      "grad_norm": 0.3969976603984833,
+      "learning_rate": 7.544025752698353e-06,
+      "loss": 0.1044,
+      "step": 32890
+    },
+    {
+      "epoch": 3.1149403522060215,
+      "grad_norm": 0.4304681718349457,
+      "learning_rate": 7.5402385911759145e-06,
+      "loss": 0.0945,
+      "step": 32900
+    },
+    {
+      "epoch": 3.1158871425866312,
+      "grad_norm": 0.4242285192012787,
+      "learning_rate": 7.536451429653475e-06,
+      "loss": 0.0909,
+      "step": 32910
+    },
+    {
+      "epoch": 3.116833932967241,
+      "grad_norm": 0.5326030254364014,
+      "learning_rate": 7.532664268131037e-06,
+      "loss": 0.101,
+      "step": 32920
+    },
+    {
+      "epoch": 3.1177807233478507,
+      "grad_norm": 0.47638463973999023,
+      "learning_rate": 7.528877106608597e-06,
+      "loss": 0.0937,
+      "step": 32930
+    },
+    {
+      "epoch": 3.1187275137284605,
+      "grad_norm": 0.32601669430732727,
+      "learning_rate": 7.525089945086158e-06,
+      "loss": 0.0916,
+      "step": 32940
+    },
+    {
+      "epoch": 3.11967430410907,
+      "grad_norm": 0.5212317109107971,
+      "learning_rate": 7.52130278356372e-06,
+      "loss": 0.0955,
+      "step": 32950
+    },
+    {
+      "epoch": 3.12062109448968,
+      "grad_norm": 0.5516402125358582,
+      "learning_rate": 7.517515622041281e-06,
+      "loss": 0.1028,
+      "step": 32960
+    },
+    {
+      "epoch": 3.1215678848702897,
+      "grad_norm": 0.4505993127822876,
+      "learning_rate": 7.5137284605188425e-06,
+      "loss": 0.0967,
+      "step": 32970
+    },
+    {
+      "epoch": 3.1225146752508994,
+      "grad_norm": 0.4699845314025879,
+      "learning_rate": 7.5099412989964025e-06,
+      "loss": 0.0915,
+      "step": 32980
+    },
+    {
+      "epoch": 3.123461465631509,
+      "grad_norm": 0.39752283692359924,
+      "learning_rate": 7.506154137473964e-06,
+      "loss": 0.0974,
+      "step": 32990
+    },
+    {
+      "epoch": 3.124408256012119,
+      "grad_norm": 0.40322551131248474,
+      "learning_rate": 7.502366975951525e-06,
+      "loss": 0.1015,
+      "step": 33000
+    },
+    {
+      "epoch": 3.1253550463927287,
+      "grad_norm": 0.38083702325820923,
+      "learning_rate": 7.498579814429085e-06,
+      "loss": 0.0952,
+      "step": 33010
+    },
+    {
+      "epoch": 3.1263018367733384,
+      "grad_norm": 0.44498032331466675,
+      "learning_rate": 7.494792652906647e-06,
+      "loss": 0.0836,
+      "step": 33020
+    },
+    {
+      "epoch": 3.127248627153948,
+      "grad_norm": 0.45370641350746155,
+      "learning_rate": 7.491005491384208e-06,
+      "loss": 0.0993,
+      "step": 33030
+    },
+    {
+      "epoch": 3.128195417534558,
+      "grad_norm": 0.49418359994888306,
+      "learning_rate": 7.48721832986177e-06,
+      "loss": 0.0945,
+      "step": 33040
+    },
+    {
+      "epoch": 3.1291422079151676,
+      "grad_norm": 0.38631823658943176,
+      "learning_rate": 7.4834311683393305e-06,
+      "loss": 0.0982,
+      "step": 33050
+    },
+    {
+      "epoch": 3.1300889982957774,
+      "grad_norm": 0.4079272150993347,
+      "learning_rate": 7.479644006816892e-06,
+      "loss": 0.0911,
+      "step": 33060
+    },
+    {
+      "epoch": 3.131035788676387,
+      "grad_norm": 0.5161040425300598,
+      "learning_rate": 7.475856845294452e-06,
+      "loss": 0.1054,
+      "step": 33070
+    },
+    {
+      "epoch": 3.131982579056997,
+      "grad_norm": 0.4277031719684601,
+      "learning_rate": 7.472069683772013e-06,
+      "loss": 0.0893,
+      "step": 33080
+    },
+    {
+      "epoch": 3.1329293694376066,
+      "grad_norm": 0.3898018002510071,
+      "learning_rate": 7.468282522249575e-06,
+      "loss": 0.1023,
+      "step": 33090
+    },
+    {
+      "epoch": 3.1338761598182163,
+      "grad_norm": 0.48282235860824585,
+      "learning_rate": 7.464495360727135e-06,
+      "loss": 0.0853,
+      "step": 33100
+    },
+    {
+      "epoch": 3.134822950198826,
+      "grad_norm": 0.5204094052314758,
+      "learning_rate": 7.460708199204697e-06,
+      "loss": 0.089,
+      "step": 33110
+    },
+    {
+      "epoch": 3.135769740579436,
+      "grad_norm": 0.49214330315589905,
+      "learning_rate": 7.456921037682258e-06,
+      "loss": 0.0978,
+      "step": 33120
+    },
+    {
+      "epoch": 3.1367165309600455,
+      "grad_norm": 0.4669604003429413,
+      "learning_rate": 7.453133876159819e-06,
+      "loss": 0.0986,
+      "step": 33130
+    },
+    {
+      "epoch": 3.1376633213406553,
+      "grad_norm": 0.4839509427547455,
+      "learning_rate": 7.44934671463738e-06,
+      "loss": 0.1034,
+      "step": 33140
+    },
+    {
+      "epoch": 3.138610111721265,
+      "grad_norm": 0.5297679901123047,
+      "learning_rate": 7.44555955311494e-06,
+      "loss": 0.1051,
+      "step": 33150
+    },
+    {
+      "epoch": 3.1395569021018748,
+      "grad_norm": 0.4383249580860138,
+      "learning_rate": 7.441772391592502e-06,
+      "loss": 0.1006,
+      "step": 33160
+    },
+    {
+      "epoch": 3.1405036924824845,
+      "grad_norm": 0.48694220185279846,
+      "learning_rate": 7.437985230070063e-06,
+      "loss": 0.097,
+      "step": 33170
+    },
+    {
+      "epoch": 3.1414504828630943,
+      "grad_norm": 0.5828351974487305,
+      "learning_rate": 7.434198068547625e-06,
+      "loss": 0.0959,
+      "step": 33180
+    },
+    {
+      "epoch": 3.142397273243704,
+      "grad_norm": 0.4270647168159485,
+      "learning_rate": 7.430410907025185e-06,
+      "loss": 0.1049,
+      "step": 33190
+    },
+    {
+      "epoch": 3.1433440636243137,
+      "grad_norm": 0.5030734539031982,
+      "learning_rate": 7.4266237455027465e-06,
+      "loss": 0.0996,
+      "step": 33200
+    },
+    {
+      "epoch": 3.1442908540049235,
+      "grad_norm": 0.46380719542503357,
+      "learning_rate": 7.422836583980307e-06,
+      "loss": 0.0947,
+      "step": 33210
+    },
+    {
+      "epoch": 3.145237644385533,
+      "grad_norm": 0.4121900796890259,
+      "learning_rate": 7.419049422457868e-06,
+      "loss": 0.0987,
+      "step": 33220
+    },
+    {
+      "epoch": 3.146184434766143,
+      "grad_norm": 0.48857617378234863,
+      "learning_rate": 7.41526226093543e-06,
+      "loss": 0.0987,
+      "step": 33230
+    },
+    {
+      "epoch": 3.1471312251467527,
+      "grad_norm": 0.4057711660861969,
+      "learning_rate": 7.41147509941299e-06,
+      "loss": 0.0938,
+      "step": 33240
+    },
+    {
+      "epoch": 3.1480780155273624,
+      "grad_norm": 0.3517865836620331,
+      "learning_rate": 7.407687937890552e-06,
+      "loss": 0.0943,
+      "step": 33250
+    },
+    {
+      "epoch": 3.149024805907972,
+      "grad_norm": 0.529962420463562,
+      "learning_rate": 7.403900776368113e-06,
+      "loss": 0.1027,
+      "step": 33260
+    },
+    {
+      "epoch": 3.149971596288582,
+      "grad_norm": 0.4298302233219147,
+      "learning_rate": 7.4001136148456744e-06,
+      "loss": 0.0916,
+      "step": 33270
+    },
+    {
+      "epoch": 3.1509183866691917,
+      "grad_norm": 0.34590843319892883,
+      "learning_rate": 7.3963264533232345e-06,
+      "loss": 0.1008,
+      "step": 33280
+    },
+    {
+      "epoch": 3.1518651770498014,
+      "grad_norm": 0.35897600650787354,
+      "learning_rate": 7.392539291800795e-06,
+      "loss": 0.0982,
+      "step": 33290
+    },
+    {
+      "epoch": 3.1528119674304107,
+      "grad_norm": 0.4598219692707062,
+      "learning_rate": 7.388752130278357e-06,
+      "loss": 0.096,
+      "step": 33300
+    },
+    {
+      "epoch": 3.1537587578110204,
+      "grad_norm": 0.5004033446311951,
+      "learning_rate": 7.384964968755918e-06,
+      "loss": 0.0921,
+      "step": 33310
+    },
+    {
+      "epoch": 3.15470554819163,
+      "grad_norm": 0.49685394763946533,
+      "learning_rate": 7.38117780723348e-06,
+      "loss": 0.1034,
+      "step": 33320
+    },
+    {
+      "epoch": 3.15565233857224,
+      "grad_norm": 0.5572829842567444,
+      "learning_rate": 7.37739064571104e-06,
+      "loss": 0.0979,
+      "step": 33330
+    },
+    {
+      "epoch": 3.1565991289528497,
+      "grad_norm": 0.4883466362953186,
+      "learning_rate": 7.3736034841886016e-06,
+      "loss": 0.0931,
+      "step": 33340
+    },
+    {
+      "epoch": 3.1575459193334594,
+      "grad_norm": 0.434589147567749,
+      "learning_rate": 7.3698163226661625e-06,
+      "loss": 0.0954,
+      "step": 33350
+    },
+    {
+      "epoch": 3.158492709714069,
+      "grad_norm": 0.4334976375102997,
+      "learning_rate": 7.3660291611437225e-06,
+      "loss": 0.0976,
+      "step": 33360
+    },
+    {
+      "epoch": 3.159439500094679,
+      "grad_norm": 0.4388626217842102,
+      "learning_rate": 7.362241999621284e-06,
+      "loss": 0.1022,
+      "step": 33370
+    },
+    {
+      "epoch": 3.1603862904752886,
+      "grad_norm": 0.5536752939224243,
+      "learning_rate": 7.358454838098845e-06,
+      "loss": 0.0956,
+      "step": 33380
+    },
+    {
+      "epoch": 3.1613330808558984,
+      "grad_norm": 0.528320848941803,
+      "learning_rate": 7.354667676576407e-06,
+      "loss": 0.0998,
+      "step": 33390
+    },
+    {
+      "epoch": 3.162279871236508,
+      "grad_norm": 0.40832021832466125,
+      "learning_rate": 7.350880515053968e-06,
+      "loss": 0.0959,
+      "step": 33400
+    },
+    {
+      "epoch": 3.163226661617118,
+      "grad_norm": 0.4212343394756317,
+      "learning_rate": 7.3470933535315295e-06,
+      "loss": 0.0933,
+      "step": 33410
+    },
+    {
+      "epoch": 3.1641734519977276,
+      "grad_norm": 0.49932220578193665,
+      "learning_rate": 7.34330619200909e-06,
+      "loss": 0.1002,
+      "step": 33420
+    },
+    {
+      "epoch": 3.1651202423783373,
+      "grad_norm": 0.4118240475654602,
+      "learning_rate": 7.3395190304866505e-06,
+      "loss": 0.0856,
+      "step": 33430
+    },
+    {
+      "epoch": 3.166067032758947,
+      "grad_norm": 0.5317455530166626,
+      "learning_rate": 7.335731868964212e-06,
+      "loss": 0.0918,
+      "step": 33440
+    },
+    {
+      "epoch": 3.167013823139557,
+      "grad_norm": 0.5176566243171692,
+      "learning_rate": 7.331944707441772e-06,
+      "loss": 0.0998,
+      "step": 33450
+    },
+    {
+      "epoch": 3.1679606135201666,
+      "grad_norm": 0.4694359600543976,
+      "learning_rate": 7.328157545919334e-06,
+      "loss": 0.1001,
+      "step": 33460
+    },
+    {
+      "epoch": 3.1689074039007763,
+      "grad_norm": 0.44935131072998047,
+      "learning_rate": 7.324370384396895e-06,
+      "loss": 0.0954,
+      "step": 33470
+    },
+    {
+      "epoch": 3.169854194281386,
+      "grad_norm": 0.49862948060035706,
+      "learning_rate": 7.320583222874457e-06,
+      "loss": 0.1028,
+      "step": 33480
+    },
+    {
+      "epoch": 3.170800984661996,
+      "grad_norm": 0.5710641145706177,
+      "learning_rate": 7.3167960613520176e-06,
+      "loss": 0.0985,
+      "step": 33490
+    },
+    {
+      "epoch": 3.1717477750426055,
+      "grad_norm": 0.44302472472190857,
+      "learning_rate": 7.313008899829578e-06,
+      "loss": 0.103,
+      "step": 33500
+    },
+    {
+      "epoch": 3.1726945654232153,
+      "grad_norm": 0.5398033857345581,
+      "learning_rate": 7.309221738307139e-06,
+      "loss": 0.0995,
+      "step": 33510
+    },
+    {
+      "epoch": 3.173641355803825,
+      "grad_norm": 0.4514923393726349,
+      "learning_rate": 7.3054345767847e-06,
+      "loss": 0.0991,
+      "step": 33520
+    },
+    {
+      "epoch": 3.1745881461844347,
+      "grad_norm": 0.5368667840957642,
+      "learning_rate": 7.301647415262262e-06,
+      "loss": 0.099,
+      "step": 33530
+    },
+    {
+      "epoch": 3.1755349365650445,
+      "grad_norm": 0.46442773938179016,
+      "learning_rate": 7.297860253739822e-06,
+      "loss": 0.1082,
+      "step": 33540
+    },
+    {
+      "epoch": 3.1764817269456542,
+      "grad_norm": 0.5406079888343811,
+      "learning_rate": 7.294073092217384e-06,
+      "loss": 0.0905,
+      "step": 33550
+    },
+    {
+      "epoch": 3.177428517326264,
+      "grad_norm": 0.531243085861206,
+      "learning_rate": 7.290285930694945e-06,
+      "loss": 0.0882,
+      "step": 33560
+    },
+    {
+      "epoch": 3.1783753077068737,
+      "grad_norm": 0.4771122336387634,
+      "learning_rate": 7.286498769172506e-06,
+      "loss": 0.101,
+      "step": 33570
+    },
+    {
+      "epoch": 3.1793220980874835,
+      "grad_norm": 0.5824273824691772,
+      "learning_rate": 7.282711607650067e-06,
+      "loss": 0.0921,
+      "step": 33580
+    },
+    {
+      "epoch": 3.180268888468093,
+      "grad_norm": 0.5940106511116028,
+      "learning_rate": 7.278924446127627e-06,
+      "loss": 0.1092,
+      "step": 33590
+    },
+    {
+      "epoch": 3.181215678848703,
+      "grad_norm": 0.4634968042373657,
+      "learning_rate": 7.275137284605189e-06,
+      "loss": 0.095,
+      "step": 33600
+    },
+    {
+      "epoch": 3.1821624692293127,
+      "grad_norm": 0.5046654939651489,
+      "learning_rate": 7.27135012308275e-06,
+      "loss": 0.0964,
+      "step": 33610
+    },
+    {
+      "epoch": 3.1831092596099224,
+      "grad_norm": 0.46652933955192566,
+      "learning_rate": 7.267562961560312e-06,
+      "loss": 0.0892,
+      "step": 33620
+    },
+    {
+      "epoch": 3.184056049990532,
+      "grad_norm": 0.5102341175079346,
+      "learning_rate": 7.263775800037872e-06,
+      "loss": 0.093,
+      "step": 33630
+    },
+    {
+      "epoch": 3.185002840371142,
+      "grad_norm": 0.5113751292228699,
+      "learning_rate": 7.259988638515433e-06,
+      "loss": 0.1,
+      "step": 33640
+    },
+    {
+      "epoch": 3.1859496307517516,
+      "grad_norm": 0.537693202495575,
+      "learning_rate": 7.2562014769929944e-06,
+      "loss": 0.096,
+      "step": 33650
+    },
+    {
+      "epoch": 3.1868964211323614,
+      "grad_norm": 0.5255269408226013,
+      "learning_rate": 7.252414315470555e-06,
+      "loss": 0.0978,
+      "step": 33660
+    },
+    {
+      "epoch": 3.187843211512971,
+      "grad_norm": 0.4506100118160248,
+      "learning_rate": 7.248627153948117e-06,
+      "loss": 0.096,
+      "step": 33670
+    },
+    {
+      "epoch": 3.188790001893581,
+      "grad_norm": 0.39353442192077637,
+      "learning_rate": 7.244839992425677e-06,
+      "loss": 0.0947,
+      "step": 33680
+    },
+    {
+      "epoch": 3.1897367922741906,
+      "grad_norm": 0.5483629703521729,
+      "learning_rate": 7.241052830903239e-06,
+      "loss": 0.0959,
+      "step": 33690
+    },
+    {
+      "epoch": 3.1906835826548003,
+      "grad_norm": 0.42552974820137024,
+      "learning_rate": 7.2372656693808e-06,
+      "loss": 0.0984,
+      "step": 33700
+    },
+    {
+      "epoch": 3.19163037303541,
+      "grad_norm": 0.6715368032455444,
+      "learning_rate": 7.233478507858361e-06,
+      "loss": 0.0949,
+      "step": 33710
+    },
+    {
+      "epoch": 3.19257716341602,
+      "grad_norm": 0.40398454666137695,
+      "learning_rate": 7.2296913463359216e-06,
+      "loss": 0.0889,
+      "step": 33720
+    },
+    {
+      "epoch": 3.1935239537966296,
+      "grad_norm": 0.42471370100975037,
+      "learning_rate": 7.2259041848134825e-06,
+      "loss": 0.0875,
+      "step": 33730
+    },
+    {
+      "epoch": 3.1944707441772393,
+      "grad_norm": 0.37097862362861633,
+      "learning_rate": 7.222117023291044e-06,
+      "loss": 0.0979,
+      "step": 33740
+    },
+    {
+      "epoch": 3.195417534557849,
+      "grad_norm": 0.514901340007782,
+      "learning_rate": 7.218329861768605e-06,
+      "loss": 0.0887,
+      "step": 33750
+    },
+    {
+      "epoch": 3.196364324938459,
+      "grad_norm": 0.5147632956504822,
+      "learning_rate": 7.214542700246167e-06,
+      "loss": 0.0991,
+      "step": 33760
+    },
+    {
+      "epoch": 3.1973111153190685,
+      "grad_norm": 0.48616480827331543,
+      "learning_rate": 7.210755538723727e-06,
+      "loss": 0.0976,
+      "step": 33770
+    },
+    {
+      "epoch": 3.1982579056996783,
+      "grad_norm": 0.5624353885650635,
+      "learning_rate": 7.206968377201288e-06,
+      "loss": 0.0962,
+      "step": 33780
+    },
+    {
+      "epoch": 3.199204696080288,
+      "grad_norm": 0.4632226526737213,
+      "learning_rate": 7.2031812156788495e-06,
+      "loss": 0.0989,
+      "step": 33790
+    },
+    {
+      "epoch": 3.2001514864608978,
+      "grad_norm": 0.403451144695282,
+      "learning_rate": 7.1993940541564104e-06,
+      "loss": 0.0923,
+      "step": 33800
+    },
+    {
+      "epoch": 3.2010982768415075,
+      "grad_norm": 0.37990331649780273,
+      "learning_rate": 7.195606892633971e-06,
+      "loss": 0.1006,
+      "step": 33810
+    },
+    {
+      "epoch": 3.202045067222117,
+      "grad_norm": 0.5115433931350708,
+      "learning_rate": 7.191819731111532e-06,
+      "loss": 0.0984,
+      "step": 33820
+    },
+    {
+      "epoch": 3.2029918576027265,
+      "grad_norm": 0.4312513470649719,
+      "learning_rate": 7.188032569589094e-06,
+      "loss": 0.0887,
+      "step": 33830
+    },
+    {
+      "epoch": 3.2039386479833363,
+      "grad_norm": 0.46648553013801575,
+      "learning_rate": 7.184245408066655e-06,
+      "loss": 0.0893,
+      "step": 33840
+    },
+    {
+      "epoch": 3.204885438363946,
+      "grad_norm": 0.4689205586910248,
+      "learning_rate": 7.180458246544215e-06,
+      "loss": 0.0991,
+      "step": 33850
+    },
+    {
+      "epoch": 3.2058322287445558,
+      "grad_norm": 0.3541924059391022,
+      "learning_rate": 7.176671085021777e-06,
+      "loss": 0.0871,
+      "step": 33860
+    },
+    {
+      "epoch": 3.2067790191251655,
+      "grad_norm": 0.45448562502861023,
+      "learning_rate": 7.1728839234993376e-06,
+      "loss": 0.0977,
+      "step": 33870
+    },
+    {
+      "epoch": 3.2077258095057752,
+      "grad_norm": 0.39853841066360474,
+      "learning_rate": 7.169096761976899e-06,
+      "loss": 0.0911,
+      "step": 33880
+    },
+    {
+      "epoch": 3.208672599886385,
+      "grad_norm": 0.6429105997085571,
+      "learning_rate": 7.16530960045446e-06,
+      "loss": 0.105,
+      "step": 33890
+    },
+    {
+      "epoch": 3.2096193902669947,
+      "grad_norm": 0.3810679614543915,
+      "learning_rate": 7.161522438932021e-06,
+      "loss": 0.0865,
+      "step": 33900
+    },
+    {
+      "epoch": 3.2105661806476045,
+      "grad_norm": 0.43609294295310974,
+      "learning_rate": 7.157735277409582e-06,
+      "loss": 0.0947,
+      "step": 33910
+    },
+    {
+      "epoch": 3.211512971028214,
+      "grad_norm": 0.644542932510376,
+      "learning_rate": 7.153948115887143e-06,
+      "loss": 0.1029,
+      "step": 33920
+    },
+    {
+      "epoch": 3.212459761408824,
+      "grad_norm": 0.4697587490081787,
+      "learning_rate": 7.150160954364705e-06,
+      "loss": 0.1001,
+      "step": 33930
+    },
+    {
+      "epoch": 3.2134065517894337,
+      "grad_norm": 0.5083484649658203,
+      "learning_rate": 7.146373792842265e-06,
+      "loss": 0.1025,
+      "step": 33940
+    },
+    {
+      "epoch": 3.2143533421700434,
+      "grad_norm": 0.4288027882575989,
+      "learning_rate": 7.142586631319826e-06,
+      "loss": 0.1035,
+      "step": 33950
+    },
+    {
+      "epoch": 3.215300132550653,
+      "grad_norm": 0.38893651962280273,
+      "learning_rate": 7.138799469797387e-06,
+      "loss": 0.0987,
+      "step": 33960
+    },
+    {
+      "epoch": 3.216246922931263,
+      "grad_norm": 0.4421164393424988,
+      "learning_rate": 7.135012308274949e-06,
+      "loss": 0.1082,
+      "step": 33970
+    },
+    {
+      "epoch": 3.2171937133118726,
+      "grad_norm": 0.4417242109775543,
+      "learning_rate": 7.13122514675251e-06,
+      "loss": 0.0949,
+      "step": 33980
+    },
+    {
+      "epoch": 3.2181405036924824,
+      "grad_norm": 0.3973984718322754,
+      "learning_rate": 7.12743798523007e-06,
+      "loss": 0.0934,
+      "step": 33990
+    },
+    {
+      "epoch": 3.219087294073092,
+      "grad_norm": 0.39709991216659546,
+      "learning_rate": 7.123650823707632e-06,
+      "loss": 0.0937,
+      "step": 34000
+    },
+    {
+      "epoch": 3.220034084453702,
+      "grad_norm": 0.42711204290390015,
+      "learning_rate": 7.119863662185193e-06,
+      "loss": 0.0929,
+      "step": 34010
+    },
+    {
+      "epoch": 3.2209808748343116,
+      "grad_norm": 0.48301854729652405,
+      "learning_rate": 7.116076500662754e-06,
+      "loss": 0.1032,
+      "step": 34020
+    },
+    {
+      "epoch": 3.2219276652149214,
+      "grad_norm": 0.47069182991981506,
+      "learning_rate": 7.1122893391403144e-06,
+      "loss": 0.0938,
+      "step": 34030
+    },
+    {
+      "epoch": 3.222874455595531,
+      "grad_norm": 0.36594197154045105,
+      "learning_rate": 7.108502177617876e-06,
+      "loss": 0.088,
+      "step": 34040
+    },
+    {
+      "epoch": 3.223821245976141,
+      "grad_norm": 0.42151713371276855,
+      "learning_rate": 7.104715016095437e-06,
+      "loss": 0.0945,
+      "step": 34050
+    },
+    {
+      "epoch": 3.2247680363567506,
+      "grad_norm": 0.37251588702201843,
+      "learning_rate": 7.100927854572998e-06,
+      "loss": 0.0957,
+      "step": 34060
+    },
+    {
+      "epoch": 3.2257148267373603,
+      "grad_norm": 0.46561744809150696,
+      "learning_rate": 7.09714069305056e-06,
+      "loss": 0.0978,
+      "step": 34070
+    },
+    {
+      "epoch": 3.22666161711797,
+      "grad_norm": 0.5953252911567688,
+      "learning_rate": 7.09335353152812e-06,
+      "loss": 0.1009,
+      "step": 34080
+    },
+    {
+      "epoch": 3.22760840749858,
+      "grad_norm": 0.5023998022079468,
+      "learning_rate": 7.0895663700056815e-06,
+      "loss": 0.1021,
+      "step": 34090
+    },
+    {
+      "epoch": 3.2285551978791895,
+      "grad_norm": 0.49502333998680115,
+      "learning_rate": 7.085779208483242e-06,
+      "loss": 0.0945,
+      "step": 34100
+    },
+    {
+      "epoch": 3.2295019882597993,
+      "grad_norm": 0.5286855697631836,
+      "learning_rate": 7.081992046960804e-06,
+      "loss": 0.0947,
+      "step": 34110
+    },
+    {
+      "epoch": 3.230448778640409,
+      "grad_norm": 0.4840443730354309,
+      "learning_rate": 7.078204885438364e-06,
+      "loss": 0.0993,
+      "step": 34120
+    },
+    {
+      "epoch": 3.2313955690210188,
+      "grad_norm": 0.4057762920856476,
+      "learning_rate": 7.074417723915925e-06,
+      "loss": 0.0879,
+      "step": 34130
+    },
+    {
+      "epoch": 3.2323423594016285,
+      "grad_norm": 0.6441464424133301,
+      "learning_rate": 7.070630562393487e-06,
+      "loss": 0.1014,
+      "step": 34140
+    },
+    {
+      "epoch": 3.2332891497822382,
+      "grad_norm": 0.4571669399738312,
+      "learning_rate": 7.066843400871048e-06,
+      "loss": 0.1046,
+      "step": 34150
+    },
+    {
+      "epoch": 3.234235940162848,
+      "grad_norm": 0.4196922779083252,
+      "learning_rate": 7.0630562393486095e-06,
+      "loss": 0.0954,
+      "step": 34160
+    },
+    {
+      "epoch": 3.2351827305434577,
+      "grad_norm": 0.3758500814437866,
+      "learning_rate": 7.0592690778261695e-06,
+      "loss": 0.0786,
+      "step": 34170
+    },
+    {
+      "epoch": 3.2361295209240675,
+      "grad_norm": 0.47714629769325256,
+      "learning_rate": 7.055481916303731e-06,
+      "loss": 0.1078,
+      "step": 34180
+    },
+    {
+      "epoch": 3.237076311304677,
+      "grad_norm": 0.4457477033138275,
+      "learning_rate": 7.051694754781292e-06,
+      "loss": 0.0975,
+      "step": 34190
+    },
+    {
+      "epoch": 3.238023101685287,
+      "grad_norm": 0.469585120677948,
+      "learning_rate": 7.047907593258852e-06,
+      "loss": 0.0982,
+      "step": 34200
+    },
+    {
+      "epoch": 3.2389698920658967,
+      "grad_norm": 0.4486069083213806,
+      "learning_rate": 7.044120431736414e-06,
+      "loss": 0.0991,
+      "step": 34210
+    },
+    {
+      "epoch": 3.2399166824465064,
+      "grad_norm": 0.5082442164421082,
+      "learning_rate": 7.040333270213975e-06,
+      "loss": 0.1033,
+      "step": 34220
+    },
+    {
+      "epoch": 3.240863472827116,
+      "grad_norm": 0.4870776832103729,
+      "learning_rate": 7.036546108691537e-06,
+      "loss": 0.0845,
+      "step": 34230
+    },
+    {
+      "epoch": 3.241810263207726,
+      "grad_norm": 0.5926299095153809,
+      "learning_rate": 7.0327589471690975e-06,
+      "loss": 0.0862,
+      "step": 34240
+    },
+    {
+      "epoch": 3.2427570535883357,
+      "grad_norm": 0.4643210470676422,
+      "learning_rate": 7.028971785646659e-06,
+      "loss": 0.0884,
+      "step": 34250
+    },
+    {
+      "epoch": 3.2437038439689454,
+      "grad_norm": 0.46026501059532166,
+      "learning_rate": 7.025184624124219e-06,
+      "loss": 0.0967,
+      "step": 34260
+    },
+    {
+      "epoch": 3.244650634349555,
+      "grad_norm": 0.4218289852142334,
+      "learning_rate": 7.02139746260178e-06,
+      "loss": 0.0884,
+      "step": 34270
+    },
+    {
+      "epoch": 3.245597424730165,
+      "grad_norm": 0.4901501536369324,
+      "learning_rate": 7.017610301079342e-06,
+      "loss": 0.1004,
+      "step": 34280
+    },
+    {
+      "epoch": 3.2465442151107746,
+      "grad_norm": 0.39414164423942566,
+      "learning_rate": 7.013823139556902e-06,
+      "loss": 0.0874,
+      "step": 34290
+    },
+    {
+      "epoch": 3.2474910054913844,
+      "grad_norm": 0.4422838091850281,
+      "learning_rate": 7.010035978034464e-06,
+      "loss": 0.0935,
+      "step": 34300
+    },
+    {
+      "epoch": 3.248437795871994,
+      "grad_norm": 0.41335728764533997,
+      "learning_rate": 7.006248816512025e-06,
+      "loss": 0.0939,
+      "step": 34310
+    },
+    {
+      "epoch": 3.249384586252604,
+      "grad_norm": 0.5088624954223633,
+      "learning_rate": 7.002461654989586e-06,
+      "loss": 0.1019,
+      "step": 34320
+    },
+    {
+      "epoch": 3.2503313766332136,
+      "grad_norm": 0.6368246078491211,
+      "learning_rate": 6.998674493467147e-06,
+      "loss": 0.1008,
+      "step": 34330
+    },
+    {
+      "epoch": 3.2512781670138233,
+      "grad_norm": 0.47923582792282104,
+      "learning_rate": 6.994887331944707e-06,
+      "loss": 0.1007,
+      "step": 34340
+    },
+    {
+      "epoch": 3.252224957394433,
+      "grad_norm": 0.3951050341129303,
+      "learning_rate": 6.991100170422269e-06,
+      "loss": 0.1036,
+      "step": 34350
+    },
+    {
+      "epoch": 3.253171747775043,
+      "grad_norm": 0.4178468883037567,
+      "learning_rate": 6.98731300889983e-06,
+      "loss": 0.0983,
+      "step": 34360
+    },
+    {
+      "epoch": 3.2541185381556526,
+      "grad_norm": 0.4253663420677185,
+      "learning_rate": 6.983525847377392e-06,
+      "loss": 0.0855,
+      "step": 34370
+    },
+    {
+      "epoch": 3.2550653285362623,
+      "grad_norm": 0.47536349296569824,
+      "learning_rate": 6.979738685854952e-06,
+      "loss": 0.0934,
+      "step": 34380
+    },
+    {
+      "epoch": 3.256012118916872,
+      "grad_norm": 0.49580034613609314,
+      "learning_rate": 6.9759515243325135e-06,
+      "loss": 0.0931,
+      "step": 34390
+    },
+    {
+      "epoch": 3.2569589092974818,
+      "grad_norm": 0.5044097900390625,
+      "learning_rate": 6.972164362810074e-06,
+      "loss": 0.0866,
+      "step": 34400
+    },
+    {
+      "epoch": 3.2579056996780915,
+      "grad_norm": 0.5280126333236694,
+      "learning_rate": 6.968377201287635e-06,
+      "loss": 0.0879,
+      "step": 34410
+    },
+    {
+      "epoch": 3.258852490058701,
+      "grad_norm": 0.5161647200584412,
+      "learning_rate": 6.964590039765197e-06,
+      "loss": 0.1006,
+      "step": 34420
+    },
+    {
+      "epoch": 3.2597992804393106,
+      "grad_norm": 0.48363789916038513,
+      "learning_rate": 6.960802878242757e-06,
+      "loss": 0.0898,
+      "step": 34430
+    },
+    {
+      "epoch": 3.2607460708199203,
+      "grad_norm": 0.47649186849594116,
+      "learning_rate": 6.957015716720319e-06,
+      "loss": 0.0884,
+      "step": 34440
+    },
+    {
+      "epoch": 3.26169286120053,
+      "grad_norm": 0.535963237285614,
+      "learning_rate": 6.95322855519788e-06,
+      "loss": 0.1064,
+      "step": 34450
+    },
+    {
+      "epoch": 3.2626396515811398,
+      "grad_norm": 0.5076953768730164,
+      "learning_rate": 6.9494413936754415e-06,
+      "loss": 0.0916,
+      "step": 34460
+    },
+    {
+      "epoch": 3.2635864419617495,
+      "grad_norm": 0.4224216639995575,
+      "learning_rate": 6.9456542321530015e-06,
+      "loss": 0.0994,
+      "step": 34470
+    },
+    {
+      "epoch": 3.2645332323423593,
+      "grad_norm": 0.6081414222717285,
+      "learning_rate": 6.941867070630562e-06,
+      "loss": 0.0992,
+      "step": 34480
+    },
+    {
+      "epoch": 3.265480022722969,
+      "grad_norm": 0.5121296644210815,
+      "learning_rate": 6.938079909108124e-06,
+      "loss": 0.0913,
+      "step": 34490
+    },
+    {
+      "epoch": 3.2664268131035787,
+      "grad_norm": 0.4258511960506439,
+      "learning_rate": 6.934292747585685e-06,
+      "loss": 0.0968,
+      "step": 34500
+    },
+    {
+      "epoch": 3.2673736034841885,
+      "grad_norm": 0.5431085228919983,
+      "learning_rate": 6.930505586063247e-06,
+      "loss": 0.0998,
+      "step": 34510
+    },
+    {
+      "epoch": 3.2683203938647982,
+      "grad_norm": 0.5170323252677917,
+      "learning_rate": 6.926718424540807e-06,
+      "loss": 0.0922,
+      "step": 34520
+    },
+    {
+      "epoch": 3.269267184245408,
+      "grad_norm": 0.5068280100822449,
+      "learning_rate": 6.922931263018369e-06,
+      "loss": 0.0962,
+      "step": 34530
+    },
+    {
+      "epoch": 3.2702139746260177,
+      "grad_norm": 0.5490415692329407,
+      "learning_rate": 6.9191441014959295e-06,
+      "loss": 0.0962,
+      "step": 34540
+    },
+    {
+      "epoch": 3.2711607650066274,
+      "grad_norm": 0.4179328382015228,
+      "learning_rate": 6.9153569399734895e-06,
+      "loss": 0.0972,
+      "step": 34550
+    },
+    {
+      "epoch": 3.272107555387237,
+      "grad_norm": 0.5323558449745178,
+      "learning_rate": 6.911569778451051e-06,
+      "loss": 0.0989,
+      "step": 34560
+    },
+    {
+      "epoch": 3.273054345767847,
+      "grad_norm": 0.4823920726776123,
+      "learning_rate": 6.907782616928612e-06,
+      "loss": 0.0944,
+      "step": 34570
+    },
+    {
+      "epoch": 3.2740011361484567,
+      "grad_norm": 0.3574821352958679,
+      "learning_rate": 6.903995455406174e-06,
+      "loss": 0.0992,
+      "step": 34580
+    },
+    {
+      "epoch": 3.2749479265290664,
+      "grad_norm": 0.4525294005870819,
+      "learning_rate": 6.900208293883735e-06,
+      "loss": 0.1004,
+      "step": 34590
+    },
+    {
+      "epoch": 3.275894716909676,
+      "grad_norm": 0.521497905254364,
+      "learning_rate": 6.8964211323612966e-06,
+      "loss": 0.0967,
+      "step": 34600
+    },
+    {
+      "epoch": 3.276841507290286,
+      "grad_norm": 0.4805905222892761,
+      "learning_rate": 6.892633970838857e-06,
+      "loss": 0.1,
+      "step": 34610
+    },
+    {
+      "epoch": 3.2777882976708956,
+      "grad_norm": 0.6537460684776306,
+      "learning_rate": 6.8888468093164175e-06,
+      "loss": 0.0977,
+      "step": 34620
+    },
+    {
+      "epoch": 3.2787350880515054,
+      "grad_norm": 0.6015987396240234,
+      "learning_rate": 6.885059647793979e-06,
+      "loss": 0.0972,
+      "step": 34630
+    },
+    {
+      "epoch": 3.279681878432115,
+      "grad_norm": 0.6600093841552734,
+      "learning_rate": 6.881272486271539e-06,
+      "loss": 0.0933,
+      "step": 34640
+    },
+    {
+      "epoch": 3.280628668812725,
+      "grad_norm": 0.5536496043205261,
+      "learning_rate": 6.877485324749101e-06,
+      "loss": 0.1031,
+      "step": 34650
+    },
+    {
+      "epoch": 3.2815754591933346,
+      "grad_norm": 0.4587319791316986,
+      "learning_rate": 6.873698163226662e-06,
+      "loss": 0.0879,
+      "step": 34660
+    },
+    {
+      "epoch": 3.2825222495739443,
+      "grad_norm": 0.6899036169052124,
+      "learning_rate": 6.869911001704224e-06,
+      "loss": 0.0961,
+      "step": 34670
+    },
+    {
+      "epoch": 3.283469039954554,
+      "grad_norm": 0.488418310880661,
+      "learning_rate": 6.8661238401817846e-06,
+      "loss": 0.1059,
+      "step": 34680
+    },
+    {
+      "epoch": 3.284415830335164,
+      "grad_norm": 0.42573970556259155,
+      "learning_rate": 6.862336678659345e-06,
+      "loss": 0.1001,
+      "step": 34690
+    },
+    {
+      "epoch": 3.2853626207157736,
+      "grad_norm": 0.5093210339546204,
+      "learning_rate": 6.858549517136906e-06,
+      "loss": 0.0984,
+      "step": 34700
+    },
+    {
+      "epoch": 3.2863094110963833,
+      "grad_norm": 0.5334013104438782,
+      "learning_rate": 6.854762355614467e-06,
+      "loss": 0.1044,
+      "step": 34710
+    },
+    {
+      "epoch": 3.287256201476993,
+      "grad_norm": 0.5470275282859802,
+      "learning_rate": 6.850975194092029e-06,
+      "loss": 0.0953,
+      "step": 34720
+    },
+    {
+      "epoch": 3.288202991857603,
+      "grad_norm": 0.4627389907836914,
+      "learning_rate": 6.847188032569589e-06,
+      "loss": 0.0926,
+      "step": 34730
+    },
+    {
+      "epoch": 3.2891497822382125,
+      "grad_norm": 0.4343864917755127,
+      "learning_rate": 6.843400871047151e-06,
+      "loss": 0.1108,
+      "step": 34740
+    },
+    {
+      "epoch": 3.2900965726188223,
+      "grad_norm": 0.4533028304576874,
+      "learning_rate": 6.839613709524712e-06,
+      "loss": 0.1036,
+      "step": 34750
+    },
+    {
+      "epoch": 3.291043362999432,
+      "grad_norm": 0.5282678008079529,
+      "learning_rate": 6.8358265480022734e-06,
+      "loss": 0.1052,
+      "step": 34760
+    },
+    {
+      "epoch": 3.2919901533800418,
+      "grad_norm": 0.4647209644317627,
+      "learning_rate": 6.832039386479834e-06,
+      "loss": 0.0928,
+      "step": 34770
+    },
+    {
+      "epoch": 3.2929369437606515,
+      "grad_norm": 0.5247597694396973,
+      "learning_rate": 6.828252224957394e-06,
+      "loss": 0.0964,
+      "step": 34780
+    },
+    {
+      "epoch": 3.2938837341412612,
+      "grad_norm": 0.5702017545700073,
+      "learning_rate": 6.824465063434956e-06,
+      "loss": 0.1065,
+      "step": 34790
+    },
+    {
+      "epoch": 3.294830524521871,
+      "grad_norm": 0.5594719052314758,
+      "learning_rate": 6.820677901912517e-06,
+      "loss": 0.0919,
+      "step": 34800
+    },
+    {
+      "epoch": 3.2957773149024807,
+      "grad_norm": 0.4289184510707855,
+      "learning_rate": 6.816890740390079e-06,
+      "loss": 0.0946,
+      "step": 34810
+    },
+    {
+      "epoch": 3.2967241052830905,
+      "grad_norm": 0.4763040542602539,
+      "learning_rate": 6.813103578867639e-06,
+      "loss": 0.0976,
+      "step": 34820
+    },
+    {
+      "epoch": 3.2976708956637,
+      "grad_norm": 0.4474315643310547,
+      "learning_rate": 6.8093164173452006e-06,
+      "loss": 0.0918,
+      "step": 34830
+    },
+    {
+      "epoch": 3.29861768604431,
+      "grad_norm": 0.5152229070663452,
+      "learning_rate": 6.8055292558227615e-06,
+      "loss": 0.0953,
+      "step": 34840
+    },
+    {
+      "epoch": 3.2995644764249197,
+      "grad_norm": 0.43370160460472107,
+      "learning_rate": 6.801742094300322e-06,
+      "loss": 0.0957,
+      "step": 34850
+    },
+    {
+      "epoch": 3.3005112668055294,
+      "grad_norm": 0.4983856976032257,
+      "learning_rate": 6.797954932777884e-06,
+      "loss": 0.0954,
+      "step": 34860
+    },
+    {
+      "epoch": 3.301458057186139,
+      "grad_norm": 0.43842777609825134,
+      "learning_rate": 6.794167771255444e-06,
+      "loss": 0.0968,
+      "step": 34870
+    },
+    {
+      "epoch": 3.302404847566749,
+      "grad_norm": 0.41550055146217346,
+      "learning_rate": 6.790380609733006e-06,
+      "loss": 0.0988,
+      "step": 34880
+    },
+    {
+      "epoch": 3.3033516379473586,
+      "grad_norm": 0.45826441049575806,
+      "learning_rate": 6.786593448210567e-06,
+      "loss": 0.0923,
+      "step": 34890
+    },
+    {
+      "epoch": 3.304298428327968,
+      "grad_norm": 0.5877769589424133,
+      "learning_rate": 6.7828062866881285e-06,
+      "loss": 0.0995,
+      "step": 34900
+    },
+    {
+      "epoch": 3.3052452187085777,
+      "grad_norm": 0.5494678616523743,
+      "learning_rate": 6.779019125165689e-06,
+      "loss": 0.0944,
+      "step": 34910
+    },
+    {
+      "epoch": 3.3061920090891874,
+      "grad_norm": 0.48389655351638794,
+      "learning_rate": 6.7752319636432495e-06,
+      "loss": 0.1098,
+      "step": 34920
+    },
+    {
+      "epoch": 3.307138799469797,
+      "grad_norm": 0.4657025635242462,
+      "learning_rate": 6.771444802120811e-06,
+      "loss": 0.0941,
+      "step": 34930
+    },
+    {
+      "epoch": 3.308085589850407,
+      "grad_norm": 0.4596572816371918,
+      "learning_rate": 6.767657640598372e-06,
+      "loss": 0.0945,
+      "step": 34940
+    },
+    {
+      "epoch": 3.3090323802310166,
+      "grad_norm": 0.5223503708839417,
+      "learning_rate": 6.763870479075934e-06,
+      "loss": 0.1009,
+      "step": 34950
+    },
+    {
+      "epoch": 3.3099791706116264,
+      "grad_norm": 0.6406366229057312,
+      "learning_rate": 6.760083317553494e-06,
+      "loss": 0.0995,
+      "step": 34960
+    },
+    {
+      "epoch": 3.310925960992236,
+      "grad_norm": 0.4907243549823761,
+      "learning_rate": 6.756296156031056e-06,
+      "loss": 0.0971,
+      "step": 34970
+    },
+    {
+      "epoch": 3.311872751372846,
+      "grad_norm": 0.49459734559059143,
+      "learning_rate": 6.7525089945086166e-06,
+      "loss": 0.0962,
+      "step": 34980
+    },
+    {
+      "epoch": 3.3128195417534556,
+      "grad_norm": 0.46839937567710876,
+      "learning_rate": 6.748721832986177e-06,
+      "loss": 0.0904,
+      "step": 34990
+    },
+    {
+      "epoch": 3.3137663321340654,
+      "grad_norm": 0.4720109701156616,
+      "learning_rate": 6.744934671463738e-06,
+      "loss": 0.1,
+      "step": 35000
+    },
+    {
+      "epoch": 3.314713122514675,
+      "grad_norm": 0.49172693490982056,
+      "learning_rate": 6.741147509941299e-06,
+      "loss": 0.0925,
+      "step": 35010
+    },
+    {
+      "epoch": 3.315659912895285,
+      "grad_norm": 0.4704832434654236,
+      "learning_rate": 6.737360348418861e-06,
+      "loss": 0.1016,
+      "step": 35020
+    },
+    {
+      "epoch": 3.3166067032758946,
+      "grad_norm": 0.4528552293777466,
+      "learning_rate": 6.733573186896422e-06,
+      "loss": 0.1003,
+      "step": 35030
+    },
+    {
+      "epoch": 3.3175534936565043,
+      "grad_norm": 0.44715917110443115,
+      "learning_rate": 6.729786025373984e-06,
+      "loss": 0.0942,
+      "step": 35040
+    },
+    {
+      "epoch": 3.318500284037114,
+      "grad_norm": 0.4122602641582489,
+      "learning_rate": 6.725998863851544e-06,
+      "loss": 0.0961,
+      "step": 35050
+    },
+    {
+      "epoch": 3.319447074417724,
+      "grad_norm": 0.4888918101787567,
+      "learning_rate": 6.7222117023291046e-06,
+      "loss": 0.1032,
+      "step": 35060
+    },
+    {
+      "epoch": 3.3203938647983335,
+      "grad_norm": 0.527289092540741,
+      "learning_rate": 6.718424540806666e-06,
+      "loss": 0.0961,
+      "step": 35070
+    },
+    {
+      "epoch": 3.3213406551789433,
+      "grad_norm": 0.47067925333976746,
+      "learning_rate": 6.714637379284226e-06,
+      "loss": 0.0933,
+      "step": 35080
+    },
+    {
+      "epoch": 3.322287445559553,
+      "grad_norm": 0.4370437562465668,
+      "learning_rate": 6.710850217761788e-06,
+      "loss": 0.0973,
+      "step": 35090
+    },
+    {
+      "epoch": 3.3232342359401628,
+      "grad_norm": 0.41316908597946167,
+      "learning_rate": 6.707063056239349e-06,
+      "loss": 0.0961,
+      "step": 35100
+    },
+    {
+      "epoch": 3.3241810263207725,
+      "grad_norm": 0.4926522672176361,
+      "learning_rate": 6.703275894716911e-06,
+      "loss": 0.1046,
+      "step": 35110
+    },
+    {
+      "epoch": 3.3251278167013822,
+      "grad_norm": 0.5030735731124878,
+      "learning_rate": 6.699488733194472e-06,
+      "loss": 0.0908,
+      "step": 35120
+    },
+    {
+      "epoch": 3.326074607081992,
+      "grad_norm": 0.4518188238143921,
+      "learning_rate": 6.695701571672032e-06,
+      "loss": 0.0955,
+      "step": 35130
+    },
+    {
+      "epoch": 3.3270213974626017,
+      "grad_norm": 0.43322181701660156,
+      "learning_rate": 6.6919144101495934e-06,
+      "loss": 0.0939,
+      "step": 35140
+    },
+    {
+      "epoch": 3.3279681878432115,
+      "grad_norm": 0.4343359172344208,
+      "learning_rate": 6.688127248627154e-06,
+      "loss": 0.0971,
+      "step": 35150
+    },
+    {
+      "epoch": 3.328914978223821,
+      "grad_norm": 0.48641207814216614,
+      "learning_rate": 6.684340087104716e-06,
+      "loss": 0.0996,
+      "step": 35160
+    },
+    {
+      "epoch": 3.329861768604431,
+      "grad_norm": 0.5804575085639954,
+      "learning_rate": 6.680552925582276e-06,
+      "loss": 0.0993,
+      "step": 35170
+    },
+    {
+      "epoch": 3.3308085589850407,
+      "grad_norm": 0.6237295866012573,
+      "learning_rate": 6.676765764059838e-06,
+      "loss": 0.0888,
+      "step": 35180
+    },
+    {
+      "epoch": 3.3317553493656504,
+      "grad_norm": 0.44823551177978516,
+      "learning_rate": 6.672978602537399e-06,
+      "loss": 0.0896,
+      "step": 35190
+    },
+    {
+      "epoch": 3.33270213974626,
+      "grad_norm": 0.617464005947113,
+      "learning_rate": 6.66919144101496e-06,
+      "loss": 0.1059,
+      "step": 35200
+    },
+    {
+      "epoch": 3.33364893012687,
+      "grad_norm": 0.5996119379997253,
+      "learning_rate": 6.665404279492521e-06,
+      "loss": 0.0877,
+      "step": 35210
+    },
+    {
+      "epoch": 3.3345957205074797,
+      "grad_norm": 0.6480814814567566,
+      "learning_rate": 6.6616171179700815e-06,
+      "loss": 0.1111,
+      "step": 35220
+    },
+    {
+      "epoch": 3.3355425108880894,
+      "grad_norm": 0.5987730622291565,
+      "learning_rate": 6.657829956447643e-06,
+      "loss": 0.0859,
+      "step": 35230
+    },
+    {
+      "epoch": 3.336489301268699,
+      "grad_norm": 0.5056741237640381,
+      "learning_rate": 6.654042794925204e-06,
+      "loss": 0.0893,
+      "step": 35240
+    },
+    {
+      "epoch": 3.337436091649309,
+      "grad_norm": 0.5164422988891602,
+      "learning_rate": 6.650255633402766e-06,
+      "loss": 0.0987,
+      "step": 35250
+    },
+    {
+      "epoch": 3.3383828820299186,
+      "grad_norm": 0.4121684432029724,
+      "learning_rate": 6.646468471880326e-06,
+      "loss": 0.0892,
+      "step": 35260
+    },
+    {
+      "epoch": 3.3393296724105284,
+      "grad_norm": 0.47107455134391785,
+      "learning_rate": 6.642681310357887e-06,
+      "loss": 0.0872,
+      "step": 35270
+    },
+    {
+      "epoch": 3.340276462791138,
+      "grad_norm": 0.5224701166152954,
+      "learning_rate": 6.6388941488354485e-06,
+      "loss": 0.1134,
+      "step": 35280
+    },
+    {
+      "epoch": 3.341223253171748,
+      "grad_norm": 0.45090168714523315,
+      "learning_rate": 6.6351069873130094e-06,
+      "loss": 0.093,
+      "step": 35290
+    },
+    {
+      "epoch": 3.3421700435523576,
+      "grad_norm": 0.2751280963420868,
+      "learning_rate": 6.631319825790571e-06,
+      "loss": 0.082,
+      "step": 35300
+    },
+    {
+      "epoch": 3.3431168339329673,
+      "grad_norm": 0.5314486622810364,
+      "learning_rate": 6.627532664268131e-06,
+      "loss": 0.0954,
+      "step": 35310
+    },
+    {
+      "epoch": 3.344063624313577,
+      "grad_norm": 0.4997088611125946,
+      "learning_rate": 6.623745502745693e-06,
+      "loss": 0.093,
+      "step": 35320
+    },
+    {
+      "epoch": 3.345010414694187,
+      "grad_norm": 0.40355512499809265,
+      "learning_rate": 6.619958341223254e-06,
+      "loss": 0.0949,
+      "step": 35330
+    },
+    {
+      "epoch": 3.3459572050747965,
+      "grad_norm": 0.4665563106536865,
+      "learning_rate": 6.616171179700815e-06,
+      "loss": 0.0909,
+      "step": 35340
+    },
+    {
+      "epoch": 3.3469039954554063,
+      "grad_norm": 0.4175027012825012,
+      "learning_rate": 6.612384018178376e-06,
+      "loss": 0.0892,
+      "step": 35350
+    },
+    {
+      "epoch": 3.347850785836016,
+      "grad_norm": 0.5048304796218872,
+      "learning_rate": 6.6085968566559366e-06,
+      "loss": 0.0952,
+      "step": 35360
+    },
+    {
+      "epoch": 3.3487975762166258,
+      "grad_norm": 0.3938494026660919,
+      "learning_rate": 6.604809695133498e-06,
+      "loss": 0.0874,
+      "step": 35370
+    },
+    {
+      "epoch": 3.3497443665972355,
+      "grad_norm": 0.4948226809501648,
+      "learning_rate": 6.601022533611059e-06,
+      "loss": 0.1051,
+      "step": 35380
+    },
+    {
+      "epoch": 3.3506911569778453,
+      "grad_norm": 0.4628227651119232,
+      "learning_rate": 6.597235372088621e-06,
+      "loss": 0.1034,
+      "step": 35390
+    },
+    {
+      "epoch": 3.351637947358455,
+      "grad_norm": 0.4768291711807251,
+      "learning_rate": 6.593448210566181e-06,
+      "loss": 0.0944,
+      "step": 35400
+    },
+    {
+      "epoch": 3.3525847377390647,
+      "grad_norm": 0.5109726190567017,
+      "learning_rate": 6.589661049043742e-06,
+      "loss": 0.098,
+      "step": 35410
+    },
+    {
+      "epoch": 3.3535315281196745,
+      "grad_norm": 0.49673202633857727,
+      "learning_rate": 6.585873887521304e-06,
+      "loss": 0.0965,
+      "step": 35420
+    },
+    {
+      "epoch": 3.354478318500284,
+      "grad_norm": 0.46202951669692993,
+      "learning_rate": 6.5820867259988645e-06,
+      "loss": 0.0868,
+      "step": 35430
+    },
+    {
+      "epoch": 3.355425108880894,
+      "grad_norm": 0.5012629628181458,
+      "learning_rate": 6.578299564476425e-06,
+      "loss": 0.1084,
+      "step": 35440
+    },
+    {
+      "epoch": 3.3563718992615037,
+      "grad_norm": 0.44407936930656433,
+      "learning_rate": 6.574512402953986e-06,
+      "loss": 0.0925,
+      "step": 35450
+    },
+    {
+      "epoch": 3.3573186896421134,
+      "grad_norm": 0.6375394463539124,
+      "learning_rate": 6.570725241431548e-06,
+      "loss": 0.1051,
+      "step": 35460
+    },
+    {
+      "epoch": 3.358265480022723,
+      "grad_norm": 0.4182000756263733,
+      "learning_rate": 6.566938079909109e-06,
+      "loss": 0.1004,
+      "step": 35470
+    },
+    {
+      "epoch": 3.359212270403333,
+      "grad_norm": 0.40456700325012207,
+      "learning_rate": 6.563150918386669e-06,
+      "loss": 0.0903,
+      "step": 35480
+    },
+    {
+      "epoch": 3.3601590607839427,
+      "grad_norm": 0.47193360328674316,
+      "learning_rate": 6.559363756864231e-06,
+      "loss": 0.1049,
+      "step": 35490
+    },
+    {
+      "epoch": 3.3611058511645524,
+      "grad_norm": 0.44254809617996216,
+      "learning_rate": 6.555576595341792e-06,
+      "loss": 0.0924,
+      "step": 35500
+    },
+    {
+      "epoch": 3.362052641545162,
+      "grad_norm": 0.44690197706222534,
+      "learning_rate": 6.551789433819353e-06,
+      "loss": 0.1009,
+      "step": 35510
+    },
+    {
+      "epoch": 3.3629994319257714,
+      "grad_norm": 0.32038065791130066,
+      "learning_rate": 6.548002272296914e-06,
+      "loss": 0.0964,
+      "step": 35520
+    },
+    {
+      "epoch": 3.363946222306381,
+      "grad_norm": 0.42337891459465027,
+      "learning_rate": 6.544215110774475e-06,
+      "loss": 0.0983,
+      "step": 35530
+    },
+    {
+      "epoch": 3.364893012686991,
+      "grad_norm": 0.45714786648750305,
+      "learning_rate": 6.540427949252036e-06,
+      "loss": 0.0878,
+      "step": 35540
+    },
+    {
+      "epoch": 3.3658398030676007,
+      "grad_norm": 0.4319290220737457,
+      "learning_rate": 6.536640787729597e-06,
+      "loss": 0.0933,
+      "step": 35550
+    },
+    {
+      "epoch": 3.3667865934482104,
+      "grad_norm": 0.5740752816200256,
+      "learning_rate": 6.532853626207159e-06,
+      "loss": 0.0961,
+      "step": 35560
+    },
+    {
+      "epoch": 3.36773338382882,
+      "grad_norm": 0.5247994661331177,
+      "learning_rate": 6.529066464684719e-06,
+      "loss": 0.0912,
+      "step": 35570
+    },
+    {
+      "epoch": 3.36868017420943,
+      "grad_norm": 0.5578654408454895,
+      "learning_rate": 6.5252793031622805e-06,
+      "loss": 0.0926,
+      "step": 35580
+    },
+    {
+      "epoch": 3.3696269645900396,
+      "grad_norm": 0.4753609895706177,
+      "learning_rate": 6.521492141639841e-06,
+      "loss": 0.0999,
+      "step": 35590
+    },
+    {
+      "epoch": 3.3705737549706494,
+      "grad_norm": 0.4210730195045471,
+      "learning_rate": 6.517704980117403e-06,
+      "loss": 0.0935,
+      "step": 35600
+    },
+    {
+      "epoch": 3.371520545351259,
+      "grad_norm": 0.5239378809928894,
+      "learning_rate": 6.513917818594964e-06,
+      "loss": 0.0998,
+      "step": 35610
+    },
+    {
+      "epoch": 3.372467335731869,
+      "grad_norm": 0.554095447063446,
+      "learning_rate": 6.510130657072524e-06,
+      "loss": 0.0968,
+      "step": 35620
+    },
+    {
+      "epoch": 3.3734141261124786,
+      "grad_norm": 0.5286445617675781,
+      "learning_rate": 6.506343495550086e-06,
+      "loss": 0.0887,
+      "step": 35630
+    },
+    {
+      "epoch": 3.3743609164930883,
+      "grad_norm": 0.4211771488189697,
+      "learning_rate": 6.502556334027647e-06,
+      "loss": 0.0887,
+      "step": 35640
+    },
+    {
+      "epoch": 3.375307706873698,
+      "grad_norm": 0.4598631262779236,
+      "learning_rate": 6.4987691725052085e-06,
+      "loss": 0.095,
+      "step": 35650
+    },
+    {
+      "epoch": 3.376254497254308,
+      "grad_norm": 0.4372754991054535,
+      "learning_rate": 6.4949820109827685e-06,
+      "loss": 0.0942,
+      "step": 35660
+    },
+    {
+      "epoch": 3.3772012876349176,
+      "grad_norm": 0.5444381833076477,
+      "learning_rate": 6.49119484946033e-06,
+      "loss": 0.09,
+      "step": 35670
+    },
+    {
+      "epoch": 3.3781480780155273,
+      "grad_norm": 0.597549557685852,
+      "learning_rate": 6.487407687937891e-06,
+      "loss": 0.0895,
+      "step": 35680
+    },
+    {
+      "epoch": 3.379094868396137,
+      "grad_norm": 0.4873766005039215,
+      "learning_rate": 6.483620526415452e-06,
+      "loss": 0.0912,
+      "step": 35690
+    },
+    {
+      "epoch": 3.380041658776747,
+      "grad_norm": 0.4911908507347107,
+      "learning_rate": 6.479833364893014e-06,
+      "loss": 0.1113,
+      "step": 35700
+    },
+    {
+      "epoch": 3.3809884491573565,
+      "grad_norm": 0.5038026571273804,
+      "learning_rate": 6.476046203370574e-06,
+      "loss": 0.0958,
+      "step": 35710
+    },
+    {
+      "epoch": 3.3819352395379663,
+      "grad_norm": 0.4452175498008728,
+      "learning_rate": 6.472259041848136e-06,
+      "loss": 0.0898,
+      "step": 35720
+    },
+    {
+      "epoch": 3.382882029918576,
+      "grad_norm": 0.4882689416408539,
+      "learning_rate": 6.4684718803256965e-06,
+      "loss": 0.0961,
+      "step": 35730
+    },
+    {
+      "epoch": 3.3838288202991857,
+      "grad_norm": 0.4825945794582367,
+      "learning_rate": 6.464684718803258e-06,
+      "loss": 0.0926,
+      "step": 35740
+    },
+    {
+      "epoch": 3.3847756106797955,
+      "grad_norm": 0.4384516477584839,
+      "learning_rate": 6.460897557280818e-06,
+      "loss": 0.0864,
+      "step": 35750
+    },
+    {
+      "epoch": 3.3857224010604052,
+      "grad_norm": 0.5509099960327148,
+      "learning_rate": 6.457110395758379e-06,
+      "loss": 0.1007,
+      "step": 35760
+    },
+    {
+      "epoch": 3.386669191441015,
+      "grad_norm": 0.6205552220344543,
+      "learning_rate": 6.453323234235941e-06,
+      "loss": 0.0955,
+      "step": 35770
+    },
+    {
+      "epoch": 3.3876159818216247,
+      "grad_norm": 0.42575883865356445,
+      "learning_rate": 6.449536072713502e-06,
+      "loss": 0.0985,
+      "step": 35780
+    },
+    {
+      "epoch": 3.3885627722022345,
+      "grad_norm": 0.5736665725708008,
+      "learning_rate": 6.4457489111910636e-06,
+      "loss": 0.0954,
+      "step": 35790
+    },
+    {
+      "epoch": 3.389509562582844,
+      "grad_norm": 0.416550874710083,
+      "learning_rate": 6.441961749668624e-06,
+      "loss": 0.0929,
+      "step": 35800
+    },
+    {
+      "epoch": 3.390456352963454,
+      "grad_norm": 0.4721170961856842,
+      "learning_rate": 6.438174588146185e-06,
+      "loss": 0.0939,
+      "step": 35810
+    },
+    {
+      "epoch": 3.3914031433440637,
+      "grad_norm": 0.44223690032958984,
+      "learning_rate": 6.434387426623746e-06,
+      "loss": 0.0905,
+      "step": 35820
+    },
+    {
+      "epoch": 3.3923499337246734,
+      "grad_norm": 0.48436668515205383,
+      "learning_rate": 6.430600265101306e-06,
+      "loss": 0.1023,
+      "step": 35830
+    },
+    {
+      "epoch": 3.393296724105283,
+      "grad_norm": 0.48289725184440613,
+      "learning_rate": 6.426813103578868e-06,
+      "loss": 0.096,
+      "step": 35840
+    },
+    {
+      "epoch": 3.394243514485893,
+      "grad_norm": 0.5580629110336304,
+      "learning_rate": 6.423025942056429e-06,
+      "loss": 0.1035,
+      "step": 35850
+    },
+    {
+      "epoch": 3.3951903048665026,
+      "grad_norm": 0.33182501792907715,
+      "learning_rate": 6.419238780533991e-06,
+      "loss": 0.0972,
+      "step": 35860
+    },
+    {
+      "epoch": 3.3961370952471124,
+      "grad_norm": 0.5468883514404297,
+      "learning_rate": 6.415451619011552e-06,
+      "loss": 0.1044,
+      "step": 35870
+    },
+    {
+      "epoch": 3.397083885627722,
+      "grad_norm": 0.4226824939250946,
+      "learning_rate": 6.411664457489113e-06,
+      "loss": 0.0929,
+      "step": 35880
+    },
+    {
+      "epoch": 3.398030676008332,
+      "grad_norm": 0.46152549982070923,
+      "learning_rate": 6.407877295966673e-06,
+      "loss": 0.1023,
+      "step": 35890
+    },
+    {
+      "epoch": 3.3989774663889416,
+      "grad_norm": 0.5353741645812988,
+      "learning_rate": 6.404090134444234e-06,
+      "loss": 0.1086,
+      "step": 35900
+    },
+    {
+      "epoch": 3.3999242567695513,
+      "grad_norm": 0.43531525135040283,
+      "learning_rate": 6.400302972921796e-06,
+      "loss": 0.1058,
+      "step": 35910
+    },
+    {
+      "epoch": 3.400871047150161,
+      "grad_norm": 0.5732455253601074,
+      "learning_rate": 6.396515811399356e-06,
+      "loss": 0.0953,
+      "step": 35920
+    },
+    {
+      "epoch": 3.401817837530771,
+      "grad_norm": 0.453078955411911,
+      "learning_rate": 6.392728649876918e-06,
+      "loss": 0.1015,
+      "step": 35930
+    },
+    {
+      "epoch": 3.4027646279113806,
+      "grad_norm": 0.48207634687423706,
+      "learning_rate": 6.388941488354479e-06,
+      "loss": 0.1028,
+      "step": 35940
+    },
+    {
+      "epoch": 3.4037114182919903,
+      "grad_norm": 0.5186272263526917,
+      "learning_rate": 6.3851543268320405e-06,
+      "loss": 0.0925,
+      "step": 35950
+    },
+    {
+      "epoch": 3.4046582086726,
+      "grad_norm": 0.5188668370246887,
+      "learning_rate": 6.381367165309601e-06,
+      "loss": 0.1047,
+      "step": 35960
+    },
+    {
+      "epoch": 3.40560499905321,
+      "grad_norm": 0.42555683851242065,
+      "learning_rate": 6.377580003787161e-06,
+      "loss": 0.1046,
+      "step": 35970
+    },
+    {
+      "epoch": 3.4065517894338195,
+      "grad_norm": 0.615644633769989,
+      "learning_rate": 6.373792842264723e-06,
+      "loss": 0.1068,
+      "step": 35980
+    },
+    {
+      "epoch": 3.4074985798144293,
+      "grad_norm": 0.5070286393165588,
+      "learning_rate": 6.370005680742284e-06,
+      "loss": 0.0975,
+      "step": 35990
+    },
+    {
+      "epoch": 3.4084453701950386,
+      "grad_norm": 0.5518382787704468,
+      "learning_rate": 6.366218519219846e-06,
+      "loss": 0.099,
+      "step": 36000
+    },
+    {
+      "epoch": 3.4093921605756483,
+      "grad_norm": 0.45891568064689636,
+      "learning_rate": 6.362431357697406e-06,
+      "loss": 0.0898,
+      "step": 36010
+    },
+    {
+      "epoch": 3.410338950956258,
+      "grad_norm": 0.5730379223823547,
+      "learning_rate": 6.358644196174968e-06,
+      "loss": 0.0962,
+      "step": 36020
+    },
+    {
+      "epoch": 3.411285741336868,
+      "grad_norm": 0.3990994691848755,
+      "learning_rate": 6.3548570346525285e-06,
+      "loss": 0.0846,
+      "step": 36030
+    },
+    {
+      "epoch": 3.4122325317174775,
+      "grad_norm": 0.5104419589042664,
+      "learning_rate": 6.351069873130089e-06,
+      "loss": 0.1066,
+      "step": 36040
+    },
+    {
+      "epoch": 3.4131793220980873,
+      "grad_norm": 0.46190160512924194,
+      "learning_rate": 6.347282711607651e-06,
+      "loss": 0.0952,
+      "step": 36050
+    },
+    {
+      "epoch": 3.414126112478697,
+      "grad_norm": 0.4073682129383087,
+      "learning_rate": 6.343495550085211e-06,
+      "loss": 0.0966,
+      "step": 36060
+    },
+    {
+      "epoch": 3.4150729028593068,
+      "grad_norm": 0.5171728730201721,
+      "learning_rate": 6.339708388562773e-06,
+      "loss": 0.1023,
+      "step": 36070
+    },
+    {
+      "epoch": 3.4160196932399165,
+      "grad_norm": 0.4601489305496216,
+      "learning_rate": 6.335921227040334e-06,
+      "loss": 0.0856,
+      "step": 36080
+    },
+    {
+      "epoch": 3.4169664836205262,
+      "grad_norm": 0.4700734317302704,
+      "learning_rate": 6.3321340655178956e-06,
+      "loss": 0.0939,
+      "step": 36090
+    },
+    {
+      "epoch": 3.417913274001136,
+      "grad_norm": 0.5556718111038208,
+      "learning_rate": 6.328346903995456e-06,
+      "loss": 0.084,
+      "step": 36100
+    },
+    {
+      "epoch": 3.4188600643817457,
+      "grad_norm": 0.5419147610664368,
+      "learning_rate": 6.3245597424730165e-06,
+      "loss": 0.0897,
+      "step": 36110
+    },
+    {
+      "epoch": 3.4198068547623555,
+      "grad_norm": 0.5299506783485413,
+      "learning_rate": 6.320772580950578e-06,
+      "loss": 0.0937,
+      "step": 36120
+    },
+    {
+      "epoch": 3.420753645142965,
+      "grad_norm": 0.4458409249782562,
+      "learning_rate": 6.316985419428139e-06,
+      "loss": 0.0833,
+      "step": 36130
+    },
+    {
+      "epoch": 3.421700435523575,
+      "grad_norm": 0.35562217235565186,
+      "learning_rate": 6.313198257905701e-06,
+      "loss": 0.1003,
+      "step": 36140
+    },
+    {
+      "epoch": 3.4226472259041847,
+      "grad_norm": 0.40143030881881714,
+      "learning_rate": 6.309411096383261e-06,
+      "loss": 0.0882,
+      "step": 36150
+    },
+    {
+      "epoch": 3.4235940162847944,
+      "grad_norm": 0.479168176651001,
+      "learning_rate": 6.305623934860823e-06,
+      "loss": 0.0951,
+      "step": 36160
+    },
+    {
+      "epoch": 3.424540806665404,
+      "grad_norm": 0.5207576751708984,
+      "learning_rate": 6.3018367733383836e-06,
+      "loss": 0.1059,
+      "step": 36170
+    },
+    {
+      "epoch": 3.425487597046014,
+      "grad_norm": 0.46082577109336853,
+      "learning_rate": 6.298049611815944e-06,
+      "loss": 0.0926,
+      "step": 36180
+    },
+    {
+      "epoch": 3.4264343874266237,
+      "grad_norm": 0.45255082845687866,
+      "learning_rate": 6.294262450293505e-06,
+      "loss": 0.0921,
+      "step": 36190
+    },
+    {
+      "epoch": 3.4273811778072334,
+      "grad_norm": 0.4582348167896271,
+      "learning_rate": 6.290475288771066e-06,
+      "loss": 0.1018,
+      "step": 36200
+    },
+    {
+      "epoch": 3.428327968187843,
+      "grad_norm": 0.5190147757530212,
+      "learning_rate": 6.286688127248628e-06,
+      "loss": 0.0997,
+      "step": 36210
+    },
+    {
+      "epoch": 3.429274758568453,
+      "grad_norm": 0.5631082653999329,
+      "learning_rate": 6.282900965726189e-06,
+      "loss": 0.094,
+      "step": 36220
+    },
+    {
+      "epoch": 3.4302215489490626,
+      "grad_norm": 0.45488786697387695,
+      "learning_rate": 6.279113804203751e-06,
+      "loss": 0.0881,
+      "step": 36230
+    },
+    {
+      "epoch": 3.4311683393296724,
+      "grad_norm": 0.4592917263507843,
+      "learning_rate": 6.275326642681311e-06,
+      "loss": 0.1066,
+      "step": 36240
+    },
+    {
+      "epoch": 3.432115129710282,
+      "grad_norm": 0.49670347571372986,
+      "learning_rate": 6.271539481158872e-06,
+      "loss": 0.0913,
+      "step": 36250
+    },
+    {
+      "epoch": 3.433061920090892,
+      "grad_norm": 0.5354011654853821,
+      "learning_rate": 6.267752319636433e-06,
+      "loss": 0.0948,
+      "step": 36260
+    },
+    {
+      "epoch": 3.4340087104715016,
+      "grad_norm": 0.5139631628990173,
+      "learning_rate": 6.263965158113993e-06,
+      "loss": 0.1055,
+      "step": 36270
+    },
+    {
+      "epoch": 3.4349555008521113,
+      "grad_norm": 0.5732100009918213,
+      "learning_rate": 6.260177996591555e-06,
+      "loss": 0.1018,
+      "step": 36280
+    },
+    {
+      "epoch": 3.435902291232721,
+      "grad_norm": 0.6240257620811462,
+      "learning_rate": 6.256390835069116e-06,
+      "loss": 0.1047,
+      "step": 36290
+    },
+    {
+      "epoch": 3.436849081613331,
+      "grad_norm": 0.4386384189128876,
+      "learning_rate": 6.252603673546678e-06,
+      "loss": 0.1011,
+      "step": 36300
+    },
+    {
+      "epoch": 3.4377958719939405,
+      "grad_norm": 0.4700325131416321,
+      "learning_rate": 6.248816512024239e-06,
+      "loss": 0.0957,
+      "step": 36310
+    },
+    {
+      "epoch": 3.4387426623745503,
+      "grad_norm": 0.41461843252182007,
+      "learning_rate": 6.245029350501799e-06,
+      "loss": 0.096,
+      "step": 36320
+    },
+    {
+      "epoch": 3.43968945275516,
+      "grad_norm": 0.4554722309112549,
+      "learning_rate": 6.2412421889793605e-06,
+      "loss": 0.0852,
+      "step": 36330
+    },
+    {
+      "epoch": 3.4406362431357698,
+      "grad_norm": 0.6045703887939453,
+      "learning_rate": 6.237455027456921e-06,
+      "loss": 0.0948,
+      "step": 36340
+    },
+    {
+      "epoch": 3.4415830335163795,
+      "grad_norm": 0.48058703541755676,
+      "learning_rate": 6.233667865934483e-06,
+      "loss": 0.094,
+      "step": 36350
+    },
+    {
+      "epoch": 3.4425298238969892,
+      "grad_norm": 0.4483596086502075,
+      "learning_rate": 6.229880704412043e-06,
+      "loss": 0.0971,
+      "step": 36360
+    },
+    {
+      "epoch": 3.443476614277599,
+      "grad_norm": 0.43034300208091736,
+      "learning_rate": 6.226093542889605e-06,
+      "loss": 0.1038,
+      "step": 36370
+    },
+    {
+      "epoch": 3.4444234046582087,
+      "grad_norm": 0.6224856972694397,
+      "learning_rate": 6.222306381367166e-06,
+      "loss": 0.1106,
+      "step": 36380
+    },
+    {
+      "epoch": 3.4453701950388185,
+      "grad_norm": 0.4863092005252838,
+      "learning_rate": 6.218519219844727e-06,
+      "loss": 0.0938,
+      "step": 36390
+    },
+    {
+      "epoch": 3.446316985419428,
+      "grad_norm": 0.38047119975090027,
+      "learning_rate": 6.2147320583222884e-06,
+      "loss": 0.0935,
+      "step": 36400
+    },
+    {
+      "epoch": 3.447263775800038,
+      "grad_norm": 0.5247942209243774,
+      "learning_rate": 6.2109448967998485e-06,
+      "loss": 0.0943,
+      "step": 36410
+    },
+    {
+      "epoch": 3.4482105661806477,
+      "grad_norm": 0.5480045080184937,
+      "learning_rate": 6.20715773527741e-06,
+      "loss": 0.098,
+      "step": 36420
+    },
+    {
+      "epoch": 3.4491573565612574,
+      "grad_norm": 0.4587515890598297,
+      "learning_rate": 6.203370573754971e-06,
+      "loss": 0.0984,
+      "step": 36430
+    },
+    {
+      "epoch": 3.450104146941867,
+      "grad_norm": 0.48768389225006104,
+      "learning_rate": 6.199583412232533e-06,
+      "loss": 0.1042,
+      "step": 36440
+    },
+    {
+      "epoch": 3.451050937322477,
+      "grad_norm": 0.4601198434829712,
+      "learning_rate": 6.195796250710093e-06,
+      "loss": 0.0974,
+      "step": 36450
+    },
+    {
+      "epoch": 3.4519977277030867,
+      "grad_norm": 0.48066088557243347,
+      "learning_rate": 6.192009089187654e-06,
+      "loss": 0.0966,
+      "step": 36460
+    },
+    {
+      "epoch": 3.4529445180836964,
+      "grad_norm": 0.42125844955444336,
+      "learning_rate": 6.1882219276652155e-06,
+      "loss": 0.0948,
+      "step": 36470
+    },
+    {
+      "epoch": 3.453891308464306,
+      "grad_norm": 0.5926153659820557,
+      "learning_rate": 6.1844347661427764e-06,
+      "loss": 0.0921,
+      "step": 36480
+    },
+    {
+      "epoch": 3.454838098844916,
+      "grad_norm": 0.5392172336578369,
+      "learning_rate": 6.180647604620338e-06,
+      "loss": 0.0964,
+      "step": 36490
+    },
+    {
+      "epoch": 3.4557848892255256,
+      "grad_norm": 0.4178329110145569,
+      "learning_rate": 6.176860443097898e-06,
+      "loss": 0.103,
+      "step": 36500
+    },
+    {
+      "epoch": 3.4567316796061354,
+      "grad_norm": 0.45059117674827576,
+      "learning_rate": 6.17307328157546e-06,
+      "loss": 0.0995,
+      "step": 36510
+    },
+    {
+      "epoch": 3.457678469986745,
+      "grad_norm": 0.5537148118019104,
+      "learning_rate": 6.169286120053021e-06,
+      "loss": 0.0894,
+      "step": 36520
+    },
+    {
+      "epoch": 3.458625260367355,
+      "grad_norm": 0.469635933637619,
+      "learning_rate": 6.165498958530581e-06,
+      "loss": 0.0955,
+      "step": 36530
+    },
+    {
+      "epoch": 3.4595720507479646,
+      "grad_norm": 0.4903915822505951,
+      "learning_rate": 6.161711797008143e-06,
+      "loss": 0.0952,
+      "step": 36540
+    },
+    {
+      "epoch": 3.4605188411285743,
+      "grad_norm": 0.4877898097038269,
+      "learning_rate": 6.1579246354857036e-06,
+      "loss": 0.0915,
+      "step": 36550
+    },
+    {
+      "epoch": 3.461465631509184,
+      "grad_norm": 0.4818441867828369,
+      "learning_rate": 6.154137473963265e-06,
+      "loss": 0.0957,
+      "step": 36560
+    },
+    {
+      "epoch": 3.462412421889794,
+      "grad_norm": 0.4799627661705017,
+      "learning_rate": 6.150350312440826e-06,
+      "loss": 0.1026,
+      "step": 36570
+    },
+    {
+      "epoch": 3.4633592122704036,
+      "grad_norm": 0.4602954685688019,
+      "learning_rate": 6.146563150918388e-06,
+      "loss": 0.0894,
+      "step": 36580
+    },
+    {
+      "epoch": 3.4643060026510133,
+      "grad_norm": 0.4297948479652405,
+      "learning_rate": 6.142775989395948e-06,
+      "loss": 0.0999,
+      "step": 36590
+    },
+    {
+      "epoch": 3.465252793031623,
+      "grad_norm": 0.4892706871032715,
+      "learning_rate": 6.138988827873509e-06,
+      "loss": 0.0953,
+      "step": 36600
+    },
+    {
+      "epoch": 3.4661995834122328,
+      "grad_norm": 0.5648135542869568,
+      "learning_rate": 6.135201666351071e-06,
+      "loss": 0.1001,
+      "step": 36610
+    },
+    {
+      "epoch": 3.467146373792842,
+      "grad_norm": 0.4892693758010864,
+      "learning_rate": 6.131414504828631e-06,
+      "loss": 0.0975,
+      "step": 36620
+    },
+    {
+      "epoch": 3.468093164173452,
+      "grad_norm": 0.4457409679889679,
+      "learning_rate": 6.1276273433061924e-06,
+      "loss": 0.0948,
+      "step": 36630
+    },
+    {
+      "epoch": 3.4690399545540616,
+      "grad_norm": 0.4559725821018219,
+      "learning_rate": 6.123840181783753e-06,
+      "loss": 0.0892,
+      "step": 36640
+    },
+    {
+      "epoch": 3.4699867449346713,
+      "grad_norm": 0.5138303637504578,
+      "learning_rate": 6.120053020261315e-06,
+      "loss": 0.0914,
+      "step": 36650
+    },
+    {
+      "epoch": 3.470933535315281,
+      "grad_norm": 0.4566517472267151,
+      "learning_rate": 6.116265858738876e-06,
+      "loss": 0.0971,
+      "step": 36660
+    },
+    {
+      "epoch": 3.471880325695891,
+      "grad_norm": 0.4594644010066986,
+      "learning_rate": 6.112478697216436e-06,
+      "loss": 0.0996,
+      "step": 36670
+    },
+    {
+      "epoch": 3.4728271160765005,
+      "grad_norm": 0.46614035964012146,
+      "learning_rate": 6.108691535693998e-06,
+      "loss": 0.1154,
+      "step": 36680
+    },
+    {
+      "epoch": 3.4737739064571103,
+      "grad_norm": 0.4599017798900604,
+      "learning_rate": 6.104904374171559e-06,
+      "loss": 0.0995,
+      "step": 36690
+    },
+    {
+      "epoch": 3.47472069683772,
+      "grad_norm": 0.4371439814567566,
+      "learning_rate": 6.10111721264912e-06,
+      "loss": 0.0844,
+      "step": 36700
+    },
+    {
+      "epoch": 3.4756674872183297,
+      "grad_norm": 0.5069402456283569,
+      "learning_rate": 6.0973300511266805e-06,
+      "loss": 0.0952,
+      "step": 36710
+    },
+    {
+      "epoch": 3.4766142775989395,
+      "grad_norm": 0.4560225307941437,
+      "learning_rate": 6.093542889604242e-06,
+      "loss": 0.0983,
+      "step": 36720
+    },
+    {
+      "epoch": 3.4775610679795492,
+      "grad_norm": 0.4633190929889679,
+      "learning_rate": 6.089755728081803e-06,
+      "loss": 0.0921,
+      "step": 36730
+    },
+    {
+      "epoch": 3.478507858360159,
+      "grad_norm": 0.3294411301612854,
+      "learning_rate": 6.085968566559364e-06,
+      "loss": 0.0891,
+      "step": 36740
+    },
+    {
+      "epoch": 3.4794546487407687,
+      "grad_norm": 0.3967805504798889,
+      "learning_rate": 6.082181405036926e-06,
+      "loss": 0.0894,
+      "step": 36750
+    },
+    {
+      "epoch": 3.4804014391213784,
+      "grad_norm": 0.4573502242565155,
+      "learning_rate": 6.078394243514486e-06,
+      "loss": 0.1026,
+      "step": 36760
+    },
+    {
+      "epoch": 3.481348229501988,
+      "grad_norm": 0.46853652596473694,
+      "learning_rate": 6.0746070819920475e-06,
+      "loss": 0.0921,
+      "step": 36770
+    },
+    {
+      "epoch": 3.482295019882598,
+      "grad_norm": 0.5473374128341675,
+      "learning_rate": 6.070819920469608e-06,
+      "loss": 0.0936,
+      "step": 36780
+    },
+    {
+      "epoch": 3.4832418102632077,
+      "grad_norm": 0.3934822976589203,
+      "learning_rate": 6.06703275894717e-06,
+      "loss": 0.0878,
+      "step": 36790
+    },
+    {
+      "epoch": 3.4841886006438174,
+      "grad_norm": 0.4433002173900604,
+      "learning_rate": 6.06324559742473e-06,
+      "loss": 0.0899,
+      "step": 36800
+    },
+    {
+      "epoch": 3.485135391024427,
+      "grad_norm": 0.5189225673675537,
+      "learning_rate": 6.059458435902291e-06,
+      "loss": 0.1038,
+      "step": 36810
+    },
+    {
+      "epoch": 3.486082181405037,
+      "grad_norm": 0.5562026500701904,
+      "learning_rate": 6.055671274379853e-06,
+      "loss": 0.0998,
+      "step": 36820
+    },
+    {
+      "epoch": 3.4870289717856466,
+      "grad_norm": 0.45497822761535645,
+      "learning_rate": 6.051884112857414e-06,
+      "loss": 0.1049,
+      "step": 36830
+    },
+    {
+      "epoch": 3.4879757621662564,
+      "grad_norm": 0.5216988921165466,
+      "learning_rate": 6.0480969513349755e-06,
+      "loss": 0.1033,
+      "step": 36840
+    },
+    {
+      "epoch": 3.488922552546866,
+      "grad_norm": 0.4457246959209442,
+      "learning_rate": 6.0443097898125355e-06,
+      "loss": 0.0898,
+      "step": 36850
+    },
+    {
+      "epoch": 3.489869342927476,
+      "grad_norm": 0.4475347697734833,
+      "learning_rate": 6.040522628290097e-06,
+      "loss": 0.1023,
+      "step": 36860
+    },
+    {
+      "epoch": 3.4908161333080856,
+      "grad_norm": 0.4940994381904602,
+      "learning_rate": 6.036735466767658e-06,
+      "loss": 0.0953,
+      "step": 36870
+    },
+    {
+      "epoch": 3.4917629236886953,
+      "grad_norm": 0.3915945291519165,
+      "learning_rate": 6.032948305245219e-06,
+      "loss": 0.0953,
+      "step": 36880
+    },
+    {
+      "epoch": 3.492709714069305,
+      "grad_norm": 0.5811683535575867,
+      "learning_rate": 6.02916114372278e-06,
+      "loss": 0.0955,
+      "step": 36890
+    },
+    {
+      "epoch": 3.493656504449915,
+      "grad_norm": 0.4573379158973694,
+      "learning_rate": 6.025373982200341e-06,
+      "loss": 0.0916,
+      "step": 36900
+    },
+    {
+      "epoch": 3.4946032948305246,
+      "grad_norm": 0.6304858922958374,
+      "learning_rate": 6.021586820677903e-06,
+      "loss": 0.0964,
+      "step": 36910
+    },
+    {
+      "epoch": 3.4955500852111343,
+      "grad_norm": 0.37654414772987366,
+      "learning_rate": 6.0177996591554635e-06,
+      "loss": 0.083,
+      "step": 36920
+    },
+    {
+      "epoch": 3.496496875591744,
+      "grad_norm": 0.5535511374473572,
+      "learning_rate": 6.014012497633025e-06,
+      "loss": 0.1011,
+      "step": 36930
+    },
+    {
+      "epoch": 3.497443665972354,
+      "grad_norm": 0.4015631675720215,
+      "learning_rate": 6.010225336110585e-06,
+      "loss": 0.087,
+      "step": 36940
+    },
+    {
+      "epoch": 3.4983904563529635,
+      "grad_norm": 0.5389937162399292,
+      "learning_rate": 6.006438174588146e-06,
+      "loss": 0.0943,
+      "step": 36950
+    },
+    {
+      "epoch": 3.4993372467335733,
+      "grad_norm": 0.48296236991882324,
+      "learning_rate": 6.002651013065708e-06,
+      "loss": 0.0984,
+      "step": 36960
+    },
+    {
+      "epoch": 3.500284037114183,
+      "grad_norm": 0.36838439106941223,
+      "learning_rate": 5.998863851543269e-06,
+      "loss": 0.0859,
+      "step": 36970
+    },
+    {
+      "epoch": 3.5012308274947928,
+      "grad_norm": 0.5506975054740906,
+      "learning_rate": 5.99507669002083e-06,
+      "loss": 0.1207,
+      "step": 36980
+    },
+    {
+      "epoch": 3.5021776178754025,
+      "grad_norm": 0.4506598114967346,
+      "learning_rate": 5.991289528498391e-06,
+      "loss": 0.0949,
+      "step": 36990
+    },
+    {
+      "epoch": 3.5031244082560122,
+      "grad_norm": 0.4446566104888916,
+      "learning_rate": 5.987502366975952e-06,
+      "loss": 0.1041,
+      "step": 37000
+    },
+    {
+      "epoch": 3.504071198636622,
+      "grad_norm": 0.43818366527557373,
+      "learning_rate": 5.983715205453513e-06,
+      "loss": 0.0986,
+      "step": 37010
+    },
+    {
+      "epoch": 3.5050179890172317,
+      "grad_norm": 0.4997681975364685,
+      "learning_rate": 5.979928043931073e-06,
+      "loss": 0.0918,
+      "step": 37020
+    },
+    {
+      "epoch": 3.5059647793978415,
+      "grad_norm": 0.5976042747497559,
+      "learning_rate": 5.976140882408635e-06,
+      "loss": 0.101,
+      "step": 37030
+    },
+    {
+      "epoch": 3.506911569778451,
+      "grad_norm": 0.43398550152778625,
+      "learning_rate": 5.972353720886196e-06,
+      "loss": 0.0949,
+      "step": 37040
+    },
+    {
+      "epoch": 3.507858360159061,
+      "grad_norm": 0.43026724457740784,
+      "learning_rate": 5.968566559363758e-06,
+      "loss": 0.0977,
+      "step": 37050
+    },
+    {
+      "epoch": 3.5088051505396707,
+      "grad_norm": 0.5097009539604187,
+      "learning_rate": 5.964779397841319e-06,
+      "loss": 0.0977,
+      "step": 37060
+    },
+    {
+      "epoch": 3.50975194092028,
+      "grad_norm": 0.36898836493492126,
+      "learning_rate": 5.9609922363188795e-06,
+      "loss": 0.0881,
+      "step": 37070
+    },
+    {
+      "epoch": 3.5106987313008897,
+      "grad_norm": 0.44765183329582214,
+      "learning_rate": 5.95720507479644e-06,
+      "loss": 0.0946,
+      "step": 37080
+    },
+    {
+      "epoch": 3.5116455216814995,
+      "grad_norm": 0.4629915654659271,
+      "learning_rate": 5.953417913274001e-06,
+      "loss": 0.0886,
+      "step": 37090
+    },
+    {
+      "epoch": 3.512592312062109,
+      "grad_norm": 0.43138405680656433,
+      "learning_rate": 5.949630751751563e-06,
+      "loss": 0.0945,
+      "step": 37100
+    },
+    {
+      "epoch": 3.513539102442719,
+      "grad_norm": 0.4578106105327606,
+      "learning_rate": 5.945843590229123e-06,
+      "loss": 0.1139,
+      "step": 37110
+    },
+    {
+      "epoch": 3.5144858928233287,
+      "grad_norm": 0.5025810599327087,
+      "learning_rate": 5.942056428706685e-06,
+      "loss": 0.1039,
+      "step": 37120
+    },
+    {
+      "epoch": 3.5154326832039384,
+      "grad_norm": 0.5416958332061768,
+      "learning_rate": 5.938269267184246e-06,
+      "loss": 0.0826,
+      "step": 37130
+    },
+    {
+      "epoch": 3.516379473584548,
+      "grad_norm": 0.4320995509624481,
+      "learning_rate": 5.9344821056618075e-06,
+      "loss": 0.1061,
+      "step": 37140
+    },
+    {
+      "epoch": 3.517326263965158,
+      "grad_norm": 0.4534364938735962,
+      "learning_rate": 5.930694944139368e-06,
+      "loss": 0.0876,
+      "step": 37150
+    },
+    {
+      "epoch": 3.5182730543457676,
+      "grad_norm": 0.5116265416145325,
+      "learning_rate": 5.926907782616928e-06,
+      "loss": 0.0907,
+      "step": 37160
+    },
+    {
+      "epoch": 3.5192198447263774,
+      "grad_norm": 0.582007646560669,
+      "learning_rate": 5.92312062109449e-06,
+      "loss": 0.1016,
+      "step": 37170
+    },
+    {
+      "epoch": 3.520166635106987,
+      "grad_norm": 0.5855867862701416,
+      "learning_rate": 5.919333459572051e-06,
+      "loss": 0.1076,
+      "step": 37180
+    },
+    {
+      "epoch": 3.521113425487597,
+      "grad_norm": 0.5251570343971252,
+      "learning_rate": 5.915546298049613e-06,
+      "loss": 0.0893,
+      "step": 37190
+    },
+    {
+      "epoch": 3.5220602158682066,
+      "grad_norm": 0.48873370885849,
+      "learning_rate": 5.911759136527173e-06,
+      "loss": 0.0926,
+      "step": 37200
+    },
+    {
+      "epoch": 3.5230070062488164,
+      "grad_norm": 0.4588128328323364,
+      "learning_rate": 5.907971975004735e-06,
+      "loss": 0.0877,
+      "step": 37210
+    },
+    {
+      "epoch": 3.523953796629426,
+      "grad_norm": 0.509213924407959,
+      "learning_rate": 5.9041848134822955e-06,
+      "loss": 0.0963,
+      "step": 37220
+    },
+    {
+      "epoch": 3.524900587010036,
+      "grad_norm": 0.4476909637451172,
+      "learning_rate": 5.900397651959856e-06,
+      "loss": 0.0957,
+      "step": 37230
+    },
+    {
+      "epoch": 3.5258473773906456,
+      "grad_norm": 0.4278342127799988,
+      "learning_rate": 5.896610490437418e-06,
+      "loss": 0.1033,
+      "step": 37240
+    },
+    {
+      "epoch": 3.5267941677712553,
+      "grad_norm": 0.41186821460723877,
+      "learning_rate": 5.892823328914978e-06,
+      "loss": 0.0943,
+      "step": 37250
+    },
+    {
+      "epoch": 3.527740958151865,
+      "grad_norm": 0.3761973977088928,
+      "learning_rate": 5.88903616739254e-06,
+      "loss": 0.1,
+      "step": 37260
+    },
+    {
+      "epoch": 3.528687748532475,
+      "grad_norm": 0.4176313281059265,
+      "learning_rate": 5.885249005870101e-06,
+      "loss": 0.0896,
+      "step": 37270
+    },
+    {
+      "epoch": 3.5296345389130845,
+      "grad_norm": 0.43153753876686096,
+      "learning_rate": 5.8814618443476626e-06,
+      "loss": 0.0949,
+      "step": 37280
+    },
+    {
+      "epoch": 3.5305813292936943,
+      "grad_norm": 0.49732184410095215,
+      "learning_rate": 5.877674682825223e-06,
+      "loss": 0.0931,
+      "step": 37290
+    },
+    {
+      "epoch": 3.531528119674304,
+      "grad_norm": 0.6734378933906555,
+      "learning_rate": 5.8738875213027835e-06,
+      "loss": 0.102,
+      "step": 37300
+    },
+    {
+      "epoch": 3.5324749100549138,
+      "grad_norm": 0.48934412002563477,
+      "learning_rate": 5.870100359780345e-06,
+      "loss": 0.0953,
+      "step": 37310
+    },
+    {
+      "epoch": 3.5334217004355235,
+      "grad_norm": 0.5085573196411133,
+      "learning_rate": 5.866313198257906e-06,
+      "loss": 0.1059,
+      "step": 37320
+    },
+    {
+      "epoch": 3.5343684908161332,
+      "grad_norm": 0.5069712996482849,
+      "learning_rate": 5.862526036735468e-06,
+      "loss": 0.0994,
+      "step": 37330
+    },
+    {
+      "epoch": 3.535315281196743,
+      "grad_norm": 0.5646600127220154,
+      "learning_rate": 5.858738875213028e-06,
+      "loss": 0.0982,
+      "step": 37340
+    },
+    {
+      "epoch": 3.5362620715773527,
+      "grad_norm": 0.3952558636665344,
+      "learning_rate": 5.85495171369059e-06,
+      "loss": 0.0928,
+      "step": 37350
+    },
+    {
+      "epoch": 3.5372088619579625,
+      "grad_norm": 0.45193013548851013,
+      "learning_rate": 5.851164552168151e-06,
+      "loss": 0.099,
+      "step": 37360
+    },
+    {
+      "epoch": 3.538155652338572,
+      "grad_norm": 0.493637353181839,
+      "learning_rate": 5.847377390645711e-06,
+      "loss": 0.0988,
+      "step": 37370
+    },
+    {
+      "epoch": 3.539102442719182,
+      "grad_norm": 0.5344982743263245,
+      "learning_rate": 5.843590229123272e-06,
+      "loss": 0.0946,
+      "step": 37380
+    },
+    {
+      "epoch": 3.5400492330997917,
+      "grad_norm": 0.4482509195804596,
+      "learning_rate": 5.839803067600833e-06,
+      "loss": 0.0998,
+      "step": 37390
+    },
+    {
+      "epoch": 3.5409960234804014,
+      "grad_norm": 0.4465737044811249,
+      "learning_rate": 5.836015906078395e-06,
+      "loss": 0.095,
+      "step": 37400
+    },
+    {
+      "epoch": 3.541942813861011,
+      "grad_norm": 0.6820048689842224,
+      "learning_rate": 5.832228744555956e-06,
+      "loss": 0.0971,
+      "step": 37410
+    },
+    {
+      "epoch": 3.542889604241621,
+      "grad_norm": 0.40104493498802185,
+      "learning_rate": 5.828441583033518e-06,
+      "loss": 0.0931,
+      "step": 37420
+    },
+    {
+      "epoch": 3.5438363946222307,
+      "grad_norm": 0.42221519351005554,
+      "learning_rate": 5.824654421511078e-06,
+      "loss": 0.0942,
+      "step": 37430
+    },
+    {
+      "epoch": 3.5447831850028404,
+      "grad_norm": 0.5034583210945129,
+      "learning_rate": 5.820867259988639e-06,
+      "loss": 0.1013,
+      "step": 37440
+    },
+    {
+      "epoch": 3.54572997538345,
+      "grad_norm": 0.42112550139427185,
+      "learning_rate": 5.8170800984662e-06,
+      "loss": 0.1039,
+      "step": 37450
+    },
+    {
+      "epoch": 3.54667676576406,
+      "grad_norm": 0.494607537984848,
+      "learning_rate": 5.81329293694376e-06,
+      "loss": 0.0986,
+      "step": 37460
+    },
+    {
+      "epoch": 3.5476235561446696,
+      "grad_norm": 0.449089378118515,
+      "learning_rate": 5.809505775421322e-06,
+      "loss": 0.0919,
+      "step": 37470
+    },
+    {
+      "epoch": 3.5485703465252794,
+      "grad_norm": 0.5358659625053406,
+      "learning_rate": 5.805718613898883e-06,
+      "loss": 0.1047,
+      "step": 37480
+    },
+    {
+      "epoch": 3.549517136905889,
+      "grad_norm": 0.43697452545166016,
+      "learning_rate": 5.801931452376445e-06,
+      "loss": 0.102,
+      "step": 37490
+    },
+    {
+      "epoch": 3.550463927286499,
+      "grad_norm": 0.48515307903289795,
+      "learning_rate": 5.798144290854006e-06,
+      "loss": 0.0946,
+      "step": 37500
+    },
+    {
+      "epoch": 3.5514107176671086,
+      "grad_norm": 0.5111203789710999,
+      "learning_rate": 5.794357129331566e-06,
+      "loss": 0.0979,
+      "step": 37510
+    },
+    {
+      "epoch": 3.5523575080477183,
+      "grad_norm": 0.47968727350234985,
+      "learning_rate": 5.7905699678091275e-06,
+      "loss": 0.0949,
+      "step": 37520
+    },
+    {
+      "epoch": 3.553304298428328,
+      "grad_norm": 0.36792320013046265,
+      "learning_rate": 5.786782806286688e-06,
+      "loss": 0.1011,
+      "step": 37530
+    },
+    {
+      "epoch": 3.554251088808938,
+      "grad_norm": 0.6238874197006226,
+      "learning_rate": 5.78299564476425e-06,
+      "loss": 0.1021,
+      "step": 37540
+    },
+    {
+      "epoch": 3.5551978791895475,
+      "grad_norm": 0.47502410411834717,
+      "learning_rate": 5.77920848324181e-06,
+      "loss": 0.0954,
+      "step": 37550
+    },
+    {
+      "epoch": 3.5561446695701573,
+      "grad_norm": 0.5178982615470886,
+      "learning_rate": 5.775421321719372e-06,
+      "loss": 0.1071,
+      "step": 37560
+    },
+    {
+      "epoch": 3.557091459950767,
+      "grad_norm": 0.5157225131988525,
+      "learning_rate": 5.771634160196933e-06,
+      "loss": 0.0962,
+      "step": 37570
+    },
+    {
+      "epoch": 3.5580382503313768,
+      "grad_norm": 0.4980148673057556,
+      "learning_rate": 5.7678469986744945e-06,
+      "loss": 0.1032,
+      "step": 37580
+    },
+    {
+      "epoch": 3.5589850407119865,
+      "grad_norm": 0.6088648438453674,
+      "learning_rate": 5.7640598371520554e-06,
+      "loss": 0.0972,
+      "step": 37590
+    },
+    {
+      "epoch": 3.5599318310925963,
+      "grad_norm": 0.42747214436531067,
+      "learning_rate": 5.7602726756296155e-06,
+      "loss": 0.0916,
+      "step": 37600
+    },
+    {
+      "epoch": 3.560878621473206,
+      "grad_norm": 0.5847950577735901,
+      "learning_rate": 5.756485514107177e-06,
+      "loss": 0.0955,
+      "step": 37610
+    },
+    {
+      "epoch": 3.5618254118538157,
+      "grad_norm": 0.5192122459411621,
+      "learning_rate": 5.752698352584738e-06,
+      "loss": 0.0986,
+      "step": 37620
+    },
+    {
+      "epoch": 3.5627722022344255,
+      "grad_norm": 0.506806492805481,
+      "learning_rate": 5.7489111910623e-06,
+      "loss": 0.0968,
+      "step": 37630
+    },
+    {
+      "epoch": 3.563718992615035,
+      "grad_norm": 0.4985431730747223,
+      "learning_rate": 5.74512402953986e-06,
+      "loss": 0.1011,
+      "step": 37640
+    },
+    {
+      "epoch": 3.564665782995645,
+      "grad_norm": 0.776935338973999,
+      "learning_rate": 5.741336868017422e-06,
+      "loss": 0.0975,
+      "step": 37650
+    },
+    {
+      "epoch": 3.5656125733762547,
+      "grad_norm": 0.46507251262664795,
+      "learning_rate": 5.7375497064949826e-06,
+      "loss": 0.0898,
+      "step": 37660
+    },
+    {
+      "epoch": 3.5665593637568644,
+      "grad_norm": 0.450090229511261,
+      "learning_rate": 5.7337625449725435e-06,
+      "loss": 0.111,
+      "step": 37670
+    },
+    {
+      "epoch": 3.567506154137474,
+      "grad_norm": 0.4301955997943878,
+      "learning_rate": 5.729975383450105e-06,
+      "loss": 0.0987,
+      "step": 37680
+    },
+    {
+      "epoch": 3.568452944518084,
+      "grad_norm": 0.36564552783966064,
+      "learning_rate": 5.726188221927665e-06,
+      "loss": 0.102,
+      "step": 37690
+    },
+    {
+      "epoch": 3.5693997348986937,
+      "grad_norm": 0.6277235150337219,
+      "learning_rate": 5.722401060405227e-06,
+      "loss": 0.1052,
+      "step": 37700
+    },
+    {
+      "epoch": 3.5703465252793034,
+      "grad_norm": 0.5736376643180847,
+      "learning_rate": 5.718613898882788e-06,
+      "loss": 0.109,
+      "step": 37710
+    },
+    {
+      "epoch": 3.571293315659913,
+      "grad_norm": 0.5228685736656189,
+      "learning_rate": 5.71482673736035e-06,
+      "loss": 0.1,
+      "step": 37720
+    },
+    {
+      "epoch": 3.572240106040523,
+      "grad_norm": 0.32450515031814575,
+      "learning_rate": 5.71103957583791e-06,
+      "loss": 0.1014,
+      "step": 37730
+    },
+    {
+      "epoch": 3.5731868964211326,
+      "grad_norm": 0.4867997169494629,
+      "learning_rate": 5.707252414315471e-06,
+      "loss": 0.0964,
+      "step": 37740
+    },
+    {
+      "epoch": 3.5741336868017424,
+      "grad_norm": 0.4821121394634247,
+      "learning_rate": 5.703465252793032e-06,
+      "loss": 0.0894,
+      "step": 37750
+    },
+    {
+      "epoch": 3.5750804771823517,
+      "grad_norm": 0.609014630317688,
+      "learning_rate": 5.699678091270593e-06,
+      "loss": 0.106,
+      "step": 37760
+    },
+    {
+      "epoch": 3.5760272675629614,
+      "grad_norm": 0.5423181056976318,
+      "learning_rate": 5.695890929748155e-06,
+      "loss": 0.1029,
+      "step": 37770
+    },
+    {
+      "epoch": 3.576974057943571,
+      "grad_norm": 0.464769572019577,
+      "learning_rate": 5.692103768225715e-06,
+      "loss": 0.1011,
+      "step": 37780
+    },
+    {
+      "epoch": 3.577920848324181,
+      "grad_norm": 0.4220525920391083,
+      "learning_rate": 5.688316606703277e-06,
+      "loss": 0.0923,
+      "step": 37790
+    },
+    {
+      "epoch": 3.5788676387047906,
+      "grad_norm": 0.4703858494758606,
+      "learning_rate": 5.684529445180838e-06,
+      "loss": 0.0998,
+      "step": 37800
+    },
+    {
+      "epoch": 3.5798144290854004,
+      "grad_norm": 0.5060818791389465,
+      "learning_rate": 5.680742283658398e-06,
+      "loss": 0.1134,
+      "step": 37810
+    },
+    {
+      "epoch": 3.58076121946601,
+      "grad_norm": 0.4083239436149597,
+      "learning_rate": 5.6769551221359594e-06,
+      "loss": 0.0953,
+      "step": 37820
+    },
+    {
+      "epoch": 3.58170800984662,
+      "grad_norm": 0.5225698947906494,
+      "learning_rate": 5.67316796061352e-06,
+      "loss": 0.0917,
+      "step": 37830
+    },
+    {
+      "epoch": 3.5826548002272296,
+      "grad_norm": 0.5536400675773621,
+      "learning_rate": 5.669380799091082e-06,
+      "loss": 0.1023,
+      "step": 37840
+    },
+    {
+      "epoch": 3.5836015906078393,
+      "grad_norm": 0.4061962366104126,
+      "learning_rate": 5.665593637568643e-06,
+      "loss": 0.0982,
+      "step": 37850
+    },
+    {
+      "epoch": 3.584548380988449,
+      "grad_norm": 0.48493853211402893,
+      "learning_rate": 5.661806476046205e-06,
+      "loss": 0.0926,
+      "step": 37860
+    },
+    {
+      "epoch": 3.585495171369059,
+      "grad_norm": 0.4256500005722046,
+      "learning_rate": 5.658019314523765e-06,
+      "loss": 0.1008,
+      "step": 37870
+    },
+    {
+      "epoch": 3.5864419617496686,
+      "grad_norm": 0.529008150100708,
+      "learning_rate": 5.654232153001326e-06,
+      "loss": 0.094,
+      "step": 37880
+    },
+    {
+      "epoch": 3.5873887521302783,
+      "grad_norm": 0.6380013227462769,
+      "learning_rate": 5.650444991478887e-06,
+      "loss": 0.0955,
+      "step": 37890
+    },
+    {
+      "epoch": 3.588335542510888,
+      "grad_norm": 0.4689798057079315,
+      "learning_rate": 5.6466578299564475e-06,
+      "loss": 0.0963,
+      "step": 37900
+    },
+    {
+      "epoch": 3.589282332891498,
+      "grad_norm": 0.5896109938621521,
+      "learning_rate": 5.642870668434009e-06,
+      "loss": 0.0981,
+      "step": 37910
+    },
+    {
+      "epoch": 3.5902291232721075,
+      "grad_norm": 0.4753442704677582,
+      "learning_rate": 5.63908350691157e-06,
+      "loss": 0.1036,
+      "step": 37920
+    },
+    {
+      "epoch": 3.5911759136527173,
+      "grad_norm": 0.40374627709388733,
+      "learning_rate": 5.635296345389132e-06,
+      "loss": 0.0946,
+      "step": 37930
+    },
+    {
+      "epoch": 3.592122704033327,
+      "grad_norm": 0.5366222858428955,
+      "learning_rate": 5.631509183866693e-06,
+      "loss": 0.0969,
+      "step": 37940
+    },
+    {
+      "epoch": 3.5930694944139367,
+      "grad_norm": 0.5004488825798035,
+      "learning_rate": 5.627722022344253e-06,
+      "loss": 0.0981,
+      "step": 37950
+    },
+    {
+      "epoch": 3.5940162847945465,
+      "grad_norm": 0.43726542592048645,
+      "learning_rate": 5.6239348608218145e-06,
+      "loss": 0.0961,
+      "step": 37960
+    },
+    {
+      "epoch": 3.5949630751751562,
+      "grad_norm": 0.4899952709674835,
+      "learning_rate": 5.6201476992993754e-06,
+      "loss": 0.103,
+      "step": 37970
+    },
+    {
+      "epoch": 3.595909865555766,
+      "grad_norm": 0.5843260288238525,
+      "learning_rate": 5.616360537776937e-06,
+      "loss": 0.1004,
+      "step": 37980
+    },
+    {
+      "epoch": 3.5968566559363757,
+      "grad_norm": 0.43663713335990906,
+      "learning_rate": 5.612573376254497e-06,
+      "loss": 0.1021,
+      "step": 37990
+    },
+    {
+      "epoch": 3.5978034463169855,
+      "grad_norm": 0.44171518087387085,
+      "learning_rate": 5.608786214732059e-06,
+      "loss": 0.1001,
+      "step": 38000
+    },
+    {
+      "epoch": 3.598750236697595,
+      "grad_norm": 0.5805891752243042,
+      "learning_rate": 5.60499905320962e-06,
+      "loss": 0.1009,
+      "step": 38010
+    },
+    {
+      "epoch": 3.599697027078205,
+      "grad_norm": 0.43290090560913086,
+      "learning_rate": 5.601211891687181e-06,
+      "loss": 0.1025,
+      "step": 38020
+    },
+    {
+      "epoch": 3.6006438174588147,
+      "grad_norm": 0.4977321922779083,
+      "learning_rate": 5.5974247301647425e-06,
+      "loss": 0.0943,
+      "step": 38030
+    },
+    {
+      "epoch": 3.6015906078394244,
+      "grad_norm": 0.4680572748184204,
+      "learning_rate": 5.5936375686423026e-06,
+      "loss": 0.0977,
+      "step": 38040
+    },
+    {
+      "epoch": 3.602537398220034,
+      "grad_norm": 0.567624568939209,
+      "learning_rate": 5.589850407119864e-06,
+      "loss": 0.0968,
+      "step": 38050
+    },
+    {
+      "epoch": 3.603484188600644,
+      "grad_norm": 0.41031312942504883,
+      "learning_rate": 5.586063245597425e-06,
+      "loss": 0.091,
+      "step": 38060
+    },
+    {
+      "epoch": 3.6044309789812536,
+      "grad_norm": 0.5081103444099426,
+      "learning_rate": 5.582276084074987e-06,
+      "loss": 0.1007,
+      "step": 38070
+    },
+    {
+      "epoch": 3.6053777693618634,
+      "grad_norm": 0.5523340106010437,
+      "learning_rate": 5.578488922552547e-06,
+      "loss": 0.0965,
+      "step": 38080
+    },
+    {
+      "epoch": 3.606324559742473,
+      "grad_norm": 0.49104392528533936,
+      "learning_rate": 5.574701761030108e-06,
+      "loss": 0.0952,
+      "step": 38090
+    },
+    {
+      "epoch": 3.607271350123083,
+      "grad_norm": 0.37221527099609375,
+      "learning_rate": 5.57091459950767e-06,
+      "loss": 0.0991,
+      "step": 38100
+    },
+    {
+      "epoch": 3.6082181405036926,
+      "grad_norm": 0.43933555483818054,
+      "learning_rate": 5.5671274379852305e-06,
+      "loss": 0.0957,
+      "step": 38110
+    },
+    {
+      "epoch": 3.6091649308843023,
+      "grad_norm": 0.5173497796058655,
+      "learning_rate": 5.563340276462792e-06,
+      "loss": 0.095,
+      "step": 38120
+    },
+    {
+      "epoch": 3.610111721264912,
+      "grad_norm": 0.43472984433174133,
+      "learning_rate": 5.559553114940352e-06,
+      "loss": 0.101,
+      "step": 38130
+    },
+    {
+      "epoch": 3.611058511645522,
+      "grad_norm": 0.5377931594848633,
+      "learning_rate": 5.555765953417914e-06,
+      "loss": 0.0956,
+      "step": 38140
+    },
+    {
+      "epoch": 3.6120053020261316,
+      "grad_norm": 0.35924702882766724,
+      "learning_rate": 5.551978791895475e-06,
+      "loss": 0.0932,
+      "step": 38150
+    },
+    {
+      "epoch": 3.6129520924067413,
+      "grad_norm": 0.6031462550163269,
+      "learning_rate": 5.548191630373035e-06,
+      "loss": 0.0976,
+      "step": 38160
+    },
+    {
+      "epoch": 3.6138988827873506,
+      "grad_norm": 0.435293048620224,
+      "learning_rate": 5.544404468850597e-06,
+      "loss": 0.1001,
+      "step": 38170
+    },
+    {
+      "epoch": 3.6148456731679603,
+      "grad_norm": 0.5811923742294312,
+      "learning_rate": 5.540617307328158e-06,
+      "loss": 0.1128,
+      "step": 38180
+    },
+    {
+      "epoch": 3.61579246354857,
+      "grad_norm": 0.49129927158355713,
+      "learning_rate": 5.536830145805719e-06,
+      "loss": 0.0932,
+      "step": 38190
+    },
+    {
+      "epoch": 3.61673925392918,
+      "grad_norm": 0.41449084877967834,
+      "learning_rate": 5.53304298428328e-06,
+      "loss": 0.097,
+      "step": 38200
+    },
+    {
+      "epoch": 3.6176860443097896,
+      "grad_norm": 0.5685412287712097,
+      "learning_rate": 5.529255822760842e-06,
+      "loss": 0.0907,
+      "step": 38210
+    },
+    {
+      "epoch": 3.6186328346903993,
+      "grad_norm": 0.5093902945518494,
+      "learning_rate": 5.525468661238402e-06,
+      "loss": 0.0942,
+      "step": 38220
+    },
+    {
+      "epoch": 3.619579625071009,
+      "grad_norm": 0.5617339015007019,
+      "learning_rate": 5.521681499715963e-06,
+      "loss": 0.1015,
+      "step": 38230
+    },
+    {
+      "epoch": 3.620526415451619,
+      "grad_norm": 0.47863510251045227,
+      "learning_rate": 5.517894338193525e-06,
+      "loss": 0.0952,
+      "step": 38240
+    },
+    {
+      "epoch": 3.6214732058322285,
+      "grad_norm": 0.6331188678741455,
+      "learning_rate": 5.514107176671085e-06,
+      "loss": 0.111,
+      "step": 38250
+    },
+    {
+      "epoch": 3.6224199962128383,
+      "grad_norm": 0.45131048560142517,
+      "learning_rate": 5.5103200151486465e-06,
+      "loss": 0.0915,
+      "step": 38260
+    },
+    {
+      "epoch": 3.623366786593448,
+      "grad_norm": 0.4484195411205292,
+      "learning_rate": 5.506532853626207e-06,
+      "loss": 0.0964,
+      "step": 38270
+    },
+    {
+      "epoch": 3.6243135769740578,
+      "grad_norm": 0.5122811198234558,
+      "learning_rate": 5.502745692103769e-06,
+      "loss": 0.1005,
+      "step": 38280
+    },
+    {
+      "epoch": 3.6252603673546675,
+      "grad_norm": 0.5720915794372559,
+      "learning_rate": 5.49895853058133e-06,
+      "loss": 0.0957,
+      "step": 38290
+    },
+    {
+      "epoch": 3.6262071577352772,
+      "grad_norm": 0.4863225221633911,
+      "learning_rate": 5.49517136905889e-06,
+      "loss": 0.0901,
+      "step": 38300
+    },
+    {
+      "epoch": 3.627153948115887,
+      "grad_norm": 0.5356199145317078,
+      "learning_rate": 5.491384207536452e-06,
+      "loss": 0.0994,
+      "step": 38310
+    },
+    {
+      "epoch": 3.6281007384964967,
+      "grad_norm": 0.46760469675064087,
+      "learning_rate": 5.487597046014013e-06,
+      "loss": 0.1005,
+      "step": 38320
+    },
+    {
+      "epoch": 3.6290475288771065,
+      "grad_norm": 0.39269837737083435,
+      "learning_rate": 5.4838098844915745e-06,
+      "loss": 0.0925,
+      "step": 38330
+    },
+    {
+      "epoch": 3.629994319257716,
+      "grad_norm": 0.49656835198402405,
+      "learning_rate": 5.4800227229691345e-06,
+      "loss": 0.0932,
+      "step": 38340
+    },
+    {
+      "epoch": 3.630941109638326,
+      "grad_norm": 0.4261341691017151,
+      "learning_rate": 5.476235561446696e-06,
+      "loss": 0.0933,
+      "step": 38350
+    },
+    {
+      "epoch": 3.6318879000189357,
+      "grad_norm": 0.5111949443817139,
+      "learning_rate": 5.472448399924257e-06,
+      "loss": 0.1001,
+      "step": 38360
+    },
+    {
+      "epoch": 3.6328346903995454,
+      "grad_norm": 0.4922689199447632,
+      "learning_rate": 5.468661238401818e-06,
+      "loss": 0.0957,
+      "step": 38370
+    },
+    {
+      "epoch": 3.633781480780155,
+      "grad_norm": 0.5869754552841187,
+      "learning_rate": 5.46487407687938e-06,
+      "loss": 0.0937,
+      "step": 38380
+    },
+    {
+      "epoch": 3.634728271160765,
+      "grad_norm": 0.6243377327919006,
+      "learning_rate": 5.46108691535694e-06,
+      "loss": 0.0976,
+      "step": 38390
+    },
+    {
+      "epoch": 3.6356750615413747,
+      "grad_norm": 0.5943287014961243,
+      "learning_rate": 5.457299753834502e-06,
+      "loss": 0.0877,
+      "step": 38400
+    },
+    {
+      "epoch": 3.6366218519219844,
+      "grad_norm": 0.5626177191734314,
+      "learning_rate": 5.4535125923120625e-06,
+      "loss": 0.1154,
+      "step": 38410
+    },
+    {
+      "epoch": 3.637568642302594,
+      "grad_norm": 0.48984989523887634,
+      "learning_rate": 5.449725430789624e-06,
+      "loss": 0.081,
+      "step": 38420
+    },
+    {
+      "epoch": 3.638515432683204,
+      "grad_norm": 0.3821601867675781,
+      "learning_rate": 5.445938269267184e-06,
+      "loss": 0.0934,
+      "step": 38430
+    },
+    {
+      "epoch": 3.6394622230638136,
+      "grad_norm": 0.38352078199386597,
+      "learning_rate": 5.442151107744745e-06,
+      "loss": 0.0917,
+      "step": 38440
+    },
+    {
+      "epoch": 3.6404090134444234,
+      "grad_norm": 0.505770206451416,
+      "learning_rate": 5.438363946222307e-06,
+      "loss": 0.1035,
+      "step": 38450
+    },
+    {
+      "epoch": 3.641355803825033,
+      "grad_norm": 0.3925401568412781,
+      "learning_rate": 5.434576784699868e-06,
+      "loss": 0.0872,
+      "step": 38460
+    },
+    {
+      "epoch": 3.642302594205643,
+      "grad_norm": 0.40184637904167175,
+      "learning_rate": 5.43078962317743e-06,
+      "loss": 0.0957,
+      "step": 38470
+    },
+    {
+      "epoch": 3.6432493845862526,
+      "grad_norm": 0.4207031726837158,
+      "learning_rate": 5.42700246165499e-06,
+      "loss": 0.095,
+      "step": 38480
+    },
+    {
+      "epoch": 3.6441961749668623,
+      "grad_norm": 0.5287842750549316,
+      "learning_rate": 5.423215300132551e-06,
+      "loss": 0.0976,
+      "step": 38490
+    },
+    {
+      "epoch": 3.645142965347472,
+      "grad_norm": 0.5332127809524536,
+      "learning_rate": 5.419428138610112e-06,
+      "loss": 0.0938,
+      "step": 38500
+    },
+    {
+      "epoch": 3.646089755728082,
+      "grad_norm": 0.5463581085205078,
+      "learning_rate": 5.415640977087673e-06,
+      "loss": 0.1012,
+      "step": 38510
+    },
+    {
+      "epoch": 3.6470365461086915,
+      "grad_norm": 0.42161843180656433,
+      "learning_rate": 5.411853815565234e-06,
+      "loss": 0.1035,
+      "step": 38520
+    },
+    {
+      "epoch": 3.6479833364893013,
+      "grad_norm": 0.42370715737342834,
+      "learning_rate": 5.408066654042795e-06,
+      "loss": 0.1005,
+      "step": 38530
+    },
+    {
+      "epoch": 3.648930126869911,
+      "grad_norm": 0.48021796345710754,
+      "learning_rate": 5.404279492520357e-06,
+      "loss": 0.0997,
+      "step": 38540
+    },
+    {
+      "epoch": 3.6498769172505208,
+      "grad_norm": 0.4591447710990906,
+      "learning_rate": 5.400492330997918e-06,
+      "loss": 0.0933,
+      "step": 38550
+    },
+    {
+      "epoch": 3.6508237076311305,
+      "grad_norm": 0.5126654505729675,
+      "learning_rate": 5.396705169475479e-06,
+      "loss": 0.0943,
+      "step": 38560
+    },
+    {
+      "epoch": 3.6517704980117403,
+      "grad_norm": 0.39282628893852234,
+      "learning_rate": 5.392918007953039e-06,
+      "loss": 0.1074,
+      "step": 38570
+    },
+    {
+      "epoch": 3.65271728839235,
+      "grad_norm": 0.5055245161056519,
+      "learning_rate": 5.3891308464306e-06,
+      "loss": 0.0951,
+      "step": 38580
+    },
+    {
+      "epoch": 3.6536640787729597,
+      "grad_norm": 0.45934176445007324,
+      "learning_rate": 5.385343684908162e-06,
+      "loss": 0.0912,
+      "step": 38590
+    },
+    {
+      "epoch": 3.6546108691535695,
+      "grad_norm": 0.5393901467323303,
+      "learning_rate": 5.381556523385723e-06,
+      "loss": 0.0963,
+      "step": 38600
+    },
+    {
+      "epoch": 3.655557659534179,
+      "grad_norm": 0.47153976559638977,
+      "learning_rate": 5.377769361863284e-06,
+      "loss": 0.0953,
+      "step": 38610
+    },
+    {
+      "epoch": 3.656504449914789,
+      "grad_norm": 0.5694108605384827,
+      "learning_rate": 5.373982200340845e-06,
+      "loss": 0.109,
+      "step": 38620
+    },
+    {
+      "epoch": 3.6574512402953987,
+      "grad_norm": 0.49688521027565,
+      "learning_rate": 5.3701950388184065e-06,
+      "loss": 0.0938,
+      "step": 38630
+    },
+    {
+      "epoch": 3.6583980306760084,
+      "grad_norm": 0.5291985273361206,
+      "learning_rate": 5.366407877295967e-06,
+      "loss": 0.0899,
+      "step": 38640
+    },
+    {
+      "epoch": 3.659344821056618,
+      "grad_norm": 0.4994890093803406,
+      "learning_rate": 5.362620715773527e-06,
+      "loss": 0.0931,
+      "step": 38650
+    },
+    {
+      "epoch": 3.660291611437228,
+      "grad_norm": 0.5147172808647156,
+      "learning_rate": 5.358833554251089e-06,
+      "loss": 0.0946,
+      "step": 38660
+    },
+    {
+      "epoch": 3.6612384018178377,
+      "grad_norm": 0.36024779081344604,
+      "learning_rate": 5.35504639272865e-06,
+      "loss": 0.085,
+      "step": 38670
+    },
+    {
+      "epoch": 3.6621851921984474,
+      "grad_norm": 0.5125046372413635,
+      "learning_rate": 5.351259231206212e-06,
+      "loss": 0.0917,
+      "step": 38680
+    },
+    {
+      "epoch": 3.663131982579057,
+      "grad_norm": 0.5628103613853455,
+      "learning_rate": 5.347472069683773e-06,
+      "loss": 0.0934,
+      "step": 38690
+    },
+    {
+      "epoch": 3.664078772959667,
+      "grad_norm": 0.390716016292572,
+      "learning_rate": 5.343684908161334e-06,
+      "loss": 0.096,
+      "step": 38700
+    },
+    {
+      "epoch": 3.6650255633402766,
+      "grad_norm": 0.41418376564979553,
+      "learning_rate": 5.3398977466388945e-06,
+      "loss": 0.095,
+      "step": 38710
+    },
+    {
+      "epoch": 3.6659723537208864,
+      "grad_norm": 0.4665653109550476,
+      "learning_rate": 5.336110585116455e-06,
+      "loss": 0.0858,
+      "step": 38720
+    },
+    {
+      "epoch": 3.666919144101496,
+      "grad_norm": 0.35321033000946045,
+      "learning_rate": 5.332323423594017e-06,
+      "loss": 0.0907,
+      "step": 38730
+    },
+    {
+      "epoch": 3.667865934482106,
+      "grad_norm": 0.5768105387687683,
+      "learning_rate": 5.328536262071577e-06,
+      "loss": 0.106,
+      "step": 38740
+    },
+    {
+      "epoch": 3.6688127248627156,
+      "grad_norm": 0.383003294467926,
+      "learning_rate": 5.324749100549139e-06,
+      "loss": 0.0893,
+      "step": 38750
+    },
+    {
+      "epoch": 3.6697595152433253,
+      "grad_norm": 0.5730847120285034,
+      "learning_rate": 5.3209619390267e-06,
+      "loss": 0.1072,
+      "step": 38760
+    },
+    {
+      "epoch": 3.670706305623935,
+      "grad_norm": 0.5745497345924377,
+      "learning_rate": 5.3171747775042616e-06,
+      "loss": 0.1006,
+      "step": 38770
+    },
+    {
+      "epoch": 3.671653096004545,
+      "grad_norm": 0.5345969796180725,
+      "learning_rate": 5.3133876159818225e-06,
+      "loss": 0.1059,
+      "step": 38780
+    },
+    {
+      "epoch": 3.6725998863851546,
+      "grad_norm": 0.5308682918548584,
+      "learning_rate": 5.3096004544593825e-06,
+      "loss": 0.1005,
+      "step": 38790
+    },
+    {
+      "epoch": 3.6735466767657643,
+      "grad_norm": 0.4583006203174591,
+      "learning_rate": 5.305813292936944e-06,
+      "loss": 0.0957,
+      "step": 38800
+    },
+    {
+      "epoch": 3.674493467146374,
+      "grad_norm": 0.511604905128479,
+      "learning_rate": 5.302026131414505e-06,
+      "loss": 0.1052,
+      "step": 38810
+    },
+    {
+      "epoch": 3.675440257526984,
+      "grad_norm": 0.42457523941993713,
+      "learning_rate": 5.298238969892067e-06,
+      "loss": 0.099,
+      "step": 38820
+    },
+    {
+      "epoch": 3.6763870479075935,
+      "grad_norm": 0.3803611993789673,
+      "learning_rate": 5.294451808369627e-06,
+      "loss": 0.102,
+      "step": 38830
+    },
+    {
+      "epoch": 3.6773338382882033,
+      "grad_norm": 0.4064808785915375,
+      "learning_rate": 5.290664646847189e-06,
+      "loss": 0.0926,
+      "step": 38840
+    },
+    {
+      "epoch": 3.678280628668813,
+      "grad_norm": 0.4240824580192566,
+      "learning_rate": 5.28687748532475e-06,
+      "loss": 0.0975,
+      "step": 38850
+    },
+    {
+      "epoch": 3.6792274190494223,
+      "grad_norm": 0.4897139370441437,
+      "learning_rate": 5.2830903238023105e-06,
+      "loss": 0.1003,
+      "step": 38860
+    },
+    {
+      "epoch": 3.680174209430032,
+      "grad_norm": 0.5453110933303833,
+      "learning_rate": 5.279303162279872e-06,
+      "loss": 0.0876,
+      "step": 38870
+    },
+    {
+      "epoch": 3.681120999810642,
+      "grad_norm": 0.5848513245582581,
+      "learning_rate": 5.275516000757432e-06,
+      "loss": 0.094,
+      "step": 38880
+    },
+    {
+      "epoch": 3.6820677901912515,
+      "grad_norm": 0.5227193832397461,
+      "learning_rate": 5.271728839234994e-06,
+      "loss": 0.0932,
+      "step": 38890
+    },
+    {
+      "epoch": 3.6830145805718613,
+      "grad_norm": 0.4724016487598419,
+      "learning_rate": 5.267941677712555e-06,
+      "loss": 0.0948,
+      "step": 38900
+    },
+    {
+      "epoch": 3.683961370952471,
+      "grad_norm": 0.39952653646469116,
+      "learning_rate": 5.264154516190117e-06,
+      "loss": 0.0876,
+      "step": 38910
+    },
+    {
+      "epoch": 3.6849081613330807,
+      "grad_norm": 0.5192437767982483,
+      "learning_rate": 5.260367354667677e-06,
+      "loss": 0.0963,
+      "step": 38920
+    },
+    {
+      "epoch": 3.6858549517136905,
+      "grad_norm": 0.4506594240665436,
+      "learning_rate": 5.256580193145238e-06,
+      "loss": 0.099,
+      "step": 38930
+    },
+    {
+      "epoch": 3.6868017420943002,
+      "grad_norm": 0.5142754316329956,
+      "learning_rate": 5.252793031622799e-06,
+      "loss": 0.09,
+      "step": 38940
+    },
+    {
+      "epoch": 3.68774853247491,
+      "grad_norm": 0.4741668105125427,
+      "learning_rate": 5.24900587010036e-06,
+      "loss": 0.0977,
+      "step": 38950
+    },
+    {
+      "epoch": 3.6886953228555197,
+      "grad_norm": 0.40883010625839233,
+      "learning_rate": 5.245218708577922e-06,
+      "loss": 0.0943,
+      "step": 38960
+    },
+    {
+      "epoch": 3.6896421132361295,
+      "grad_norm": 0.4715437591075897,
+      "learning_rate": 5.241431547055482e-06,
+      "loss": 0.0937,
+      "step": 38970
+    },
+    {
+      "epoch": 3.690588903616739,
+      "grad_norm": 0.44934457540512085,
+      "learning_rate": 5.237644385533044e-06,
+      "loss": 0.0993,
+      "step": 38980
+    },
+    {
+      "epoch": 3.691535693997349,
+      "grad_norm": 0.5743240714073181,
+      "learning_rate": 5.233857224010605e-06,
+      "loss": 0.0994,
+      "step": 38990
+    },
+    {
+      "epoch": 3.6924824843779587,
+      "grad_norm": 0.5706230998039246,
+      "learning_rate": 5.230070062488165e-06,
+      "loss": 0.1102,
+      "step": 39000
+    },
+    {
+      "epoch": 3.6934292747585684,
+      "grad_norm": 0.6386612057685852,
+      "learning_rate": 5.2262829009657265e-06,
+      "loss": 0.1075,
+      "step": 39010
+    },
+    {
+      "epoch": 3.694376065139178,
+      "grad_norm": 0.5190657377243042,
+      "learning_rate": 5.222495739443287e-06,
+      "loss": 0.0912,
+      "step": 39020
+    },
+    {
+      "epoch": 3.695322855519788,
+      "grad_norm": 0.4563543200492859,
+      "learning_rate": 5.218708577920849e-06,
+      "loss": 0.0967,
+      "step": 39030
+    },
+    {
+      "epoch": 3.6962696459003976,
+      "grad_norm": 0.477840393781662,
+      "learning_rate": 5.21492141639841e-06,
+      "loss": 0.0921,
+      "step": 39040
+    },
+    {
+      "epoch": 3.6972164362810074,
+      "grad_norm": 0.46329066157341003,
+      "learning_rate": 5.211134254875972e-06,
+      "loss": 0.0922,
+      "step": 39050
+    },
+    {
+      "epoch": 3.698163226661617,
+      "grad_norm": 0.46231696009635925,
+      "learning_rate": 5.207347093353532e-06,
+      "loss": 0.0902,
+      "step": 39060
+    },
+    {
+      "epoch": 3.699110017042227,
+      "grad_norm": 0.4845748245716095,
+      "learning_rate": 5.203559931831093e-06,
+      "loss": 0.0913,
+      "step": 39070
+    },
+    {
+      "epoch": 3.7000568074228366,
+      "grad_norm": 0.4496608078479767,
+      "learning_rate": 5.1997727703086544e-06,
+      "loss": 0.1088,
+      "step": 39080
+    },
+    {
+      "epoch": 3.7010035978034463,
+      "grad_norm": 0.5793826580047607,
+      "learning_rate": 5.1959856087862145e-06,
+      "loss": 0.0963,
+      "step": 39090
+    },
+    {
+      "epoch": 3.701950388184056,
+      "grad_norm": 0.43485772609710693,
+      "learning_rate": 5.192198447263776e-06,
+      "loss": 0.0919,
+      "step": 39100
+    },
+    {
+      "epoch": 3.702897178564666,
+      "grad_norm": 0.47092369198799133,
+      "learning_rate": 5.188411285741337e-06,
+      "loss": 0.1123,
+      "step": 39110
+    },
+    {
+      "epoch": 3.7038439689452756,
+      "grad_norm": 0.5703817009925842,
+      "learning_rate": 5.184624124218899e-06,
+      "loss": 0.1002,
+      "step": 39120
+    },
+    {
+      "epoch": 3.7047907593258853,
+      "grad_norm": 0.5016136765480042,
+      "learning_rate": 5.18083696269646e-06,
+      "loss": 0.091,
+      "step": 39130
+    },
+    {
+      "epoch": 3.705737549706495,
+      "grad_norm": 0.5749192237854004,
+      "learning_rate": 5.17704980117402e-06,
+      "loss": 0.0987,
+      "step": 39140
+    },
+    {
+      "epoch": 3.706684340087105,
+      "grad_norm": 0.4935661554336548,
+      "learning_rate": 5.1732626396515816e-06,
+      "loss": 0.103,
+      "step": 39150
+    },
+    {
+      "epoch": 3.7076311304677145,
+      "grad_norm": 0.4583258032798767,
+      "learning_rate": 5.1694754781291425e-06,
+      "loss": 0.1013,
+      "step": 39160
+    },
+    {
+      "epoch": 3.7085779208483243,
+      "grad_norm": 0.6873877644538879,
+      "learning_rate": 5.165688316606704e-06,
+      "loss": 0.1042,
+      "step": 39170
+    },
+    {
+      "epoch": 3.709524711228934,
+      "grad_norm": 0.4328051507472992,
+      "learning_rate": 5.161901155084264e-06,
+      "loss": 0.0933,
+      "step": 39180
+    },
+    {
+      "epoch": 3.7104715016095438,
+      "grad_norm": 0.4692210257053375,
+      "learning_rate": 5.158113993561826e-06,
+      "loss": 0.0967,
+      "step": 39190
+    },
+    {
+      "epoch": 3.7114182919901535,
+      "grad_norm": 0.48374322056770325,
+      "learning_rate": 5.154326832039387e-06,
+      "loss": 0.1008,
+      "step": 39200
+    },
+    {
+      "epoch": 3.7123650823707632,
+      "grad_norm": 0.3715035021305084,
+      "learning_rate": 5.150539670516948e-06,
+      "loss": 0.1025,
+      "step": 39210
+    },
+    {
+      "epoch": 3.713311872751373,
+      "grad_norm": 0.4053986370563507,
+      "learning_rate": 5.1467525089945095e-06,
+      "loss": 0.0927,
+      "step": 39220
+    },
+    {
+      "epoch": 3.7142586631319827,
+      "grad_norm": 0.44326600432395935,
+      "learning_rate": 5.14296534747207e-06,
+      "loss": 0.1005,
+      "step": 39230
+    },
+    {
+      "epoch": 3.7152054535125925,
+      "grad_norm": 0.4053700268268585,
+      "learning_rate": 5.139178185949631e-06,
+      "loss": 0.0878,
+      "step": 39240
+    },
+    {
+      "epoch": 3.716152243893202,
+      "grad_norm": 0.6459446549415588,
+      "learning_rate": 5.135391024427192e-06,
+      "loss": 0.098,
+      "step": 39250
+    },
+    {
+      "epoch": 3.717099034273812,
+      "grad_norm": 0.5322126746177673,
+      "learning_rate": 5.131603862904754e-06,
+      "loss": 0.0916,
+      "step": 39260
+    },
+    {
+      "epoch": 3.7180458246544212,
+      "grad_norm": 0.4378093481063843,
+      "learning_rate": 5.127816701382314e-06,
+      "loss": 0.0913,
+      "step": 39270
+    },
+    {
+      "epoch": 3.718992615035031,
+      "grad_norm": 0.46981629729270935,
+      "learning_rate": 5.124029539859875e-06,
+      "loss": 0.1043,
+      "step": 39280
+    },
+    {
+      "epoch": 3.7199394054156407,
+      "grad_norm": 0.5976932048797607,
+      "learning_rate": 5.120242378337437e-06,
+      "loss": 0.0986,
+      "step": 39290
+    },
+    {
+      "epoch": 3.7208861957962505,
+      "grad_norm": 0.48355865478515625,
+      "learning_rate": 5.1164552168149975e-06,
+      "loss": 0.0873,
+      "step": 39300
+    },
+    {
+      "epoch": 3.72183298617686,
+      "grad_norm": 0.49303942918777466,
+      "learning_rate": 5.112668055292559e-06,
+      "loss": 0.0943,
+      "step": 39310
+    },
+    {
+      "epoch": 3.72277977655747,
+      "grad_norm": 0.41613462567329407,
+      "learning_rate": 5.108880893770119e-06,
+      "loss": 0.0976,
+      "step": 39320
+    },
+    {
+      "epoch": 3.7237265669380797,
+      "grad_norm": 0.5080744624137878,
+      "learning_rate": 5.105093732247681e-06,
+      "loss": 0.0889,
+      "step": 39330
+    },
+    {
+      "epoch": 3.7246733573186894,
+      "grad_norm": 0.47723624110221863,
+      "learning_rate": 5.101306570725242e-06,
+      "loss": 0.1033,
+      "step": 39340
+    },
+    {
+      "epoch": 3.725620147699299,
+      "grad_norm": 0.37608417868614197,
+      "learning_rate": 5.097519409202802e-06,
+      "loss": 0.0966,
+      "step": 39350
+    },
+    {
+      "epoch": 3.726566938079909,
+      "grad_norm": 0.38140979409217834,
+      "learning_rate": 5.093732247680364e-06,
+      "loss": 0.0972,
+      "step": 39360
+    },
+    {
+      "epoch": 3.7275137284605186,
+      "grad_norm": 0.39315181970596313,
+      "learning_rate": 5.089945086157925e-06,
+      "loss": 0.0953,
+      "step": 39370
+    },
+    {
+      "epoch": 3.7284605188411284,
+      "grad_norm": 0.525129497051239,
+      "learning_rate": 5.086157924635486e-06,
+      "loss": 0.0933,
+      "step": 39380
+    },
+    {
+      "epoch": 3.729407309221738,
+      "grad_norm": 0.464629203081131,
+      "learning_rate": 5.082370763113047e-06,
+      "loss": 0.095,
+      "step": 39390
+    },
+    {
+      "epoch": 3.730354099602348,
+      "grad_norm": 0.46547815203666687,
+      "learning_rate": 5.078583601590609e-06,
+      "loss": 0.0894,
+      "step": 39400
+    },
+    {
+      "epoch": 3.7313008899829576,
+      "grad_norm": 0.5702111124992371,
+      "learning_rate": 5.074796440068169e-06,
+      "loss": 0.0897,
+      "step": 39410
+    },
+    {
+      "epoch": 3.7322476803635674,
+      "grad_norm": 0.45302221179008484,
+      "learning_rate": 5.07100927854573e-06,
+      "loss": 0.0917,
+      "step": 39420
+    },
+    {
+      "epoch": 3.733194470744177,
+      "grad_norm": 0.5478273034095764,
+      "learning_rate": 5.067222117023292e-06,
+      "loss": 0.0968,
+      "step": 39430
+    },
+    {
+      "epoch": 3.734141261124787,
+      "grad_norm": 0.5107770562171936,
+      "learning_rate": 5.063434955500852e-06,
+      "loss": 0.0977,
+      "step": 39440
+    },
+    {
+      "epoch": 3.7350880515053966,
+      "grad_norm": 0.6436827182769775,
+      "learning_rate": 5.0596477939784135e-06,
+      "loss": 0.1011,
+      "step": 39450
+    },
+    {
+      "epoch": 3.7360348418860063,
+      "grad_norm": 0.49567675590515137,
+      "learning_rate": 5.0558606324559744e-06,
+      "loss": 0.0942,
+      "step": 39460
+    },
+    {
+      "epoch": 3.736981632266616,
+      "grad_norm": 0.5632124543190002,
+      "learning_rate": 5.052073470933536e-06,
+      "loss": 0.0965,
+      "step": 39470
+    },
+    {
+      "epoch": 3.737928422647226,
+      "grad_norm": 0.47855645418167114,
+      "learning_rate": 5.048286309411097e-06,
+      "loss": 0.0938,
+      "step": 39480
+    },
+    {
+      "epoch": 3.7388752130278355,
+      "grad_norm": 0.4823874533176422,
+      "learning_rate": 5.044499147888657e-06,
+      "loss": 0.1016,
+      "step": 39490
+    },
+    {
+      "epoch": 3.7398220034084453,
+      "grad_norm": 0.4572794437408447,
+      "learning_rate": 5.040711986366219e-06,
+      "loss": 0.0936,
+      "step": 39500
+    },
+    {
+      "epoch": 3.740768793789055,
+      "grad_norm": 0.4439483880996704,
+      "learning_rate": 5.03692482484378e-06,
+      "loss": 0.0971,
+      "step": 39510
+    },
+    {
+      "epoch": 3.7417155841696648,
+      "grad_norm": 0.3832974433898926,
+      "learning_rate": 5.0331376633213415e-06,
+      "loss": 0.0943,
+      "step": 39520
+    },
+    {
+      "epoch": 3.7426623745502745,
+      "grad_norm": 0.40241408348083496,
+      "learning_rate": 5.0293505017989016e-06,
+      "loss": 0.0983,
+      "step": 39530
+    },
+    {
+      "epoch": 3.7436091649308842,
+      "grad_norm": 0.43205514550209045,
+      "learning_rate": 5.025563340276463e-06,
+      "loss": 0.0973,
+      "step": 39540
+    },
+    {
+      "epoch": 3.744555955311494,
+      "grad_norm": 0.6235865950584412,
+      "learning_rate": 5.021776178754024e-06,
+      "loss": 0.0899,
+      "step": 39550
+    },
+    {
+      "epoch": 3.7455027456921037,
+      "grad_norm": 0.48281237483024597,
+      "learning_rate": 5.017989017231585e-06,
+      "loss": 0.0924,
+      "step": 39560
+    },
+    {
+      "epoch": 3.7464495360727135,
+      "grad_norm": 0.4790968894958496,
+      "learning_rate": 5.014201855709147e-06,
+      "loss": 0.0955,
+      "step": 39570
+    },
+    {
+      "epoch": 3.747396326453323,
+      "grad_norm": 0.46759265661239624,
+      "learning_rate": 5.010414694186707e-06,
+      "loss": 0.0945,
+      "step": 39580
+    },
+    {
+      "epoch": 3.748343116833933,
+      "grad_norm": 0.5322169661521912,
+      "learning_rate": 5.006627532664269e-06,
+      "loss": 0.1096,
+      "step": 39590
+    },
+    {
+      "epoch": 3.7492899072145427,
+      "grad_norm": 0.46110084652900696,
+      "learning_rate": 5.0028403711418295e-06,
+      "loss": 0.0896,
+      "step": 39600
+    },
+    {
+      "epoch": 3.7502366975951524,
+      "grad_norm": 0.4692782461643219,
+      "learning_rate": 4.99905320961939e-06,
+      "loss": 0.1023,
+      "step": 39610
+    },
+    {
+      "epoch": 3.751183487975762,
+      "grad_norm": 0.5090218186378479,
+      "learning_rate": 4.995266048096951e-06,
+      "loss": 0.0952,
+      "step": 39620
+    },
+    {
+      "epoch": 3.752130278356372,
+      "grad_norm": 0.41407740116119385,
+      "learning_rate": 4.991478886574513e-06,
+      "loss": 0.0829,
+      "step": 39630
+    },
+    {
+      "epoch": 3.7530770687369817,
+      "grad_norm": 0.5581246614456177,
+      "learning_rate": 4.987691725052074e-06,
+      "loss": 0.0991,
+      "step": 39640
+    },
+    {
+      "epoch": 3.7540238591175914,
+      "grad_norm": 0.5363439321517944,
+      "learning_rate": 4.983904563529635e-06,
+      "loss": 0.0975,
+      "step": 39650
+    },
+    {
+      "epoch": 3.754970649498201,
+      "grad_norm": 0.45378658175468445,
+      "learning_rate": 4.980117402007197e-06,
+      "loss": 0.0937,
+      "step": 39660
+    },
+    {
+      "epoch": 3.755917439878811,
+      "grad_norm": 0.41703352332115173,
+      "learning_rate": 4.976330240484757e-06,
+      "loss": 0.0931,
+      "step": 39670
+    },
+    {
+      "epoch": 3.7568642302594206,
+      "grad_norm": 0.5335865020751953,
+      "learning_rate": 4.972543078962318e-06,
+      "loss": 0.1045,
+      "step": 39680
+    },
+    {
+      "epoch": 3.7578110206400304,
+      "grad_norm": 0.6401326060295105,
+      "learning_rate": 4.968755917439879e-06,
+      "loss": 0.0875,
+      "step": 39690
+    },
+    {
+      "epoch": 3.75875781102064,
+      "grad_norm": 0.5189956426620483,
+      "learning_rate": 4.96496875591744e-06,
+      "loss": 0.0914,
+      "step": 39700
+    },
+    {
+      "epoch": 3.75970460140125,
+      "grad_norm": 0.5471692085266113,
+      "learning_rate": 4.961181594395001e-06,
+      "loss": 0.1065,
+      "step": 39710
+    },
+    {
+      "epoch": 3.7606513917818596,
+      "grad_norm": 0.5299888253211975,
+      "learning_rate": 4.957394432872563e-06,
+      "loss": 0.1024,
+      "step": 39720
+    },
+    {
+      "epoch": 3.7615981821624693,
+      "grad_norm": 0.38434287905693054,
+      "learning_rate": 4.953607271350124e-06,
+      "loss": 0.0994,
+      "step": 39730
+    },
+    {
+      "epoch": 3.762544972543079,
+      "grad_norm": 0.45277974009513855,
+      "learning_rate": 4.949820109827685e-06,
+      "loss": 0.0953,
+      "step": 39740
+    },
+    {
+      "epoch": 3.763491762923689,
+      "grad_norm": 0.5209444761276245,
+      "learning_rate": 4.9460329483052455e-06,
+      "loss": 0.0917,
+      "step": 39750
+    },
+    {
+      "epoch": 3.7644385533042986,
+      "grad_norm": 0.37321043014526367,
+      "learning_rate": 4.942245786782806e-06,
+      "loss": 0.0951,
+      "step": 39760
+    },
+    {
+      "epoch": 3.7653853436849083,
+      "grad_norm": 0.5020799040794373,
+      "learning_rate": 4.938458625260368e-06,
+      "loss": 0.0967,
+      "step": 39770
+    },
+    {
+      "epoch": 3.766332134065518,
+      "grad_norm": 0.624764084815979,
+      "learning_rate": 4.934671463737929e-06,
+      "loss": 0.1014,
+      "step": 39780
+    },
+    {
+      "epoch": 3.7672789244461278,
+      "grad_norm": 0.4501070976257324,
+      "learning_rate": 4.93088430221549e-06,
+      "loss": 0.0932,
+      "step": 39790
+    },
+    {
+      "epoch": 3.7682257148267375,
+      "grad_norm": 0.4746125042438507,
+      "learning_rate": 4.927097140693051e-06,
+      "loss": 0.0907,
+      "step": 39800
+    },
+    {
+      "epoch": 3.7691725052073473,
+      "grad_norm": 0.5339996814727783,
+      "learning_rate": 4.923309979170612e-06,
+      "loss": 0.0976,
+      "step": 39810
+    },
+    {
+      "epoch": 3.770119295587957,
+      "grad_norm": 0.495650976896286,
+      "learning_rate": 4.919522817648173e-06,
+      "loss": 0.1006,
+      "step": 39820
+    },
+    {
+      "epoch": 3.7710660859685667,
+      "grad_norm": 0.4445793926715851,
+      "learning_rate": 4.915735656125734e-06,
+      "loss": 0.0875,
+      "step": 39830
+    },
+    {
+      "epoch": 3.7720128763491765,
+      "grad_norm": 0.3669644892215729,
+      "learning_rate": 4.911948494603295e-06,
+      "loss": 0.0893,
+      "step": 39840
+    },
+    {
+      "epoch": 3.772959666729786,
+      "grad_norm": 0.5043411254882812,
+      "learning_rate": 4.908161333080856e-06,
+      "loss": 0.103,
+      "step": 39850
+    },
+    {
+      "epoch": 3.773906457110396,
+      "grad_norm": 0.4899156987667084,
+      "learning_rate": 4.904374171558418e-06,
+      "loss": 0.1061,
+      "step": 39860
+    },
+    {
+      "epoch": 3.7748532474910057,
+      "grad_norm": 0.4879787564277649,
+      "learning_rate": 4.900587010035979e-06,
+      "loss": 0.0988,
+      "step": 39870
+    },
+    {
+      "epoch": 3.7758000378716154,
+      "grad_norm": 0.48512783646583557,
+      "learning_rate": 4.89679984851354e-06,
+      "loss": 0.1002,
+      "step": 39880
+    },
+    {
+      "epoch": 3.776746828252225,
+      "grad_norm": 0.5323088765144348,
+      "learning_rate": 4.893012686991101e-06,
+      "loss": 0.0943,
+      "step": 39890
+    },
+    {
+      "epoch": 3.777693618632835,
+      "grad_norm": 0.44328346848487854,
+      "learning_rate": 4.8892255254686615e-06,
+      "loss": 0.1019,
+      "step": 39900
+    },
+    {
+      "epoch": 3.7786404090134447,
+      "grad_norm": 0.4099396765232086,
+      "learning_rate": 4.885438363946222e-06,
+      "loss": 0.0961,
+      "step": 39910
+    },
+    {
+      "epoch": 3.7795871993940544,
+      "grad_norm": 0.4763758182525635,
+      "learning_rate": 4.881651202423784e-06,
+      "loss": 0.0974,
+      "step": 39920
+    },
+    {
+      "epoch": 3.780533989774664,
+      "grad_norm": 0.417901873588562,
+      "learning_rate": 4.877864040901345e-06,
+      "loss": 0.1075,
+      "step": 39930
+    },
+    {
+      "epoch": 3.781480780155274,
+      "grad_norm": 0.5062112808227539,
+      "learning_rate": 4.874076879378906e-06,
+      "loss": 0.0879,
+      "step": 39940
+    },
+    {
+      "epoch": 3.7824275705358836,
+      "grad_norm": 0.4354439377784729,
+      "learning_rate": 4.870289717856467e-06,
+      "loss": 0.1048,
+      "step": 39950
+    },
+    {
+      "epoch": 3.783374360916493,
+      "grad_norm": 0.3767815828323364,
+      "learning_rate": 4.866502556334028e-06,
+      "loss": 0.0985,
+      "step": 39960
+    },
+    {
+      "epoch": 3.7843211512971027,
+      "grad_norm": 0.48620229959487915,
+      "learning_rate": 4.8627153948115895e-06,
+      "loss": 0.1029,
+      "step": 39970
+    },
+    {
+      "epoch": 3.7852679416777124,
+      "grad_norm": 0.40215253829956055,
+      "learning_rate": 4.85892823328915e-06,
+      "loss": 0.0974,
+      "step": 39980
+    },
+    {
+      "epoch": 3.786214732058322,
+      "grad_norm": 0.43390804529190063,
+      "learning_rate": 4.855141071766711e-06,
+      "loss": 0.1016,
+      "step": 39990
+    },
+    {
+      "epoch": 3.787161522438932,
+      "grad_norm": 0.48430293798446655,
+      "learning_rate": 4.851353910244272e-06,
+      "loss": 0.0981,
+      "step": 40000
+    },
+    {
+      "epoch": 3.7881083128195416,
+      "grad_norm": 0.4889768362045288,
+      "learning_rate": 4.847566748721834e-06,
+      "loss": 0.0992,
+      "step": 40010
+    },
+    {
+      "epoch": 3.7890551032001514,
+      "grad_norm": 0.5483484268188477,
+      "learning_rate": 4.843779587199394e-06,
+      "loss": 0.096,
+      "step": 40020
+    },
+    {
+      "epoch": 3.790001893580761,
+      "grad_norm": 0.4289323687553406,
+      "learning_rate": 4.839992425676956e-06,
+      "loss": 0.0913,
+      "step": 40030
+    },
+    {
+      "epoch": 3.790948683961371,
+      "grad_norm": 0.5495353937149048,
+      "learning_rate": 4.836205264154517e-06,
+      "loss": 0.1083,
+      "step": 40040
+    },
+    {
+      "epoch": 3.7918954743419806,
+      "grad_norm": 0.5764693021774292,
+      "learning_rate": 4.8324181026320775e-06,
+      "loss": 0.0974,
+      "step": 40050
+    },
+    {
+      "epoch": 3.7928422647225903,
+      "grad_norm": 0.46960946917533875,
+      "learning_rate": 4.828630941109639e-06,
+      "loss": 0.0879,
+      "step": 40060
+    },
+    {
+      "epoch": 3.7937890551032,
+      "grad_norm": 0.4713224768638611,
+      "learning_rate": 4.8248437795872e-06,
+      "loss": 0.0894,
+      "step": 40070
+    },
+    {
+      "epoch": 3.79473584548381,
+      "grad_norm": 0.5814439654350281,
+      "learning_rate": 4.821056618064761e-06,
+      "loss": 0.0995,
+      "step": 40080
+    },
+    {
+      "epoch": 3.7956826358644196,
+      "grad_norm": 0.5569701790809631,
+      "learning_rate": 4.817269456542322e-06,
+      "loss": 0.103,
+      "step": 40090
+    },
+    {
+      "epoch": 3.7966294262450293,
+      "grad_norm": 0.4573192894458771,
+      "learning_rate": 4.813482295019883e-06,
+      "loss": 0.0901,
+      "step": 40100
+    },
+    {
+      "epoch": 3.797576216625639,
+      "grad_norm": 0.5137414932250977,
+      "learning_rate": 4.809695133497444e-06,
+      "loss": 0.0958,
+      "step": 40110
+    },
+    {
+      "epoch": 3.798523007006249,
+      "grad_norm": 0.5065782070159912,
+      "learning_rate": 4.8059079719750055e-06,
+      "loss": 0.1015,
+      "step": 40120
+    },
+    {
+      "epoch": 3.7994697973868585,
+      "grad_norm": 0.512101411819458,
+      "learning_rate": 4.802120810452566e-06,
+      "loss": 0.0904,
+      "step": 40130
+    },
+    {
+      "epoch": 3.8004165877674683,
+      "grad_norm": 0.4687511920928955,
+      "learning_rate": 4.798333648930127e-06,
+      "loss": 0.0917,
+      "step": 40140
+    },
+    {
+      "epoch": 3.801363378148078,
+      "grad_norm": 0.5853365063667297,
+      "learning_rate": 4.794546487407689e-06,
+      "loss": 0.1092,
+      "step": 40150
+    },
+    {
+      "epoch": 3.8023101685286878,
+      "grad_norm": 0.4754956066608429,
+      "learning_rate": 4.790759325885249e-06,
+      "loss": 0.0892,
+      "step": 40160
+    },
+    {
+      "epoch": 3.8032569589092975,
+      "grad_norm": 0.48234477639198303,
+      "learning_rate": 4.78697216436281e-06,
+      "loss": 0.1014,
+      "step": 40170
+    },
+    {
+      "epoch": 3.8042037492899072,
+      "grad_norm": 0.4432053864002228,
+      "learning_rate": 4.783185002840372e-06,
+      "loss": 0.0928,
+      "step": 40180
+    },
+    {
+      "epoch": 3.805150539670517,
+      "grad_norm": 0.6404855847358704,
+      "learning_rate": 4.779397841317933e-06,
+      "loss": 0.1047,
+      "step": 40190
+    },
+    {
+      "epoch": 3.8060973300511267,
+      "grad_norm": 0.509357750415802,
+      "learning_rate": 4.7756106797954935e-06,
+      "loss": 0.1058,
+      "step": 40200
+    },
+    {
+      "epoch": 3.8070441204317365,
+      "grad_norm": 0.6530634760856628,
+      "learning_rate": 4.771823518273055e-06,
+      "loss": 0.0989,
+      "step": 40210
+    },
+    {
+      "epoch": 3.807990910812346,
+      "grad_norm": 0.5133205652236938,
+      "learning_rate": 4.768036356750616e-06,
+      "loss": 0.1008,
+      "step": 40220
+    },
+    {
+      "epoch": 3.808937701192956,
+      "grad_norm": 0.4522642493247986,
+      "learning_rate": 4.764249195228177e-06,
+      "loss": 0.0926,
+      "step": 40230
+    },
+    {
+      "epoch": 3.8098844915735657,
+      "grad_norm": 0.5123722553253174,
+      "learning_rate": 4.760462033705738e-06,
+      "loss": 0.0945,
+      "step": 40240
+    },
+    {
+      "epoch": 3.8108312819541754,
+      "grad_norm": 0.5599709749221802,
+      "learning_rate": 4.756674872183299e-06,
+      "loss": 0.0934,
+      "step": 40250
+    },
+    {
+      "epoch": 3.811778072334785,
+      "grad_norm": 0.5186089873313904,
+      "learning_rate": 4.75288771066086e-06,
+      "loss": 0.0897,
+      "step": 40260
+    },
+    {
+      "epoch": 3.812724862715395,
+      "grad_norm": 0.4753962755203247,
+      "learning_rate": 4.7491005491384215e-06,
+      "loss": 0.0854,
+      "step": 40270
+    },
+    {
+      "epoch": 3.8136716530960046,
+      "grad_norm": 0.5974533557891846,
+      "learning_rate": 4.745313387615982e-06,
+      "loss": 0.1008,
+      "step": 40280
+    },
+    {
+      "epoch": 3.8146184434766144,
+      "grad_norm": 0.39401939511299133,
+      "learning_rate": 4.741526226093543e-06,
+      "loss": 0.0982,
+      "step": 40290
+    },
+    {
+      "epoch": 3.815565233857224,
+      "grad_norm": 0.3924501836299896,
+      "learning_rate": 4.737739064571104e-06,
+      "loss": 0.0974,
+      "step": 40300
+    },
+    {
+      "epoch": 3.816512024237834,
+      "grad_norm": 0.4209947884082794,
+      "learning_rate": 4.733951903048665e-06,
+      "loss": 0.0874,
+      "step": 40310
+    },
+    {
+      "epoch": 3.8174588146184436,
+      "grad_norm": 0.3790121078491211,
+      "learning_rate": 4.730164741526227e-06,
+      "loss": 0.0893,
+      "step": 40320
+    },
+    {
+      "epoch": 3.8184056049990533,
+      "grad_norm": 0.4306885302066803,
+      "learning_rate": 4.726377580003788e-06,
+      "loss": 0.094,
+      "step": 40330
+    },
+    {
+      "epoch": 3.819352395379663,
+      "grad_norm": 0.49594056606292725,
+      "learning_rate": 4.7225904184813486e-06,
+      "loss": 0.091,
+      "step": 40340
+    },
+    {
+      "epoch": 3.820299185760273,
+      "grad_norm": 0.47653913497924805,
+      "learning_rate": 4.7188032569589095e-06,
+      "loss": 0.0933,
+      "step": 40350
+    },
+    {
+      "epoch": 3.8212459761408826,
+      "grad_norm": 0.5557363629341125,
+      "learning_rate": 4.715016095436471e-06,
+      "loss": 0.0967,
+      "step": 40360
+    },
+    {
+      "epoch": 3.822192766521492,
+      "grad_norm": 0.5122710466384888,
+      "learning_rate": 4.711228933914031e-06,
+      "loss": 0.0988,
+      "step": 40370
+    },
+    {
+      "epoch": 3.8231395569021016,
+      "grad_norm": 0.4322662353515625,
+      "learning_rate": 4.707441772391593e-06,
+      "loss": 0.1034,
+      "step": 40380
+    },
+    {
+      "epoch": 3.8240863472827114,
+      "grad_norm": 0.4525582492351532,
+      "learning_rate": 4.703654610869154e-06,
+      "loss": 0.0991,
+      "step": 40390
+    },
+    {
+      "epoch": 3.825033137663321,
+      "grad_norm": 0.473119854927063,
+      "learning_rate": 4.699867449346715e-06,
+      "loss": 0.0967,
+      "step": 40400
+    },
+    {
+      "epoch": 3.825979928043931,
+      "grad_norm": 0.43417078256607056,
+      "learning_rate": 4.6960802878242765e-06,
+      "loss": 0.0953,
+      "step": 40410
+    },
+    {
+      "epoch": 3.8269267184245406,
+      "grad_norm": 0.4016440510749817,
+      "learning_rate": 4.6922931263018374e-06,
+      "loss": 0.0943,
+      "step": 40420
+    },
+    {
+      "epoch": 3.8278735088051503,
+      "grad_norm": 0.4091476798057556,
+      "learning_rate": 4.688505964779398e-06,
+      "loss": 0.0974,
+      "step": 40430
+    },
+    {
+      "epoch": 3.82882029918576,
+      "grad_norm": 0.44425806403160095,
+      "learning_rate": 4.684718803256959e-06,
+      "loss": 0.0925,
+      "step": 40440
+    },
+    {
+      "epoch": 3.82976708956637,
+      "grad_norm": 0.4543251693248749,
+      "learning_rate": 4.68093164173452e-06,
+      "loss": 0.0906,
+      "step": 40450
+    },
+    {
+      "epoch": 3.8307138799469795,
+      "grad_norm": 0.4300362169742584,
+      "learning_rate": 4.677144480212081e-06,
+      "loss": 0.0908,
+      "step": 40460
+    },
+    {
+      "epoch": 3.8316606703275893,
+      "grad_norm": 0.5083745718002319,
+      "learning_rate": 4.673357318689643e-06,
+      "loss": 0.0934,
+      "step": 40470
+    },
+    {
+      "epoch": 3.832607460708199,
+      "grad_norm": 0.45083117485046387,
+      "learning_rate": 4.669570157167204e-06,
+      "loss": 0.0906,
+      "step": 40480
+    },
+    {
+      "epoch": 3.8335542510888088,
+      "grad_norm": 0.43083450198173523,
+      "learning_rate": 4.6657829956447646e-06,
+      "loss": 0.0904,
+      "step": 40490
+    },
+    {
+      "epoch": 3.8345010414694185,
+      "grad_norm": 0.48674798011779785,
+      "learning_rate": 4.661995834122326e-06,
+      "loss": 0.0954,
+      "step": 40500
+    },
+    {
+      "epoch": 3.8354478318500282,
+      "grad_norm": 0.5845592021942139,
+      "learning_rate": 4.658208672599886e-06,
+      "loss": 0.1006,
+      "step": 40510
+    },
+    {
+      "epoch": 3.836394622230638,
+      "grad_norm": 0.4620465934276581,
+      "learning_rate": 4.654421511077447e-06,
+      "loss": 0.0905,
+      "step": 40520
+    },
+    {
+      "epoch": 3.8373414126112477,
+      "grad_norm": 0.5082583427429199,
+      "learning_rate": 4.650634349555009e-06,
+      "loss": 0.1009,
+      "step": 40530
+    },
+    {
+      "epoch": 3.8382882029918575,
+      "grad_norm": 0.4657653272151947,
+      "learning_rate": 4.64684718803257e-06,
+      "loss": 0.0937,
+      "step": 40540
+    },
+    {
+      "epoch": 3.839234993372467,
+      "grad_norm": 0.4716366231441498,
+      "learning_rate": 4.643060026510131e-06,
+      "loss": 0.0924,
+      "step": 40550
+    },
+    {
+      "epoch": 3.840181783753077,
+      "grad_norm": 0.5509023666381836,
+      "learning_rate": 4.6392728649876925e-06,
+      "loss": 0.0934,
+      "step": 40560
+    },
+    {
+      "epoch": 3.8411285741336867,
+      "grad_norm": 0.4932687282562256,
+      "learning_rate": 4.6354857034652534e-06,
+      "loss": 0.0921,
+      "step": 40570
+    },
+    {
+      "epoch": 3.8420753645142964,
+      "grad_norm": 0.5577378869056702,
+      "learning_rate": 4.631698541942814e-06,
+      "loss": 0.0957,
+      "step": 40580
+    },
+    {
+      "epoch": 3.843022154894906,
+      "grad_norm": 0.4173696041107178,
+      "learning_rate": 4.627911380420375e-06,
+      "loss": 0.0981,
+      "step": 40590
+    },
+    {
+      "epoch": 3.843968945275516,
+      "grad_norm": 0.5011758208274841,
+      "learning_rate": 4.624124218897936e-06,
+      "loss": 0.0972,
+      "step": 40600
+    },
+    {
+      "epoch": 3.8449157356561257,
+      "grad_norm": 0.4696127772331238,
+      "learning_rate": 4.620337057375497e-06,
+      "loss": 0.0893,
+      "step": 40610
+    },
+    {
+      "epoch": 3.8458625260367354,
+      "grad_norm": 0.4647819697856903,
+      "learning_rate": 4.616549895853059e-06,
+      "loss": 0.0876,
+      "step": 40620
+    },
+    {
+      "epoch": 3.846809316417345,
+      "grad_norm": 0.4026613235473633,
+      "learning_rate": 4.61276273433062e-06,
+      "loss": 0.0959,
+      "step": 40630
+    },
+    {
+      "epoch": 3.847756106797955,
+      "grad_norm": 0.5512098670005798,
+      "learning_rate": 4.6089755728081806e-06,
+      "loss": 0.0947,
+      "step": 40640
+    },
+    {
+      "epoch": 3.8487028971785646,
+      "grad_norm": 0.44850537180900574,
+      "learning_rate": 4.6051884112857414e-06,
+      "loss": 0.0863,
+      "step": 40650
+    },
+    {
+      "epoch": 3.8496496875591744,
+      "grad_norm": 0.4348522126674652,
+      "learning_rate": 4.601401249763302e-06,
+      "loss": 0.1017,
+      "step": 40660
+    },
+    {
+      "epoch": 3.850596477939784,
+      "grad_norm": 0.5444629192352295,
+      "learning_rate": 4.597614088240864e-06,
+      "loss": 0.0927,
+      "step": 40670
+    },
+    {
+      "epoch": 3.851543268320394,
+      "grad_norm": 0.5211396813392639,
+      "learning_rate": 4.593826926718425e-06,
+      "loss": 0.1035,
+      "step": 40680
+    },
+    {
+      "epoch": 3.8524900587010036,
+      "grad_norm": 0.4562847912311554,
+      "learning_rate": 4.590039765195986e-06,
+      "loss": 0.089,
+      "step": 40690
+    },
+    {
+      "epoch": 3.8534368490816133,
+      "grad_norm": 0.46698418259620667,
+      "learning_rate": 4.586252603673547e-06,
+      "loss": 0.1002,
+      "step": 40700
+    },
+    {
+      "epoch": 3.854383639462223,
+      "grad_norm": 0.421149879693985,
+      "learning_rate": 4.5824654421511085e-06,
+      "loss": 0.0984,
+      "step": 40710
+    },
+    {
+      "epoch": 3.855330429842833,
+      "grad_norm": 0.4744303822517395,
+      "learning_rate": 4.5786782806286686e-06,
+      "loss": 0.0939,
+      "step": 40720
+    },
+    {
+      "epoch": 3.8562772202234425,
+      "grad_norm": 0.498382568359375,
+      "learning_rate": 4.57489111910623e-06,
+      "loss": 0.1012,
+      "step": 40730
+    },
+    {
+      "epoch": 3.8572240106040523,
+      "grad_norm": 0.4662265479564667,
+      "learning_rate": 4.571103957583791e-06,
+      "loss": 0.1025,
+      "step": 40740
+    },
+    {
+      "epoch": 3.858170800984662,
+      "grad_norm": 0.4596439003944397,
+      "learning_rate": 4.567316796061352e-06,
+      "loss": 0.1004,
+      "step": 40750
+    },
+    {
+      "epoch": 3.8591175913652718,
+      "grad_norm": 0.4291379153728485,
+      "learning_rate": 4.563529634538914e-06,
+      "loss": 0.0911,
+      "step": 40760
+    },
+    {
+      "epoch": 3.8600643817458815,
+      "grad_norm": 0.5993068218231201,
+      "learning_rate": 4.559742473016475e-06,
+      "loss": 0.1015,
+      "step": 40770
+    },
+    {
+      "epoch": 3.8610111721264913,
+      "grad_norm": 0.4954267740249634,
+      "learning_rate": 4.555955311494036e-06,
+      "loss": 0.0964,
+      "step": 40780
+    },
+    {
+      "epoch": 3.861957962507101,
+      "grad_norm": 0.5256617069244385,
+      "learning_rate": 4.5521681499715965e-06,
+      "loss": 0.1025,
+      "step": 40790
+    },
+    {
+      "epoch": 3.8629047528877107,
+      "grad_norm": 0.4482620060443878,
+      "learning_rate": 4.5483809884491574e-06,
+      "loss": 0.092,
+      "step": 40800
+    },
+    {
+      "epoch": 3.8638515432683205,
+      "grad_norm": 0.4422076344490051,
+      "learning_rate": 4.544593826926718e-06,
+      "loss": 0.0924,
+      "step": 40810
+    },
+    {
+      "epoch": 3.86479833364893,
+      "grad_norm": 0.4300210773944855,
+      "learning_rate": 4.54080666540428e-06,
+      "loss": 0.0921,
+      "step": 40820
+    },
+    {
+      "epoch": 3.86574512402954,
+      "grad_norm": 0.6585009694099426,
+      "learning_rate": 4.537019503881841e-06,
+      "loss": 0.0913,
+      "step": 40830
+    },
+    {
+      "epoch": 3.8666919144101497,
+      "grad_norm": 0.4743376672267914,
+      "learning_rate": 4.533232342359402e-06,
+      "loss": 0.0921,
+      "step": 40840
+    },
+    {
+      "epoch": 3.8676387047907594,
+      "grad_norm": 0.49737921357154846,
+      "learning_rate": 4.529445180836964e-06,
+      "loss": 0.1085,
+      "step": 40850
+    },
+    {
+      "epoch": 3.868585495171369,
+      "grad_norm": 0.49208909273147583,
+      "learning_rate": 4.525658019314524e-06,
+      "loss": 0.0999,
+      "step": 40860
+    },
+    {
+      "epoch": 3.869532285551979,
+      "grad_norm": 0.5970167517662048,
+      "learning_rate": 4.521870857792085e-06,
+      "loss": 0.1033,
+      "step": 40870
+    },
+    {
+      "epoch": 3.8704790759325887,
+      "grad_norm": 0.5238867998123169,
+      "learning_rate": 4.518083696269646e-06,
+      "loss": 0.0958,
+      "step": 40880
+    },
+    {
+      "epoch": 3.8714258663131984,
+      "grad_norm": 0.4827011227607727,
+      "learning_rate": 4.514296534747207e-06,
+      "loss": 0.1031,
+      "step": 40890
+    },
+    {
+      "epoch": 3.872372656693808,
+      "grad_norm": 0.5186099410057068,
+      "learning_rate": 4.510509373224768e-06,
+      "loss": 0.1,
+      "step": 40900
+    },
+    {
+      "epoch": 3.873319447074418,
+      "grad_norm": 0.41698890924453735,
+      "learning_rate": 4.50672221170233e-06,
+      "loss": 0.0982,
+      "step": 40910
+    },
+    {
+      "epoch": 3.8742662374550276,
+      "grad_norm": 0.5530847311019897,
+      "learning_rate": 4.502935050179891e-06,
+      "loss": 0.0983,
+      "step": 40920
+    },
+    {
+      "epoch": 3.8752130278356374,
+      "grad_norm": 0.519140899181366,
+      "learning_rate": 4.499147888657452e-06,
+      "loss": 0.0957,
+      "step": 40930
+    },
+    {
+      "epoch": 3.876159818216247,
+      "grad_norm": 0.46212148666381836,
+      "learning_rate": 4.4953607271350125e-06,
+      "loss": 0.0904,
+      "step": 40940
+    },
+    {
+      "epoch": 3.877106608596857,
+      "grad_norm": 0.5408146381378174,
+      "learning_rate": 4.4915735656125734e-06,
+      "loss": 0.1085,
+      "step": 40950
+    },
+    {
+      "epoch": 3.8780533989774666,
+      "grad_norm": 0.5476594567298889,
+      "learning_rate": 4.487786404090135e-06,
+      "loss": 0.0984,
+      "step": 40960
+    },
+    {
+      "epoch": 3.8790001893580763,
+      "grad_norm": 0.45062705874443054,
+      "learning_rate": 4.483999242567696e-06,
+      "loss": 0.0981,
+      "step": 40970
+    },
+    {
+      "epoch": 3.879946979738686,
+      "grad_norm": 0.47842487692832947,
+      "learning_rate": 4.480212081045257e-06,
+      "loss": 0.0979,
+      "step": 40980
+    },
+    {
+      "epoch": 3.880893770119296,
+      "grad_norm": 0.47004956007003784,
+      "learning_rate": 4.476424919522818e-06,
+      "loss": 0.0996,
+      "step": 40990
+    },
+    {
+      "epoch": 3.8818405604999056,
+      "grad_norm": 0.5024536848068237,
+      "learning_rate": 4.47263775800038e-06,
+      "loss": 0.0903,
+      "step": 41000
+    },
+    {
+      "epoch": 3.8827873508805153,
+      "grad_norm": 0.4794340133666992,
+      "learning_rate": 4.46885059647794e-06,
+      "loss": 0.0961,
+      "step": 41010
+    },
+    {
+      "epoch": 3.883734141261125,
+      "grad_norm": 0.4689330756664276,
+      "learning_rate": 4.465063434955501e-06,
+      "loss": 0.1038,
+      "step": 41020
+    },
+    {
+      "epoch": 3.884680931641735,
+      "grad_norm": 0.6886982321739197,
+      "learning_rate": 4.461276273433062e-06,
+      "loss": 0.1095,
+      "step": 41030
+    },
+    {
+      "epoch": 3.8856277220223445,
+      "grad_norm": 0.48080509901046753,
+      "learning_rate": 4.457489111910623e-06,
+      "loss": 0.0917,
+      "step": 41040
+    },
+    {
+      "epoch": 3.8865745124029543,
+      "grad_norm": 0.4674771726131439,
+      "learning_rate": 4.453701950388185e-06,
+      "loss": 0.0889,
+      "step": 41050
+    },
+    {
+      "epoch": 3.8875213027835636,
+      "grad_norm": 0.6061664819717407,
+      "learning_rate": 4.449914788865746e-06,
+      "loss": 0.0952,
+      "step": 41060
+    },
+    {
+      "epoch": 3.8884680931641733,
+      "grad_norm": 0.5249764323234558,
+      "learning_rate": 4.446127627343307e-06,
+      "loss": 0.1027,
+      "step": 41070
+    },
+    {
+      "epoch": 3.889414883544783,
+      "grad_norm": 0.5209566354751587,
+      "learning_rate": 4.442340465820868e-06,
+      "loss": 0.0985,
+      "step": 41080
+    },
+    {
+      "epoch": 3.890361673925393,
+      "grad_norm": 0.6610578894615173,
+      "learning_rate": 4.4385533042984285e-06,
+      "loss": 0.1029,
+      "step": 41090
+    },
+    {
+      "epoch": 3.8913084643060025,
+      "grad_norm": 0.46757304668426514,
+      "learning_rate": 4.434766142775989e-06,
+      "loss": 0.0956,
+      "step": 41100
+    },
+    {
+      "epoch": 3.8922552546866123,
+      "grad_norm": 0.47179147601127625,
+      "learning_rate": 4.430978981253551e-06,
+      "loss": 0.106,
+      "step": 41110
+    },
+    {
+      "epoch": 3.893202045067222,
+      "grad_norm": 0.6236637830734253,
+      "learning_rate": 4.427191819731112e-06,
+      "loss": 0.1017,
+      "step": 41120
+    },
+    {
+      "epoch": 3.8941488354478317,
+      "grad_norm": 0.48480746150016785,
+      "learning_rate": 4.423404658208673e-06,
+      "loss": 0.0922,
+      "step": 41130
+    },
+    {
+      "epoch": 3.8950956258284415,
+      "grad_norm": 0.4346686601638794,
+      "learning_rate": 4.419617496686235e-06,
+      "loss": 0.0944,
+      "step": 41140
+    },
+    {
+      "epoch": 3.8960424162090512,
+      "grad_norm": 0.39827197790145874,
+      "learning_rate": 4.415830335163795e-06,
+      "loss": 0.0913,
+      "step": 41150
+    },
+    {
+      "epoch": 3.896989206589661,
+      "grad_norm": 0.395149827003479,
+      "learning_rate": 4.412043173641356e-06,
+      "loss": 0.0964,
+      "step": 41160
+    },
+    {
+      "epoch": 3.8979359969702707,
+      "grad_norm": 0.47335392236709595,
+      "learning_rate": 4.408256012118917e-06,
+      "loss": 0.0892,
+      "step": 41170
+    },
+    {
+      "epoch": 3.8988827873508805,
+      "grad_norm": 0.47912123799324036,
+      "learning_rate": 4.404468850596478e-06,
+      "loss": 0.0965,
+      "step": 41180
+    },
+    {
+      "epoch": 3.89982957773149,
+      "grad_norm": 0.4129624664783478,
+      "learning_rate": 4.400681689074039e-06,
+      "loss": 0.0887,
+      "step": 41190
+    },
+    {
+      "epoch": 3.9007763681121,
+      "grad_norm": 0.5240882635116577,
+      "learning_rate": 4.396894527551601e-06,
+      "loss": 0.0967,
+      "step": 41200
+    },
+    {
+      "epoch": 3.9017231584927097,
+      "grad_norm": 0.5136657953262329,
+      "learning_rate": 4.393107366029162e-06,
+      "loss": 0.0956,
+      "step": 41210
+    },
+    {
+      "epoch": 3.9026699488733194,
+      "grad_norm": 0.5962952971458435,
+      "learning_rate": 4.389320204506723e-06,
+      "loss": 0.0984,
+      "step": 41220
+    },
+    {
+      "epoch": 3.903616739253929,
+      "grad_norm": 0.37900274991989136,
+      "learning_rate": 4.385533042984284e-06,
+      "loss": 0.092,
+      "step": 41230
+    },
+    {
+      "epoch": 3.904563529634539,
+      "grad_norm": 0.5577122569084167,
+      "learning_rate": 4.3817458814618445e-06,
+      "loss": 0.1054,
+      "step": 41240
+    },
+    {
+      "epoch": 3.9055103200151486,
+      "grad_norm": 0.4894963800907135,
+      "learning_rate": 4.377958719939405e-06,
+      "loss": 0.1018,
+      "step": 41250
+    },
+    {
+      "epoch": 3.9064571103957584,
+      "grad_norm": 0.43864747881889343,
+      "learning_rate": 4.374171558416967e-06,
+      "loss": 0.0965,
+      "step": 41260
+    },
+    {
+      "epoch": 3.907403900776368,
+      "grad_norm": 0.45537394285202026,
+      "learning_rate": 4.370384396894528e-06,
+      "loss": 0.0887,
+      "step": 41270
+    },
+    {
+      "epoch": 3.908350691156978,
+      "grad_norm": 0.5147911310195923,
+      "learning_rate": 4.366597235372089e-06,
+      "loss": 0.0912,
+      "step": 41280
+    },
+    {
+      "epoch": 3.9092974815375876,
+      "grad_norm": 0.5520839691162109,
+      "learning_rate": 4.36281007384965e-06,
+      "loss": 0.099,
+      "step": 41290
+    },
+    {
+      "epoch": 3.9102442719181973,
+      "grad_norm": 0.5957738757133484,
+      "learning_rate": 4.359022912327211e-06,
+      "loss": 0.1005,
+      "step": 41300
+    },
+    {
+      "epoch": 3.911191062298807,
+      "grad_norm": 0.5196476578712463,
+      "learning_rate": 4.3552357508047725e-06,
+      "loss": 0.1033,
+      "step": 41310
+    },
+    {
+      "epoch": 3.912137852679417,
+      "grad_norm": 0.59003746509552,
+      "learning_rate": 4.351448589282333e-06,
+      "loss": 0.0987,
+      "step": 41320
+    },
+    {
+      "epoch": 3.9130846430600266,
+      "grad_norm": 0.5245180726051331,
+      "learning_rate": 4.347661427759894e-06,
+      "loss": 0.0974,
+      "step": 41330
+    },
+    {
+      "epoch": 3.9140314334406363,
+      "grad_norm": 0.5251783728599548,
+      "learning_rate": 4.343874266237455e-06,
+      "loss": 0.0985,
+      "step": 41340
+    },
+    {
+      "epoch": 3.914978223821246,
+      "grad_norm": 0.46796226501464844,
+      "learning_rate": 4.340087104715017e-06,
+      "loss": 0.1035,
+      "step": 41350
+    },
+    {
+      "epoch": 3.915925014201856,
+      "grad_norm": 0.5118393898010254,
+      "learning_rate": 4.336299943192577e-06,
+      "loss": 0.0924,
+      "step": 41360
+    },
+    {
+      "epoch": 3.9168718045824655,
+      "grad_norm": 0.5003322958946228,
+      "learning_rate": 4.332512781670139e-06,
+      "loss": 0.0982,
+      "step": 41370
+    },
+    {
+      "epoch": 3.9178185949630753,
+      "grad_norm": 0.6285452246665955,
+      "learning_rate": 4.3287256201477e-06,
+      "loss": 0.0941,
+      "step": 41380
+    },
+    {
+      "epoch": 3.918765385343685,
+      "grad_norm": 0.4589633345603943,
+      "learning_rate": 4.3249384586252605e-06,
+      "loss": 0.0941,
+      "step": 41390
+    },
+    {
+      "epoch": 3.9197121757242948,
+      "grad_norm": 0.48504549264907837,
+      "learning_rate": 4.321151297102822e-06,
+      "loss": 0.0986,
+      "step": 41400
+    },
+    {
+      "epoch": 3.9206589661049045,
+      "grad_norm": 0.5638332366943359,
+      "learning_rate": 4.317364135580383e-06,
+      "loss": 0.1092,
+      "step": 41410
+    },
+    {
+      "epoch": 3.9216057564855142,
+      "grad_norm": 0.42241472005844116,
+      "learning_rate": 4.313576974057944e-06,
+      "loss": 0.1062,
+      "step": 41420
+    },
+    {
+      "epoch": 3.922552546866124,
+      "grad_norm": 0.46967965364456177,
+      "learning_rate": 4.309789812535505e-06,
+      "loss": 0.0994,
+      "step": 41430
+    },
+    {
+      "epoch": 3.9234993372467337,
+      "grad_norm": 0.4713064432144165,
+      "learning_rate": 4.306002651013066e-06,
+      "loss": 0.1011,
+      "step": 41440
+    },
+    {
+      "epoch": 3.9244461276273435,
+      "grad_norm": 0.5808030366897583,
+      "learning_rate": 4.302215489490627e-06,
+      "loss": 0.0967,
+      "step": 41450
+    },
+    {
+      "epoch": 3.925392918007953,
+      "grad_norm": 0.5431656241416931,
+      "learning_rate": 4.2984283279681885e-06,
+      "loss": 0.0967,
+      "step": 41460
+    },
+    {
+      "epoch": 3.9263397083885625,
+      "grad_norm": 0.5076019167900085,
+      "learning_rate": 4.294641166445749e-06,
+      "loss": 0.0957,
+      "step": 41470
+    },
+    {
+      "epoch": 3.9272864987691722,
+      "grad_norm": 0.6491959691047668,
+      "learning_rate": 4.29085400492331e-06,
+      "loss": 0.0995,
+      "step": 41480
+    },
+    {
+      "epoch": 3.928233289149782,
+      "grad_norm": 0.46067243814468384,
+      "learning_rate": 4.287066843400872e-06,
+      "loss": 0.0989,
+      "step": 41490
+    },
+    {
+      "epoch": 3.9291800795303917,
+      "grad_norm": 0.5454583764076233,
+      "learning_rate": 4.283279681878432e-06,
+      "loss": 0.1021,
+      "step": 41500
+    },
+    {
+      "epoch": 3.9301268699110015,
+      "grad_norm": 0.6414952278137207,
+      "learning_rate": 4.279492520355994e-06,
+      "loss": 0.0944,
+      "step": 41510
+    },
+    {
+      "epoch": 3.931073660291611,
+      "grad_norm": 0.4030686914920807,
+      "learning_rate": 4.275705358833555e-06,
+      "loss": 0.0901,
+      "step": 41520
+    },
+    {
+      "epoch": 3.932020450672221,
+      "grad_norm": 0.4818320572376251,
+      "learning_rate": 4.271918197311116e-06,
+      "loss": 0.0925,
+      "step": 41530
+    },
+    {
+      "epoch": 3.9329672410528307,
+      "grad_norm": 0.5245049595832825,
+      "learning_rate": 4.2681310357886765e-06,
+      "loss": 0.0972,
+      "step": 41540
+    },
+    {
+      "epoch": 3.9339140314334404,
+      "grad_norm": 0.492740273475647,
+      "learning_rate": 4.264343874266238e-06,
+      "loss": 0.0995,
+      "step": 41550
+    },
+    {
+      "epoch": 3.93486082181405,
+      "grad_norm": 0.5344408750534058,
+      "learning_rate": 4.260556712743799e-06,
+      "loss": 0.0982,
+      "step": 41560
+    },
+    {
+      "epoch": 3.93580761219466,
+      "grad_norm": 0.4735774099826813,
+      "learning_rate": 4.25676955122136e-06,
+      "loss": 0.0895,
+      "step": 41570
+    },
+    {
+      "epoch": 3.9367544025752697,
+      "grad_norm": 0.4924392104148865,
+      "learning_rate": 4.252982389698921e-06,
+      "loss": 0.0966,
+      "step": 41580
+    },
+    {
+      "epoch": 3.9377011929558794,
+      "grad_norm": 0.4154967963695526,
+      "learning_rate": 4.249195228176482e-06,
+      "loss": 0.0941,
+      "step": 41590
+    },
+    {
+      "epoch": 3.938647983336489,
+      "grad_norm": 0.5360844731330872,
+      "learning_rate": 4.2454080666540436e-06,
+      "loss": 0.0954,
+      "step": 41600
+    },
+    {
+      "epoch": 3.939594773717099,
+      "grad_norm": 0.5082783102989197,
+      "learning_rate": 4.2416209051316045e-06,
+      "loss": 0.0958,
+      "step": 41610
+    },
+    {
+      "epoch": 3.9405415640977086,
+      "grad_norm": 0.5777152180671692,
+      "learning_rate": 4.237833743609165e-06,
+      "loss": 0.1034,
+      "step": 41620
+    },
+    {
+      "epoch": 3.9414883544783184,
+      "grad_norm": 0.40907520055770874,
+      "learning_rate": 4.234046582086726e-06,
+      "loss": 0.0997,
+      "step": 41630
+    },
+    {
+      "epoch": 3.942435144858928,
+      "grad_norm": 0.34812426567077637,
+      "learning_rate": 4.230259420564287e-06,
+      "loss": 0.0912,
+      "step": 41640
+    },
+    {
+      "epoch": 3.943381935239538,
+      "grad_norm": 0.5555279850959778,
+      "learning_rate": 4.226472259041848e-06,
+      "loss": 0.0933,
+      "step": 41650
+    },
+    {
+      "epoch": 3.9443287256201476,
+      "grad_norm": 0.45173120498657227,
+      "learning_rate": 4.22268509751941e-06,
+      "loss": 0.09,
+      "step": 41660
+    },
+    {
+      "epoch": 3.9452755160007573,
+      "grad_norm": 0.4487028121948242,
+      "learning_rate": 4.218897935996971e-06,
+      "loss": 0.0951,
+      "step": 41670
+    },
+    {
+      "epoch": 3.946222306381367,
+      "grad_norm": 0.4836462736129761,
+      "learning_rate": 4.215110774474532e-06,
+      "loss": 0.0978,
+      "step": 41680
+    },
+    {
+      "epoch": 3.947169096761977,
+      "grad_norm": 0.521132230758667,
+      "learning_rate": 4.211323612952093e-06,
+      "loss": 0.0945,
+      "step": 41690
+    },
+    {
+      "epoch": 3.9481158871425865,
+      "grad_norm": 0.6219900846481323,
+      "learning_rate": 4.207536451429654e-06,
+      "loss": 0.1016,
+      "step": 41700
+    },
+    {
+      "epoch": 3.9490626775231963,
+      "grad_norm": 0.4954199194908142,
+      "learning_rate": 4.203749289907214e-06,
+      "loss": 0.0984,
+      "step": 41710
+    },
+    {
+      "epoch": 3.950009467903806,
+      "grad_norm": 0.5494492053985596,
+      "learning_rate": 4.199962128384776e-06,
+      "loss": 0.0959,
+      "step": 41720
+    },
+    {
+      "epoch": 3.9509562582844158,
+      "grad_norm": 0.4500981569290161,
+      "learning_rate": 4.196174966862337e-06,
+      "loss": 0.0954,
+      "step": 41730
+    },
+    {
+      "epoch": 3.9519030486650255,
+      "grad_norm": 0.4959108531475067,
+      "learning_rate": 4.192387805339898e-06,
+      "loss": 0.1015,
+      "step": 41740
+    },
+    {
+      "epoch": 3.9528498390456352,
+      "grad_norm": 0.514580249786377,
+      "learning_rate": 4.1886006438174595e-06,
+      "loss": 0.1017,
+      "step": 41750
+    },
+    {
+      "epoch": 3.953796629426245,
+      "grad_norm": 0.6242176294326782,
+      "learning_rate": 4.1848134822950204e-06,
+      "loss": 0.1032,
+      "step": 41760
+    },
+    {
+      "epoch": 3.9547434198068547,
+      "grad_norm": 0.5583734512329102,
+      "learning_rate": 4.181026320772581e-06,
+      "loss": 0.0922,
+      "step": 41770
+    },
+    {
+      "epoch": 3.9556902101874645,
+      "grad_norm": 0.5744973421096802,
+      "learning_rate": 4.177239159250142e-06,
+      "loss": 0.1132,
+      "step": 41780
+    },
+    {
+      "epoch": 3.956637000568074,
+      "grad_norm": 0.48197564482688904,
+      "learning_rate": 4.173451997727703e-06,
+      "loss": 0.1009,
+      "step": 41790
+    },
+    {
+      "epoch": 3.957583790948684,
+      "grad_norm": 0.44193461537361145,
+      "learning_rate": 4.169664836205264e-06,
+      "loss": 0.089,
+      "step": 41800
+    },
+    {
+      "epoch": 3.9585305813292937,
+      "grad_norm": 0.5327918529510498,
+      "learning_rate": 4.165877674682826e-06,
+      "loss": 0.0966,
+      "step": 41810
+    },
+    {
+      "epoch": 3.9594773717099034,
+      "grad_norm": 0.5483490824699402,
+      "learning_rate": 4.162090513160387e-06,
+      "loss": 0.1069,
+      "step": 41820
+    },
+    {
+      "epoch": 3.960424162090513,
+      "grad_norm": 0.4684647023677826,
+      "learning_rate": 4.1583033516379476e-06,
+      "loss": 0.0917,
+      "step": 41830
+    },
+    {
+      "epoch": 3.961370952471123,
+      "grad_norm": 0.5732463598251343,
+      "learning_rate": 4.154516190115509e-06,
+      "loss": 0.0979,
+      "step": 41840
+    },
+    {
+      "epoch": 3.9623177428517327,
+      "grad_norm": 0.4299609959125519,
+      "learning_rate": 4.150729028593069e-06,
+      "loss": 0.0904,
+      "step": 41850
+    },
+    {
+      "epoch": 3.9632645332323424,
+      "grad_norm": 0.5590096116065979,
+      "learning_rate": 4.146941867070631e-06,
+      "loss": 0.0963,
+      "step": 41860
+    },
+    {
+      "epoch": 3.964211323612952,
+      "grad_norm": 0.47193509340286255,
+      "learning_rate": 4.143154705548192e-06,
+      "loss": 0.1036,
+      "step": 41870
+    },
+    {
+      "epoch": 3.965158113993562,
+      "grad_norm": 0.4498583674430847,
+      "learning_rate": 4.139367544025753e-06,
+      "loss": 0.0872,
+      "step": 41880
+    },
+    {
+      "epoch": 3.9661049043741716,
+      "grad_norm": 0.5629134178161621,
+      "learning_rate": 4.135580382503314e-06,
+      "loss": 0.0999,
+      "step": 41890
+    },
+    {
+      "epoch": 3.9670516947547814,
+      "grad_norm": 0.5969022512435913,
+      "learning_rate": 4.1317932209808755e-06,
+      "loss": 0.1046,
+      "step": 41900
+    },
+    {
+      "epoch": 3.967998485135391,
+      "grad_norm": 0.7509692311286926,
+      "learning_rate": 4.1280060594584364e-06,
+      "loss": 0.1002,
+      "step": 41910
+    },
+    {
+      "epoch": 3.968945275516001,
+      "grad_norm": 0.5245600938796997,
+      "learning_rate": 4.124218897935997e-06,
+      "loss": 0.1006,
+      "step": 41920
+    },
+    {
+      "epoch": 3.9698920658966106,
+      "grad_norm": 0.5322967767715454,
+      "learning_rate": 4.120431736413558e-06,
+      "loss": 0.0946,
+      "step": 41930
+    },
+    {
+      "epoch": 3.9708388562772203,
+      "grad_norm": 0.3742385804653168,
+      "learning_rate": 4.116644574891119e-06,
+      "loss": 0.0973,
+      "step": 41940
+    },
+    {
+      "epoch": 3.97178564665783,
+      "grad_norm": 0.7984851002693176,
+      "learning_rate": 4.112857413368681e-06,
+      "loss": 0.1026,
+      "step": 41950
+    },
+    {
+      "epoch": 3.97273243703844,
+      "grad_norm": 0.5663356184959412,
+      "learning_rate": 4.109070251846242e-06,
+      "loss": 0.0899,
+      "step": 41960
+    },
+    {
+      "epoch": 3.9736792274190496,
+      "grad_norm": 0.596892237663269,
+      "learning_rate": 4.105283090323803e-06,
+      "loss": 0.0942,
+      "step": 41970
+    },
+    {
+      "epoch": 3.9746260177996593,
+      "grad_norm": 0.4649079144001007,
+      "learning_rate": 4.1014959288013636e-06,
+      "loss": 0.1016,
+      "step": 41980
+    },
+    {
+      "epoch": 3.975572808180269,
+      "grad_norm": 0.4916474521160126,
+      "learning_rate": 4.0977087672789245e-06,
+      "loss": 0.0997,
+      "step": 41990
+    },
+    {
+      "epoch": 3.9765195985608788,
+      "grad_norm": 0.5276491641998291,
+      "learning_rate": 4.093921605756485e-06,
+      "loss": 0.0989,
+      "step": 42000
+    },
+    {
+      "epoch": 3.9774663889414885,
+      "grad_norm": 0.4202950596809387,
+      "learning_rate": 4.090134444234047e-06,
+      "loss": 0.0949,
+      "step": 42010
+    },
+    {
+      "epoch": 3.9784131793220983,
+      "grad_norm": 0.4367360770702362,
+      "learning_rate": 4.086347282711608e-06,
+      "loss": 0.0954,
+      "step": 42020
+    },
+    {
+      "epoch": 3.979359969702708,
+      "grad_norm": 0.4967721700668335,
+      "learning_rate": 4.082560121189169e-06,
+      "loss": 0.1072,
+      "step": 42030
+    },
+    {
+      "epoch": 3.9803067600833177,
+      "grad_norm": 0.4512579143047333,
+      "learning_rate": 4.078772959666731e-06,
+      "loss": 0.0905,
+      "step": 42040
+    },
+    {
+      "epoch": 3.9812535504639275,
+      "grad_norm": 0.5061561465263367,
+      "learning_rate": 4.0749857981442915e-06,
+      "loss": 0.0954,
+      "step": 42050
+    },
+    {
+      "epoch": 3.9822003408445372,
+      "grad_norm": 0.4698218107223511,
+      "learning_rate": 4.0711986366218516e-06,
+      "loss": 0.1007,
+      "step": 42060
+    },
+    {
+      "epoch": 3.983147131225147,
+      "grad_norm": 0.6213154196739197,
+      "learning_rate": 4.067411475099413e-06,
+      "loss": 0.0922,
+      "step": 42070
+    },
+    {
+      "epoch": 3.9840939216057567,
+      "grad_norm": 0.5239683985710144,
+      "learning_rate": 4.063624313576974e-06,
+      "loss": 0.0936,
+      "step": 42080
+    },
+    {
+      "epoch": 3.9850407119863664,
+      "grad_norm": 0.4962012469768524,
+      "learning_rate": 4.059837152054535e-06,
+      "loss": 0.1056,
+      "step": 42090
+    },
+    {
+      "epoch": 3.985987502366976,
+      "grad_norm": 0.3595285415649414,
+      "learning_rate": 4.056049990532097e-06,
+      "loss": 0.0968,
+      "step": 42100
+    },
+    {
+      "epoch": 3.986934292747586,
+      "grad_norm": 0.40510043501853943,
+      "learning_rate": 4.052262829009658e-06,
+      "loss": 0.0974,
+      "step": 42110
+    },
+    {
+      "epoch": 3.9878810831281957,
+      "grad_norm": 0.6077454090118408,
+      "learning_rate": 4.048475667487219e-06,
+      "loss": 0.0937,
+      "step": 42120
+    },
+    {
+      "epoch": 3.9888278735088054,
+      "grad_norm": 0.5463760495185852,
+      "learning_rate": 4.0446885059647795e-06,
+      "loss": 0.0941,
+      "step": 42130
+    },
+    {
+      "epoch": 3.989774663889415,
+      "grad_norm": 0.548797070980072,
+      "learning_rate": 4.0409013444423404e-06,
+      "loss": 0.0925,
+      "step": 42140
+    },
+    {
+      "epoch": 3.990721454270025,
+      "grad_norm": 0.5181418657302856,
+      "learning_rate": 4.037114182919901e-06,
+      "loss": 0.1075,
+      "step": 42150
+    },
+    {
+      "epoch": 3.991668244650634,
+      "grad_norm": 0.49558815360069275,
+      "learning_rate": 4.033327021397463e-06,
+      "loss": 0.0938,
+      "step": 42160
+    },
+    {
+      "epoch": 3.992615035031244,
+      "grad_norm": 0.4663916230201721,
+      "learning_rate": 4.029539859875024e-06,
+      "loss": 0.0906,
+      "step": 42170
+    },
+    {
+      "epoch": 3.9935618254118537,
+      "grad_norm": 0.5485295057296753,
+      "learning_rate": 4.025752698352585e-06,
+      "loss": 0.0894,
+      "step": 42180
+    },
+    {
+      "epoch": 3.9945086157924634,
+      "grad_norm": 0.6362839341163635,
+      "learning_rate": 4.021965536830147e-06,
+      "loss": 0.0995,
+      "step": 42190
+    },
+    {
+      "epoch": 3.995455406173073,
+      "grad_norm": 0.499687522649765,
+      "learning_rate": 4.018178375307707e-06,
+      "loss": 0.0889,
+      "step": 42200
+    },
+    {
+      "epoch": 3.996402196553683,
+      "grad_norm": 0.4681762158870697,
+      "learning_rate": 4.014391213785268e-06,
+      "loss": 0.0959,
+      "step": 42210
+    },
+    {
+      "epoch": 3.9973489869342926,
+      "grad_norm": 0.400115430355072,
+      "learning_rate": 4.010604052262829e-06,
+      "loss": 0.0889,
+      "step": 42220
+    },
+    {
+      "epoch": 3.9982957773149024,
+      "grad_norm": 0.5818225145339966,
+      "learning_rate": 4.00681689074039e-06,
+      "loss": 0.0921,
+      "step": 42230
+    },
+    {
+      "epoch": 3.999242567695512,
+      "grad_norm": 0.4878033995628357,
+      "learning_rate": 4.003029729217951e-06,
+      "loss": 0.1001,
+      "step": 42240
+    },
+    {
+      "epoch": 4.0,
+      "eval_f1_micro": 0.3810792804796802,
+      "eval_loss": 0.11376741528511047,
+      "eval_precision": 0.5667181865018032,
+      "eval_recall": 0.28705060521508724,
+      "eval_runtime": 344.9502,
+      "eval_samples_per_second": 122.467,
+      "eval_steps_per_second": 7.656,
+      "step": 42248
+    },
+    {
+      "epoch": 4.000189358076122,
+      "grad_norm": 0.49115651845932007,
+      "learning_rate": 3.999242567695513e-06,
+      "loss": 0.0971,
+      "step": 42250
+    },
+    {
+      "epoch": 4.001136148456732,
+      "grad_norm": 0.39359888434410095,
+      "learning_rate": 3.995455406173074e-06,
+      "loss": 0.0859,
+      "step": 42260
+    },
+    {
+      "epoch": 4.002082938837342,
+      "grad_norm": 0.5498154759407043,
+      "learning_rate": 3.991668244650635e-06,
+      "loss": 0.0983,
+      "step": 42270
+    },
+    {
+      "epoch": 4.0030297292179515,
+      "grad_norm": 0.3986906409263611,
+      "learning_rate": 3.9878810831281955e-06,
+      "loss": 0.087,
+      "step": 42280
+    },
+    {
+      "epoch": 4.003976519598561,
+      "grad_norm": 0.49595123529434204,
+      "learning_rate": 3.9840939216057564e-06,
+      "loss": 0.0904,
+      "step": 42290
+    },
+    {
+      "epoch": 4.004923309979171,
+      "grad_norm": 0.5356326103210449,
+      "learning_rate": 3.980306760083318e-06,
+      "loss": 0.0977,
+      "step": 42300
+    },
+    {
+      "epoch": 4.005870100359781,
+      "grad_norm": 0.4728410840034485,
+      "learning_rate": 3.976519598560879e-06,
+      "loss": 0.0972,
+      "step": 42310
+    },
+    {
+      "epoch": 4.0068168907403905,
+      "grad_norm": 0.5084603428840637,
+      "learning_rate": 3.97273243703844e-06,
+      "loss": 0.0911,
+      "step": 42320
+    },
+    {
+      "epoch": 4.007763681121,
+      "grad_norm": 0.4655855596065521,
+      "learning_rate": 3.968945275516001e-06,
+      "loss": 0.0995,
+      "step": 42330
+    },
+    {
+      "epoch": 4.00871047150161,
+      "grad_norm": 0.4194011092185974,
+      "learning_rate": 3.965158113993562e-06,
+      "loss": 0.0991,
+      "step": 42340
+    },
+    {
+      "epoch": 4.00965726188222,
+      "grad_norm": 0.42159825563430786,
+      "learning_rate": 3.961370952471123e-06,
+      "loss": 0.0902,
+      "step": 42350
+    },
+    {
+      "epoch": 4.010604052262829,
+      "grad_norm": 0.4718884825706482,
+      "learning_rate": 3.957583790948684e-06,
+      "loss": 0.0951,
+      "step": 42360
+    },
+    {
+      "epoch": 4.011550842643438,
+      "grad_norm": 0.4827013909816742,
+      "learning_rate": 3.953796629426245e-06,
+      "loss": 0.0913,
+      "step": 42370
+    },
+    {
+      "epoch": 4.012497633024048,
+      "grad_norm": 0.517824113368988,
+      "learning_rate": 3.950009467903806e-06,
+      "loss": 0.1011,
+      "step": 42380
+    },
+    {
+      "epoch": 4.013444423404658,
+      "grad_norm": 0.6215113401412964,
+      "learning_rate": 3.946222306381368e-06,
+      "loss": 0.09,
+      "step": 42390
+    },
+    {
+      "epoch": 4.0143912137852675,
+      "grad_norm": 0.39003387093544006,
+      "learning_rate": 3.942435144858929e-06,
+      "loss": 0.097,
+      "step": 42400
+    },
+    {
+      "epoch": 4.015338004165877,
+      "grad_norm": 0.6039709448814392,
+      "learning_rate": 3.93864798333649e-06,
+      "loss": 0.0943,
+      "step": 42410
+    },
+    {
+      "epoch": 4.016284794546487,
+      "grad_norm": 0.3276531398296356,
+      "learning_rate": 3.934860821814051e-06,
+      "loss": 0.0876,
+      "step": 42420
+    },
+    {
+      "epoch": 4.017231584927097,
+      "grad_norm": 0.5303285717964172,
+      "learning_rate": 3.9310736602916115e-06,
+      "loss": 0.0961,
+      "step": 42430
+    },
+    {
+      "epoch": 4.0181783753077065,
+      "grad_norm": 0.4780145585536957,
+      "learning_rate": 3.927286498769172e-06,
+      "loss": 0.1001,
+      "step": 42440
+    },
+    {
+      "epoch": 4.019125165688316,
+      "grad_norm": 0.4328151345252991,
+      "learning_rate": 3.923499337246734e-06,
+      "loss": 0.0844,
+      "step": 42450
+    },
+    {
+      "epoch": 4.020071956068926,
+      "grad_norm": 0.4836941957473755,
+      "learning_rate": 3.919712175724295e-06,
+      "loss": 0.0974,
+      "step": 42460
+    },
+    {
+      "epoch": 4.021018746449536,
+      "grad_norm": 0.42942121624946594,
+      "learning_rate": 3.915925014201856e-06,
+      "loss": 0.0867,
+      "step": 42470
+    },
+    {
+      "epoch": 4.0219655368301455,
+      "grad_norm": 0.5348581671714783,
+      "learning_rate": 3.912137852679418e-06,
+      "loss": 0.0902,
+      "step": 42480
+    },
+    {
+      "epoch": 4.022912327210755,
+      "grad_norm": 0.4421156644821167,
+      "learning_rate": 3.908350691156978e-06,
+      "loss": 0.0909,
+      "step": 42490
+    },
+    {
+      "epoch": 4.023859117591365,
+      "grad_norm": 0.4744870662689209,
+      "learning_rate": 3.9045635296345395e-06,
+      "loss": 0.0938,
+      "step": 42500
+    },
+    {
+      "epoch": 4.024805907971975,
+      "grad_norm": 0.4022984206676483,
+      "learning_rate": 3.9007763681121e-06,
+      "loss": 0.0963,
+      "step": 42510
+    },
+    {
+      "epoch": 4.025752698352584,
+      "grad_norm": 0.5710201859474182,
+      "learning_rate": 3.896989206589661e-06,
+      "loss": 0.085,
+      "step": 42520
+    },
+    {
+      "epoch": 4.026699488733194,
+      "grad_norm": 0.46365416049957275,
+      "learning_rate": 3.893202045067222e-06,
+      "loss": 0.0856,
+      "step": 42530
+    },
+    {
+      "epoch": 4.027646279113804,
+      "grad_norm": 0.5811256170272827,
+      "learning_rate": 3.889414883544784e-06,
+      "loss": 0.0884,
+      "step": 42540
+    },
+    {
+      "epoch": 4.028593069494414,
+      "grad_norm": 0.6684948205947876,
+      "learning_rate": 3.885627722022345e-06,
+      "loss": 0.0943,
+      "step": 42550
+    },
+    {
+      "epoch": 4.029539859875023,
+      "grad_norm": 0.5227164626121521,
+      "learning_rate": 3.881840560499906e-06,
+      "loss": 0.0913,
+      "step": 42560
+    },
+    {
+      "epoch": 4.030486650255633,
+      "grad_norm": 0.47879600524902344,
+      "learning_rate": 3.878053398977467e-06,
+      "loss": 0.0914,
+      "step": 42570
+    },
+    {
+      "epoch": 4.031433440636243,
+      "grad_norm": 0.48690658807754517,
+      "learning_rate": 3.8742662374550275e-06,
+      "loss": 0.0975,
+      "step": 42580
+    },
+    {
+      "epoch": 4.032380231016853,
+      "grad_norm": 0.44016072154045105,
+      "learning_rate": 3.870479075932589e-06,
+      "loss": 0.0899,
+      "step": 42590
+    },
+    {
+      "epoch": 4.033327021397462,
+      "grad_norm": 0.6146271824836731,
+      "learning_rate": 3.86669191441015e-06,
+      "loss": 0.0915,
+      "step": 42600
+    },
+    {
+      "epoch": 4.034273811778072,
+      "grad_norm": 0.46333184838294983,
+      "learning_rate": 3.862904752887711e-06,
+      "loss": 0.0877,
+      "step": 42610
+    },
+    {
+      "epoch": 4.035220602158682,
+      "grad_norm": 0.6286950707435608,
+      "learning_rate": 3.859117591365272e-06,
+      "loss": 0.0972,
+      "step": 42620
+    },
+    {
+      "epoch": 4.036167392539292,
+      "grad_norm": 0.4000542163848877,
+      "learning_rate": 3.855330429842833e-06,
+      "loss": 0.0875,
+      "step": 42630
+    },
+    {
+      "epoch": 4.037114182919901,
+      "grad_norm": 0.44887304306030273,
+      "learning_rate": 3.851543268320394e-06,
+      "loss": 0.0939,
+      "step": 42640
+    },
+    {
+      "epoch": 4.038060973300511,
+      "grad_norm": 0.4717208743095398,
+      "learning_rate": 3.8477561067979555e-06,
+      "loss": 0.0938,
+      "step": 42650
+    },
+    {
+      "epoch": 4.039007763681121,
+      "grad_norm": 0.5275483131408691,
+      "learning_rate": 3.843968945275516e-06,
+      "loss": 0.0834,
+      "step": 42660
+    },
+    {
+      "epoch": 4.0399545540617305,
+      "grad_norm": 0.5291388630867004,
+      "learning_rate": 3.840181783753077e-06,
+      "loss": 0.0809,
+      "step": 42670
+    },
+    {
+      "epoch": 4.04090134444234,
+      "grad_norm": 0.49929943680763245,
+      "learning_rate": 3.836394622230639e-06,
+      "loss": 0.0963,
+      "step": 42680
+    },
+    {
+      "epoch": 4.04184813482295,
+      "grad_norm": 0.65031498670578,
+      "learning_rate": 3.8326074607082e-06,
+      "loss": 0.0855,
+      "step": 42690
+    },
+    {
+      "epoch": 4.04279492520356,
+      "grad_norm": 0.4142380654811859,
+      "learning_rate": 3.82882029918576e-06,
+      "loss": 0.0999,
+      "step": 42700
+    },
+    {
+      "epoch": 4.0437417155841695,
+      "grad_norm": 0.4350908100605011,
+      "learning_rate": 3.825033137663322e-06,
+      "loss": 0.0862,
+      "step": 42710
+    },
+    {
+      "epoch": 4.044688505964779,
+      "grad_norm": 0.5821089744567871,
+      "learning_rate": 3.821245976140883e-06,
+      "loss": 0.0968,
+      "step": 42720
+    },
+    {
+      "epoch": 4.045635296345389,
+      "grad_norm": 0.5138930678367615,
+      "learning_rate": 3.8174588146184435e-06,
+      "loss": 0.0857,
+      "step": 42730
+    },
+    {
+      "epoch": 4.046582086725999,
+      "grad_norm": 0.4821796417236328,
+      "learning_rate": 3.813671653096005e-06,
+      "loss": 0.0989,
+      "step": 42740
+    },
+    {
+      "epoch": 4.0475288771066085,
+      "grad_norm": 0.41542673110961914,
+      "learning_rate": 3.809884491573566e-06,
+      "loss": 0.0864,
+      "step": 42750
+    },
+    {
+      "epoch": 4.048475667487218,
+      "grad_norm": 0.49766823649406433,
+      "learning_rate": 3.806097330051127e-06,
+      "loss": 0.0998,
+      "step": 42760
+    },
+    {
+      "epoch": 4.049422457867828,
+      "grad_norm": 0.4959116280078888,
+      "learning_rate": 3.802310168528688e-06,
+      "loss": 0.0874,
+      "step": 42770
+    },
+    {
+      "epoch": 4.050369248248438,
+      "grad_norm": 0.40544602274894714,
+      "learning_rate": 3.798523007006249e-06,
+      "loss": 0.0962,
+      "step": 42780
+    },
+    {
+      "epoch": 4.051316038629047,
+      "grad_norm": 0.33507898449897766,
+      "learning_rate": 3.79473584548381e-06,
+      "loss": 0.0804,
+      "step": 42790
+    },
+    {
+      "epoch": 4.052262829009657,
+      "grad_norm": 0.55490642786026,
+      "learning_rate": 3.7909486839613715e-06,
+      "loss": 0.0981,
+      "step": 42800
+    },
+    {
+      "epoch": 4.053209619390267,
+      "grad_norm": 0.5510737299919128,
+      "learning_rate": 3.7871615224389324e-06,
+      "loss": 0.088,
+      "step": 42810
+    },
+    {
+      "epoch": 4.054156409770877,
+      "grad_norm": 0.43902432918548584,
+      "learning_rate": 3.7833743609164937e-06,
+      "loss": 0.0914,
+      "step": 42820
+    },
+    {
+      "epoch": 4.055103200151486,
+      "grad_norm": 0.6550730466842651,
+      "learning_rate": 3.7795871993940546e-06,
+      "loss": 0.0906,
+      "step": 42830
+    },
+    {
+      "epoch": 4.056049990532096,
+      "grad_norm": 0.5181649923324585,
+      "learning_rate": 3.7758000378716155e-06,
+      "loss": 0.0989,
+      "step": 42840
+    },
+    {
+      "epoch": 4.056996780912706,
+      "grad_norm": 0.479004830121994,
+      "learning_rate": 3.7720128763491764e-06,
+      "loss": 0.0845,
+      "step": 42850
+    },
+    {
+      "epoch": 4.057943571293316,
+      "grad_norm": 0.565980076789856,
+      "learning_rate": 3.7682257148267377e-06,
+      "loss": 0.0969,
+      "step": 42860
+    },
+    {
+      "epoch": 4.058890361673925,
+      "grad_norm": 0.5291234254837036,
+      "learning_rate": 3.7644385533042986e-06,
+      "loss": 0.0919,
+      "step": 42870
+    },
+    {
+      "epoch": 4.059837152054535,
+      "grad_norm": 0.4307625889778137,
+      "learning_rate": 3.76065139178186e-06,
+      "loss": 0.0916,
+      "step": 42880
+    },
+    {
+      "epoch": 4.060783942435145,
+      "grad_norm": 0.5084070563316345,
+      "learning_rate": 3.7568642302594212e-06,
+      "loss": 0.0903,
+      "step": 42890
+    },
+    {
+      "epoch": 4.061730732815755,
+      "grad_norm": 0.4158044159412384,
+      "learning_rate": 3.753077068736982e-06,
+      "loss": 0.0969,
+      "step": 42900
+    },
+    {
+      "epoch": 4.062677523196364,
+      "grad_norm": 0.4868123233318329,
+      "learning_rate": 3.7492899072145426e-06,
+      "loss": 0.0872,
+      "step": 42910
+    },
+    {
+      "epoch": 4.063624313576974,
+      "grad_norm": 0.5304853320121765,
+      "learning_rate": 3.745502745692104e-06,
+      "loss": 0.0999,
+      "step": 42920
+    },
+    {
+      "epoch": 4.064571103957584,
+      "grad_norm": 0.5248091816902161,
+      "learning_rate": 3.7417155841696652e-06,
+      "loss": 0.0868,
+      "step": 42930
+    },
+    {
+      "epoch": 4.0655178943381935,
+      "grad_norm": 0.4787554144859314,
+      "learning_rate": 3.737928422647226e-06,
+      "loss": 0.0874,
+      "step": 42940
+    },
+    {
+      "epoch": 4.066464684718803,
+      "grad_norm": 0.5181282758712769,
+      "learning_rate": 3.7341412611247875e-06,
+      "loss": 0.088,
+      "step": 42950
+    },
+    {
+      "epoch": 4.067411475099413,
+      "grad_norm": 0.49244225025177,
+      "learning_rate": 3.7303540996023484e-06,
+      "loss": 0.0858,
+      "step": 42960
+    },
+    {
+      "epoch": 4.068358265480023,
+      "grad_norm": 0.6987329721450806,
+      "learning_rate": 3.7265669380799097e-06,
+      "loss": 0.0864,
+      "step": 42970
+    },
+    {
+      "epoch": 4.0693050558606325,
+      "grad_norm": 0.5377588868141174,
+      "learning_rate": 3.72277977655747e-06,
+      "loss": 0.0928,
+      "step": 42980
+    },
+    {
+      "epoch": 4.070251846241242,
+      "grad_norm": 0.4637696146965027,
+      "learning_rate": 3.7189926150350315e-06,
+      "loss": 0.0864,
+      "step": 42990
+    },
+    {
+      "epoch": 4.071198636621852,
+      "grad_norm": 0.43769657611846924,
+      "learning_rate": 3.7152054535125924e-06,
+      "loss": 0.085,
+      "step": 43000
+    },
+    {
+      "epoch": 4.072145427002462,
+      "grad_norm": 0.5931176543235779,
+      "learning_rate": 3.7114182919901537e-06,
+      "loss": 0.1043,
+      "step": 43010
+    },
+    {
+      "epoch": 4.0730922173830715,
+      "grad_norm": 0.45635858178138733,
+      "learning_rate": 3.707631130467715e-06,
+      "loss": 0.0967,
+      "step": 43020
+    },
+    {
+      "epoch": 4.074039007763681,
+      "grad_norm": 0.49210211634635925,
+      "learning_rate": 3.703843968945276e-06,
+      "loss": 0.0806,
+      "step": 43030
+    },
+    {
+      "epoch": 4.074985798144291,
+      "grad_norm": 0.6789916157722473,
+      "learning_rate": 3.7000568074228372e-06,
+      "loss": 0.0991,
+      "step": 43040
+    },
+    {
+      "epoch": 4.075932588524901,
+      "grad_norm": 0.44278761744499207,
+      "learning_rate": 3.6962696459003977e-06,
+      "loss": 0.0774,
+      "step": 43050
+    },
+    {
+      "epoch": 4.07687937890551,
+      "grad_norm": 0.5981815457344055,
+      "learning_rate": 3.692482484377959e-06,
+      "loss": 0.0935,
+      "step": 43060
+    },
+    {
+      "epoch": 4.07782616928612,
+      "grad_norm": 0.6542117595672607,
+      "learning_rate": 3.68869532285552e-06,
+      "loss": 0.098,
+      "step": 43070
+    },
+    {
+      "epoch": 4.07877295966673,
+      "grad_norm": 0.5480513572692871,
+      "learning_rate": 3.6849081613330812e-06,
+      "loss": 0.1004,
+      "step": 43080
+    },
+    {
+      "epoch": 4.07971975004734,
+      "grad_norm": 0.4663204848766327,
+      "learning_rate": 3.681120999810642e-06,
+      "loss": 0.0915,
+      "step": 43090
+    },
+    {
+      "epoch": 4.080666540427949,
+      "grad_norm": 0.5434608459472656,
+      "learning_rate": 3.6773338382882035e-06,
+      "loss": 0.0831,
+      "step": 43100
+    },
+    {
+      "epoch": 4.081613330808559,
+      "grad_norm": 0.44176313281059265,
+      "learning_rate": 3.6735466767657648e-06,
+      "loss": 0.0872,
+      "step": 43110
+    },
+    {
+      "epoch": 4.082560121189169,
+      "grad_norm": 0.5268293619155884,
+      "learning_rate": 3.6697595152433252e-06,
+      "loss": 0.0984,
+      "step": 43120
+    },
+    {
+      "epoch": 4.083506911569779,
+      "grad_norm": 0.6170424222946167,
+      "learning_rate": 3.665972353720886e-06,
+      "loss": 0.0946,
+      "step": 43130
+    },
+    {
+      "epoch": 4.084453701950388,
+      "grad_norm": 0.45425689220428467,
+      "learning_rate": 3.6621851921984475e-06,
+      "loss": 0.0731,
+      "step": 43140
+    },
+    {
+      "epoch": 4.085400492330998,
+      "grad_norm": 0.4029372036457062,
+      "learning_rate": 3.6583980306760088e-06,
+      "loss": 0.0899,
+      "step": 43150
+    },
+    {
+      "epoch": 4.086347282711608,
+      "grad_norm": 0.6184468865394592,
+      "learning_rate": 3.6546108691535697e-06,
+      "loss": 0.0877,
+      "step": 43160
+    },
+    {
+      "epoch": 4.087294073092218,
+      "grad_norm": 0.49466800689697266,
+      "learning_rate": 3.650823707631131e-06,
+      "loss": 0.0795,
+      "step": 43170
+    },
+    {
+      "epoch": 4.088240863472827,
+      "grad_norm": 0.625234842300415,
+      "learning_rate": 3.647036546108692e-06,
+      "loss": 0.0829,
+      "step": 43180
+    },
+    {
+      "epoch": 4.089187653853437,
+      "grad_norm": 0.5789068937301636,
+      "learning_rate": 3.643249384586253e-06,
+      "loss": 0.0903,
+      "step": 43190
+    },
+    {
+      "epoch": 4.090134444234047,
+      "grad_norm": 0.5432450175285339,
+      "learning_rate": 3.6394622230638137e-06,
+      "loss": 0.0932,
+      "step": 43200
+    },
+    {
+      "epoch": 4.091081234614657,
+      "grad_norm": 0.4496850371360779,
+      "learning_rate": 3.635675061541375e-06,
+      "loss": 0.1028,
+      "step": 43210
+    },
+    {
+      "epoch": 4.092028024995266,
+      "grad_norm": 0.5556248426437378,
+      "learning_rate": 3.631887900018936e-06,
+      "loss": 0.0884,
+      "step": 43220
+    },
+    {
+      "epoch": 4.092974815375876,
+      "grad_norm": 0.4554558992385864,
+      "learning_rate": 3.6281007384964972e-06,
+      "loss": 0.0915,
+      "step": 43230
+    },
+    {
+      "epoch": 4.093921605756486,
+      "grad_norm": 0.5424495339393616,
+      "learning_rate": 3.6243135769740585e-06,
+      "loss": 0.0902,
+      "step": 43240
+    },
+    {
+      "epoch": 4.0948683961370955,
+      "grad_norm": 0.5642138719558716,
+      "learning_rate": 3.6205264154516194e-06,
+      "loss": 0.0981,
+      "step": 43250
+    },
+    {
+      "epoch": 4.095815186517705,
+      "grad_norm": 0.45746803283691406,
+      "learning_rate": 3.6167392539291803e-06,
+      "loss": 0.085,
+      "step": 43260
+    },
+    {
+      "epoch": 4.096761976898315,
+      "grad_norm": 0.34574341773986816,
+      "learning_rate": 3.6129520924067412e-06,
+      "loss": 0.0862,
+      "step": 43270
+    },
+    {
+      "epoch": 4.097708767278925,
+      "grad_norm": 0.5801050662994385,
+      "learning_rate": 3.6091649308843026e-06,
+      "loss": 0.0977,
+      "step": 43280
+    },
+    {
+      "epoch": 4.0986555576595345,
+      "grad_norm": 0.4668993353843689,
+      "learning_rate": 3.6053777693618634e-06,
+      "loss": 0.0949,
+      "step": 43290
+    },
+    {
+      "epoch": 4.099602348040144,
+      "grad_norm": 0.4316027760505676,
+      "learning_rate": 3.6015906078394248e-06,
+      "loss": 0.0861,
+      "step": 43300
+    },
+    {
+      "epoch": 4.100549138420754,
+      "grad_norm": 0.4293339252471924,
+      "learning_rate": 3.5978034463169857e-06,
+      "loss": 0.0851,
+      "step": 43310
+    },
+    {
+      "epoch": 4.101495928801364,
+      "grad_norm": 0.5610676407814026,
+      "learning_rate": 3.594016284794547e-06,
+      "loss": 0.0951,
+      "step": 43320
+    },
+    {
+      "epoch": 4.1024427191819735,
+      "grad_norm": 0.5009633302688599,
+      "learning_rate": 3.5902291232721075e-06,
+      "loss": 0.0776,
+      "step": 43330
+    },
+    {
+      "epoch": 4.103389509562583,
+      "grad_norm": 0.4935799241065979,
+      "learning_rate": 3.5864419617496688e-06,
+      "loss": 0.0954,
+      "step": 43340
+    },
+    {
+      "epoch": 4.104336299943193,
+      "grad_norm": 0.5310682058334351,
+      "learning_rate": 3.58265480022723e-06,
+      "loss": 0.0908,
+      "step": 43350
+    },
+    {
+      "epoch": 4.105283090323803,
+      "grad_norm": 0.5049786567687988,
+      "learning_rate": 3.578867638704791e-06,
+      "loss": 0.0912,
+      "step": 43360
+    },
+    {
+      "epoch": 4.106229880704412,
+      "grad_norm": 0.5558485984802246,
+      "learning_rate": 3.5750804771823523e-06,
+      "loss": 0.0943,
+      "step": 43370
+    },
+    {
+      "epoch": 4.107176671085022,
+      "grad_norm": 0.4917635917663574,
+      "learning_rate": 3.571293315659913e-06,
+      "loss": 0.0899,
+      "step": 43380
+    },
+    {
+      "epoch": 4.108123461465632,
+      "grad_norm": 0.4739043712615967,
+      "learning_rate": 3.5675061541374745e-06,
+      "loss": 0.0873,
+      "step": 43390
+    },
+    {
+      "epoch": 4.109070251846242,
+      "grad_norm": 0.5204480886459351,
+      "learning_rate": 3.563718992615035e-06,
+      "loss": 0.1029,
+      "step": 43400
+    },
+    {
+      "epoch": 4.110017042226851,
+      "grad_norm": 0.4711787700653076,
+      "learning_rate": 3.5599318310925963e-06,
+      "loss": 0.0947,
+      "step": 43410
+    },
+    {
+      "epoch": 4.110963832607461,
+      "grad_norm": 0.5298397541046143,
+      "learning_rate": 3.5561446695701572e-06,
+      "loss": 0.0915,
+      "step": 43420
+    },
+    {
+      "epoch": 4.111910622988071,
+      "grad_norm": 0.6017003059387207,
+      "learning_rate": 3.5523575080477185e-06,
+      "loss": 0.0902,
+      "step": 43430
+    },
+    {
+      "epoch": 4.112857413368681,
+      "grad_norm": 0.4537605345249176,
+      "learning_rate": 3.54857034652528e-06,
+      "loss": 0.0843,
+      "step": 43440
+    },
+    {
+      "epoch": 4.1138042037492895,
+      "grad_norm": 0.5435331463813782,
+      "learning_rate": 3.5447831850028408e-06,
+      "loss": 0.0977,
+      "step": 43450
+    },
+    {
+      "epoch": 4.114750994129899,
+      "grad_norm": 0.5554385185241699,
+      "learning_rate": 3.540996023480402e-06,
+      "loss": 0.0895,
+      "step": 43460
+    },
+    {
+      "epoch": 4.115697784510509,
+      "grad_norm": 0.5493789315223694,
+      "learning_rate": 3.5372088619579626e-06,
+      "loss": 0.0863,
+      "step": 43470
+    },
+    {
+      "epoch": 4.116644574891119,
+      "grad_norm": 0.5053249597549438,
+      "learning_rate": 3.533421700435524e-06,
+      "loss": 0.0876,
+      "step": 43480
+    },
+    {
+      "epoch": 4.117591365271728,
+      "grad_norm": 0.5336869955062866,
+      "learning_rate": 3.5296345389130848e-06,
+      "loss": 0.0879,
+      "step": 43490
+    },
+    {
+      "epoch": 4.118538155652338,
+      "grad_norm": 0.5733828544616699,
+      "learning_rate": 3.525847377390646e-06,
+      "loss": 0.0983,
+      "step": 43500
+    },
+    {
+      "epoch": 4.119484946032948,
+      "grad_norm": 0.6210169792175293,
+      "learning_rate": 3.522060215868207e-06,
+      "loss": 0.0854,
+      "step": 43510
+    },
+    {
+      "epoch": 4.120431736413558,
+      "grad_norm": 0.4751220941543579,
+      "learning_rate": 3.5182730543457683e-06,
+      "loss": 0.0949,
+      "step": 43520
+    },
+    {
+      "epoch": 4.121378526794167,
+      "grad_norm": 0.5846556425094604,
+      "learning_rate": 3.5144858928233296e-06,
+      "loss": 0.0801,
+      "step": 43530
+    },
+    {
+      "epoch": 4.122325317174777,
+      "grad_norm": 0.5414957404136658,
+      "learning_rate": 3.51069873130089e-06,
+      "loss": 0.0845,
+      "step": 43540
+    },
+    {
+      "epoch": 4.123272107555387,
+      "grad_norm": 0.47205281257629395,
+      "learning_rate": 3.506911569778451e-06,
+      "loss": 0.0982,
+      "step": 43550
+    },
+    {
+      "epoch": 4.124218897935997,
+      "grad_norm": 0.5458006858825684,
+      "learning_rate": 3.5031244082560123e-06,
+      "loss": 0.0885,
+      "step": 43560
+    },
+    {
+      "epoch": 4.125165688316606,
+      "grad_norm": 0.5724362730979919,
+      "learning_rate": 3.4993372467335736e-06,
+      "loss": 0.086,
+      "step": 43570
+    },
+    {
+      "epoch": 4.126112478697216,
+      "grad_norm": 0.3912672996520996,
+      "learning_rate": 3.4955500852111345e-06,
+      "loss": 0.0897,
+      "step": 43580
+    },
+    {
+      "epoch": 4.127059269077826,
+      "grad_norm": 0.4875558316707611,
+      "learning_rate": 3.491762923688696e-06,
+      "loss": 0.0947,
+      "step": 43590
+    },
+    {
+      "epoch": 4.128006059458436,
+      "grad_norm": 0.46745261549949646,
+      "learning_rate": 3.4879757621662567e-06,
+      "loss": 0.0883,
+      "step": 43600
+    },
+    {
+      "epoch": 4.128952849839045,
+      "grad_norm": 0.6316899657249451,
+      "learning_rate": 3.4841886006438176e-06,
+      "loss": 0.1003,
+      "step": 43610
+    },
+    {
+      "epoch": 4.129899640219655,
+      "grad_norm": 0.44248083233833313,
+      "learning_rate": 3.4804014391213785e-06,
+      "loss": 0.0972,
+      "step": 43620
+    },
+    {
+      "epoch": 4.130846430600265,
+      "grad_norm": 0.6286366581916809,
+      "learning_rate": 3.47661427759894e-06,
+      "loss": 0.0857,
+      "step": 43630
+    },
+    {
+      "epoch": 4.1317932209808745,
+      "grad_norm": 0.46254751086235046,
+      "learning_rate": 3.4728271160765008e-06,
+      "loss": 0.0873,
+      "step": 43640
+    },
+    {
+      "epoch": 4.132740011361484,
+      "grad_norm": 0.591060221195221,
+      "learning_rate": 3.469039954554062e-06,
+      "loss": 0.0979,
+      "step": 43650
+    },
+    {
+      "epoch": 4.133686801742094,
+      "grad_norm": 0.5578216910362244,
+      "learning_rate": 3.4652527930316234e-06,
+      "loss": 0.1034,
+      "step": 43660
+    },
+    {
+      "epoch": 4.134633592122704,
+      "grad_norm": 0.7366402745246887,
+      "learning_rate": 3.4614656315091843e-06,
+      "loss": 0.1012,
+      "step": 43670
+    },
+    {
+      "epoch": 4.1355803825033135,
+      "grad_norm": 0.5355960130691528,
+      "learning_rate": 3.4576784699867448e-06,
+      "loss": 0.0971,
+      "step": 43680
+    },
+    {
+      "epoch": 4.136527172883923,
+      "grad_norm": 0.5104373693466187,
+      "learning_rate": 3.453891308464306e-06,
+      "loss": 0.0926,
+      "step": 43690
+    },
+    {
+      "epoch": 4.137473963264533,
+      "grad_norm": 0.48333603143692017,
+      "learning_rate": 3.4501041469418674e-06,
+      "loss": 0.099,
+      "step": 43700
+    },
+    {
+      "epoch": 4.138420753645143,
+      "grad_norm": 0.5138222575187683,
+      "learning_rate": 3.4463169854194283e-06,
+      "loss": 0.0983,
+      "step": 43710
+    },
+    {
+      "epoch": 4.1393675440257525,
+      "grad_norm": 0.49473053216934204,
+      "learning_rate": 3.4425298238969896e-06,
+      "loss": 0.0889,
+      "step": 43720
+    },
+    {
+      "epoch": 4.140314334406362,
+      "grad_norm": 0.6268186569213867,
+      "learning_rate": 3.4387426623745505e-06,
+      "loss": 0.0898,
+      "step": 43730
+    },
+    {
+      "epoch": 4.141261124786972,
+      "grad_norm": 0.4563818871974945,
+      "learning_rate": 3.434955500852112e-06,
+      "loss": 0.0878,
+      "step": 43740
+    },
+    {
+      "epoch": 4.142207915167582,
+      "grad_norm": 0.4766491651535034,
+      "learning_rate": 3.4311683393296723e-06,
+      "loss": 0.0866,
+      "step": 43750
+    },
+    {
+      "epoch": 4.143154705548191,
+      "grad_norm": 0.5177786946296692,
+      "learning_rate": 3.4273811778072336e-06,
+      "loss": 0.087,
+      "step": 43760
+    },
+    {
+      "epoch": 4.144101495928801,
+      "grad_norm": 0.5356361865997314,
+      "learning_rate": 3.4235940162847945e-06,
+      "loss": 0.0851,
+      "step": 43770
+    },
+    {
+      "epoch": 4.145048286309411,
+      "grad_norm": 0.5455963015556335,
+      "learning_rate": 3.419806854762356e-06,
+      "loss": 0.0932,
+      "step": 43780
+    },
+    {
+      "epoch": 4.145995076690021,
+      "grad_norm": 0.501149594783783,
+      "learning_rate": 3.416019693239917e-06,
+      "loss": 0.0876,
+      "step": 43790
+    },
+    {
+      "epoch": 4.14694186707063,
+      "grad_norm": 0.5715905427932739,
+      "learning_rate": 3.412232531717478e-06,
+      "loss": 0.0959,
+      "step": 43800
+    },
+    {
+      "epoch": 4.14788865745124,
+      "grad_norm": 0.5404168367385864,
+      "learning_rate": 3.4084453701950394e-06,
+      "loss": 0.0905,
+      "step": 43810
+    },
+    {
+      "epoch": 4.14883544783185,
+      "grad_norm": 0.4844471514225006,
+      "learning_rate": 3.4046582086726003e-06,
+      "loss": 0.0814,
+      "step": 43820
+    },
+    {
+      "epoch": 4.14978223821246,
+      "grad_norm": 0.4519600570201874,
+      "learning_rate": 3.400871047150161e-06,
+      "loss": 0.0831,
+      "step": 43830
+    },
+    {
+      "epoch": 4.150729028593069,
+      "grad_norm": 0.5062319040298462,
+      "learning_rate": 3.397083885627722e-06,
+      "loss": 0.0879,
+      "step": 43840
+    },
+    {
+      "epoch": 4.151675818973679,
+      "grad_norm": 0.4504513740539551,
+      "learning_rate": 3.3932967241052834e-06,
+      "loss": 0.0981,
+      "step": 43850
+    },
+    {
+      "epoch": 4.152622609354289,
+      "grad_norm": 0.6103249788284302,
+      "learning_rate": 3.3895095625828443e-06,
+      "loss": 0.0992,
+      "step": 43860
+    },
+    {
+      "epoch": 4.153569399734899,
+      "grad_norm": 0.5387426018714905,
+      "learning_rate": 3.3857224010604056e-06,
+      "loss": 0.0953,
+      "step": 43870
+    },
+    {
+      "epoch": 4.154516190115508,
+      "grad_norm": 0.5584139823913574,
+      "learning_rate": 3.381935239537967e-06,
+      "loss": 0.0872,
+      "step": 43880
+    },
+    {
+      "epoch": 4.155462980496118,
+      "grad_norm": 0.4367479383945465,
+      "learning_rate": 3.378148078015528e-06,
+      "loss": 0.0927,
+      "step": 43890
+    },
+    {
+      "epoch": 4.156409770876728,
+      "grad_norm": 0.5904217958450317,
+      "learning_rate": 3.3743609164930883e-06,
+      "loss": 0.0887,
+      "step": 43900
+    },
+    {
+      "epoch": 4.1573565612573375,
+      "grad_norm": 0.6446182131767273,
+      "learning_rate": 3.3705737549706496e-06,
+      "loss": 0.0908,
+      "step": 43910
+    },
+    {
+      "epoch": 4.158303351637947,
+      "grad_norm": 0.38347458839416504,
+      "learning_rate": 3.366786593448211e-06,
+      "loss": 0.0941,
+      "step": 43920
+    },
+    {
+      "epoch": 4.159250142018557,
+      "grad_norm": 0.5349393486976624,
+      "learning_rate": 3.362999431925772e-06,
+      "loss": 0.0929,
+      "step": 43930
+    },
+    {
+      "epoch": 4.160196932399167,
+      "grad_norm": 0.5757917165756226,
+      "learning_rate": 3.359212270403333e-06,
+      "loss": 0.0875,
+      "step": 43940
+    },
+    {
+      "epoch": 4.1611437227797765,
+      "grad_norm": 0.34581947326660156,
+      "learning_rate": 3.355425108880894e-06,
+      "loss": 0.0951,
+      "step": 43950
+    },
+    {
+      "epoch": 4.162090513160386,
+      "grad_norm": 0.4834064245223999,
+      "learning_rate": 3.3516379473584554e-06,
+      "loss": 0.0894,
+      "step": 43960
+    },
+    {
+      "epoch": 4.163037303540996,
+      "grad_norm": 0.5049462914466858,
+      "learning_rate": 3.347850785836016e-06,
+      "loss": 0.0941,
+      "step": 43970
+    },
+    {
+      "epoch": 4.163984093921606,
+      "grad_norm": 0.6113541722297668,
+      "learning_rate": 3.344063624313577e-06,
+      "loss": 0.093,
+      "step": 43980
+    },
+    {
+      "epoch": 4.1649308843022155,
+      "grad_norm": 0.6100361943244934,
+      "learning_rate": 3.340276462791138e-06,
+      "loss": 0.0983,
+      "step": 43990
+    },
+    {
+      "epoch": 4.165877674682825,
+      "grad_norm": 0.7216998934745789,
+      "learning_rate": 3.3364893012686994e-06,
+      "loss": 0.0979,
+      "step": 44000
+    },
+    {
+      "epoch": 4.166824465063435,
+      "grad_norm": 0.789558470249176,
+      "learning_rate": 3.3327021397462607e-06,
+      "loss": 0.1005,
+      "step": 44010
+    },
+    {
+      "epoch": 4.167771255444045,
+      "grad_norm": 0.4637594223022461,
+      "learning_rate": 3.3289149782238216e-06,
+      "loss": 0.0846,
+      "step": 44020
+    },
+    {
+      "epoch": 4.168718045824654,
+      "grad_norm": 0.49285727739334106,
+      "learning_rate": 3.325127816701383e-06,
+      "loss": 0.0794,
+      "step": 44030
+    },
+    {
+      "epoch": 4.169664836205264,
+      "grad_norm": 0.43726247549057007,
+      "learning_rate": 3.3213406551789434e-06,
+      "loss": 0.081,
+      "step": 44040
+    },
+    {
+      "epoch": 4.170611626585874,
+      "grad_norm": 0.5689728856086731,
+      "learning_rate": 3.3175534936565047e-06,
+      "loss": 0.093,
+      "step": 44050
+    },
+    {
+      "epoch": 4.171558416966484,
+      "grad_norm": 0.5433076620101929,
+      "learning_rate": 3.3137663321340656e-06,
+      "loss": 0.0962,
+      "step": 44060
+    },
+    {
+      "epoch": 4.172505207347093,
+      "grad_norm": 0.47096943855285645,
+      "learning_rate": 3.309979170611627e-06,
+      "loss": 0.0893,
+      "step": 44070
+    },
+    {
+      "epoch": 4.173451997727703,
+      "grad_norm": 0.50560462474823,
+      "learning_rate": 3.306192009089188e-06,
+      "loss": 0.0866,
+      "step": 44080
+    },
+    {
+      "epoch": 4.174398788108313,
+      "grad_norm": 0.6455307602882385,
+      "learning_rate": 3.302404847566749e-06,
+      "loss": 0.102,
+      "step": 44090
+    },
+    {
+      "epoch": 4.175345578488923,
+      "grad_norm": 0.5444678664207458,
+      "learning_rate": 3.2986176860443105e-06,
+      "loss": 0.0829,
+      "step": 44100
+    },
+    {
+      "epoch": 4.176292368869532,
+      "grad_norm": 0.5178965330123901,
+      "learning_rate": 3.294830524521871e-06,
+      "loss": 0.0884,
+      "step": 44110
+    },
+    {
+      "epoch": 4.177239159250142,
+      "grad_norm": 0.5928695797920227,
+      "learning_rate": 3.2910433629994323e-06,
+      "loss": 0.0952,
+      "step": 44120
+    },
+    {
+      "epoch": 4.178185949630752,
+      "grad_norm": 0.47902634739875793,
+      "learning_rate": 3.287256201476993e-06,
+      "loss": 0.1003,
+      "step": 44130
+    },
+    {
+      "epoch": 4.179132740011362,
+      "grad_norm": 0.4959571957588196,
+      "learning_rate": 3.2834690399545545e-06,
+      "loss": 0.091,
+      "step": 44140
+    },
+    {
+      "epoch": 4.180079530391971,
+      "grad_norm": 0.5497419834136963,
+      "learning_rate": 3.2796818784321154e-06,
+      "loss": 0.0969,
+      "step": 44150
+    },
+    {
+      "epoch": 4.181026320772581,
+      "grad_norm": 0.4828396141529083,
+      "learning_rate": 3.2758947169096767e-06,
+      "loss": 0.0869,
+      "step": 44160
+    },
+    {
+      "epoch": 4.181973111153191,
+      "grad_norm": 0.46545299887657166,
+      "learning_rate": 3.2721075553872376e-06,
+      "loss": 0.0841,
+      "step": 44170
+    },
+    {
+      "epoch": 4.1829199015338006,
+      "grad_norm": 0.6381782293319702,
+      "learning_rate": 3.2683203938647985e-06,
+      "loss": 0.0926,
+      "step": 44180
+    },
+    {
+      "epoch": 4.18386669191441,
+      "grad_norm": 0.5599507093429565,
+      "learning_rate": 3.2645332323423594e-06,
+      "loss": 0.0921,
+      "step": 44190
+    },
+    {
+      "epoch": 4.18481348229502,
+      "grad_norm": 0.5431086421012878,
+      "learning_rate": 3.2607460708199207e-06,
+      "loss": 0.095,
+      "step": 44200
+    },
+    {
+      "epoch": 4.18576027267563,
+      "grad_norm": 0.6155442595481873,
+      "learning_rate": 3.256958909297482e-06,
+      "loss": 0.101,
+      "step": 44210
+    },
+    {
+      "epoch": 4.1867070630562395,
+      "grad_norm": 0.489976704120636,
+      "learning_rate": 3.253171747775043e-06,
+      "loss": 0.0864,
+      "step": 44220
+    },
+    {
+      "epoch": 4.187653853436849,
+      "grad_norm": 0.6099218130111694,
+      "learning_rate": 3.2493845862526042e-06,
+      "loss": 0.0933,
+      "step": 44230
+    },
+    {
+      "epoch": 4.188600643817459,
+      "grad_norm": 0.5284793376922607,
+      "learning_rate": 3.245597424730165e-06,
+      "loss": 0.0897,
+      "step": 44240
+    },
+    {
+      "epoch": 4.189547434198069,
+      "grad_norm": 0.5145465731620789,
+      "learning_rate": 3.241810263207726e-06,
+      "loss": 0.0909,
+      "step": 44250
+    },
+    {
+      "epoch": 4.1904942245786785,
+      "grad_norm": 0.6255049109458923,
+      "learning_rate": 3.238023101685287e-06,
+      "loss": 0.0878,
+      "step": 44260
+    },
+    {
+      "epoch": 4.191441014959288,
+      "grad_norm": 0.5650139451026917,
+      "learning_rate": 3.2342359401628482e-06,
+      "loss": 0.09,
+      "step": 44270
+    },
+    {
+      "epoch": 4.192387805339898,
+      "grad_norm": 0.5437065362930298,
+      "learning_rate": 3.230448778640409e-06,
+      "loss": 0.0881,
+      "step": 44280
+    },
+    {
+      "epoch": 4.193334595720508,
+      "grad_norm": 0.5027974843978882,
+      "learning_rate": 3.2266616171179705e-06,
+      "loss": 0.0874,
+      "step": 44290
+    },
+    {
+      "epoch": 4.1942813861011174,
+      "grad_norm": 0.484749436378479,
+      "learning_rate": 3.2228744555955318e-06,
+      "loss": 0.0866,
+      "step": 44300
+    },
+    {
+      "epoch": 4.195228176481727,
+      "grad_norm": 0.46499377489089966,
+      "learning_rate": 3.2190872940730927e-06,
+      "loss": 0.0879,
+      "step": 44310
+    },
+    {
+      "epoch": 4.196174966862337,
+      "grad_norm": 0.6362327933311462,
+      "learning_rate": 3.215300132550653e-06,
+      "loss": 0.1016,
+      "step": 44320
+    },
+    {
+      "epoch": 4.197121757242947,
+      "grad_norm": 0.5938698053359985,
+      "learning_rate": 3.2115129710282145e-06,
+      "loss": 0.0904,
+      "step": 44330
+    },
+    {
+      "epoch": 4.198068547623556,
+      "grad_norm": 0.5677269101142883,
+      "learning_rate": 3.207725809505776e-06,
+      "loss": 0.096,
+      "step": 44340
+    },
+    {
+      "epoch": 4.199015338004166,
+      "grad_norm": 0.5697394609451294,
+      "learning_rate": 3.2039386479833367e-06,
+      "loss": 0.0902,
+      "step": 44350
+    },
+    {
+      "epoch": 4.199962128384776,
+      "grad_norm": 0.7053269743919373,
+      "learning_rate": 3.200151486460898e-06,
+      "loss": 0.0922,
+      "step": 44360
+    },
+    {
+      "epoch": 4.200908918765386,
+      "grad_norm": 0.5635038018226624,
+      "learning_rate": 3.196364324938459e-06,
+      "loss": 0.0905,
+      "step": 44370
+    },
+    {
+      "epoch": 4.201855709145995,
+      "grad_norm": 0.7025290131568909,
+      "learning_rate": 3.1925771634160202e-06,
+      "loss": 0.0978,
+      "step": 44380
+    },
+    {
+      "epoch": 4.202802499526605,
+      "grad_norm": 0.44185054302215576,
+      "learning_rate": 3.1887900018935807e-06,
+      "loss": 0.0942,
+      "step": 44390
+    },
+    {
+      "epoch": 4.203749289907215,
+      "grad_norm": 0.5605813264846802,
+      "learning_rate": 3.185002840371142e-06,
+      "loss": 0.0846,
+      "step": 44400
+    },
+    {
+      "epoch": 4.204696080287825,
+      "grad_norm": 0.541060745716095,
+      "learning_rate": 3.181215678848703e-06,
+      "loss": 0.0915,
+      "step": 44410
+    },
+    {
+      "epoch": 4.205642870668434,
+      "grad_norm": 0.6130930185317993,
+      "learning_rate": 3.1774285173262642e-06,
+      "loss": 0.088,
+      "step": 44420
+    },
+    {
+      "epoch": 4.206589661049044,
+      "grad_norm": 0.4869426190853119,
+      "learning_rate": 3.1736413558038256e-06,
+      "loss": 0.086,
+      "step": 44430
+    },
+    {
+      "epoch": 4.207536451429654,
+      "grad_norm": 0.5549150109291077,
+      "learning_rate": 3.1698541942813865e-06,
+      "loss": 0.1032,
+      "step": 44440
+    },
+    {
+      "epoch": 4.208483241810264,
+      "grad_norm": 0.4869174063205719,
+      "learning_rate": 3.1660670327589478e-06,
+      "loss": 0.0957,
+      "step": 44450
+    },
+    {
+      "epoch": 4.209430032190873,
+      "grad_norm": 0.43977847695350647,
+      "learning_rate": 3.1622798712365082e-06,
+      "loss": 0.0936,
+      "step": 44460
+    },
+    {
+      "epoch": 4.210376822571483,
+      "grad_norm": 0.502215564250946,
+      "learning_rate": 3.1584927097140696e-06,
+      "loss": 0.0909,
+      "step": 44470
+    },
+    {
+      "epoch": 4.211323612952093,
+      "grad_norm": 0.53812175989151,
+      "learning_rate": 3.1547055481916305e-06,
+      "loss": 0.0854,
+      "step": 44480
+    },
+    {
+      "epoch": 4.2122704033327025,
+      "grad_norm": 0.4930614233016968,
+      "learning_rate": 3.1509183866691918e-06,
+      "loss": 0.0879,
+      "step": 44490
+    },
+    {
+      "epoch": 4.213217193713312,
+      "grad_norm": 0.4674643576145172,
+      "learning_rate": 3.1471312251467527e-06,
+      "loss": 0.0847,
+      "step": 44500
+    },
+    {
+      "epoch": 4.214163984093922,
+      "grad_norm": 0.5616251826286316,
+      "learning_rate": 3.143344063624314e-06,
+      "loss": 0.0863,
+      "step": 44510
+    },
+    {
+      "epoch": 4.215110774474532,
+      "grad_norm": 0.6773747801780701,
+      "learning_rate": 3.1395569021018753e-06,
+      "loss": 0.0955,
+      "step": 44520
+    },
+    {
+      "epoch": 4.2160575648551415,
+      "grad_norm": 0.4281054735183716,
+      "learning_rate": 3.135769740579436e-06,
+      "loss": 0.0897,
+      "step": 44530
+    },
+    {
+      "epoch": 4.217004355235751,
+      "grad_norm": 0.48651763796806335,
+      "learning_rate": 3.1319825790569967e-06,
+      "loss": 0.0889,
+      "step": 44540
+    },
+    {
+      "epoch": 4.217951145616361,
+      "grad_norm": 0.4472315013408661,
+      "learning_rate": 3.128195417534558e-06,
+      "loss": 0.0892,
+      "step": 44550
+    },
+    {
+      "epoch": 4.218897935996971,
+      "grad_norm": 0.732990026473999,
+      "learning_rate": 3.1244082560121193e-06,
+      "loss": 0.0944,
+      "step": 44560
+    },
+    {
+      "epoch": 4.2198447263775805,
+      "grad_norm": 0.6399824023246765,
+      "learning_rate": 3.1206210944896802e-06,
+      "loss": 0.0925,
+      "step": 44570
+    },
+    {
+      "epoch": 4.220791516758189,
+      "grad_norm": 0.5704659819602966,
+      "learning_rate": 3.1168339329672415e-06,
+      "loss": 0.0967,
+      "step": 44580
+    },
+    {
+      "epoch": 4.221738307138799,
+      "grad_norm": 0.4731642007827759,
+      "learning_rate": 3.1130467714448024e-06,
+      "loss": 0.0763,
+      "step": 44590
+    },
+    {
+      "epoch": 4.222685097519409,
+      "grad_norm": 0.4535263180732727,
+      "learning_rate": 3.1092596099223633e-06,
+      "loss": 0.0871,
+      "step": 44600
+    },
+    {
+      "epoch": 4.2236318879000185,
+      "grad_norm": 0.5828526020050049,
+      "learning_rate": 3.1054724483999242e-06,
+      "loss": 0.0829,
+      "step": 44610
+    },
+    {
+      "epoch": 4.224578678280628,
+      "grad_norm": 0.6800204515457153,
+      "learning_rate": 3.1016852868774856e-06,
+      "loss": 0.0786,
+      "step": 44620
+    },
+    {
+      "epoch": 4.225525468661238,
+      "grad_norm": 0.49945423007011414,
+      "learning_rate": 3.0978981253550465e-06,
+      "loss": 0.09,
+      "step": 44630
+    },
+    {
+      "epoch": 4.226472259041848,
+      "grad_norm": 0.5540609955787659,
+      "learning_rate": 3.0941109638326078e-06,
+      "loss": 0.0906,
+      "step": 44640
+    },
+    {
+      "epoch": 4.2274190494224575,
+      "grad_norm": 0.47390490770339966,
+      "learning_rate": 3.090323802310169e-06,
+      "loss": 0.0908,
+      "step": 44650
+    },
+    {
+      "epoch": 4.228365839803067,
+      "grad_norm": 0.5915796756744385,
+      "learning_rate": 3.08653664078773e-06,
+      "loss": 0.0917,
+      "step": 44660
+    },
+    {
+      "epoch": 4.229312630183677,
+      "grad_norm": 0.5430005192756653,
+      "learning_rate": 3.0827494792652905e-06,
+      "loss": 0.0953,
+      "step": 44670
+    },
+    {
+      "epoch": 4.230259420564287,
+      "grad_norm": 0.6260116696357727,
+      "learning_rate": 3.0789623177428518e-06,
+      "loss": 0.0943,
+      "step": 44680
+    },
+    {
+      "epoch": 4.2312062109448965,
+      "grad_norm": 0.6242892146110535,
+      "learning_rate": 3.075175156220413e-06,
+      "loss": 0.0951,
+      "step": 44690
+    },
+    {
+      "epoch": 4.232153001325506,
+      "grad_norm": 0.5777116417884827,
+      "learning_rate": 3.071387994697974e-06,
+      "loss": 0.0939,
+      "step": 44700
+    },
+    {
+      "epoch": 4.233099791706116,
+      "grad_norm": 0.638270914554596,
+      "learning_rate": 3.0676008331755353e-06,
+      "loss": 0.0942,
+      "step": 44710
+    },
+    {
+      "epoch": 4.234046582086726,
+      "grad_norm": 0.5501429438591003,
+      "learning_rate": 3.0638136716530962e-06,
+      "loss": 0.1015,
+      "step": 44720
+    },
+    {
+      "epoch": 4.234993372467335,
+      "grad_norm": 0.5100410580635071,
+      "learning_rate": 3.0600265101306575e-06,
+      "loss": 0.0945,
+      "step": 44730
+    },
+    {
+      "epoch": 4.235940162847945,
+      "grad_norm": 0.4826267659664154,
+      "learning_rate": 3.056239348608218e-06,
+      "loss": 0.0921,
+      "step": 44740
+    },
+    {
+      "epoch": 4.236886953228555,
+      "grad_norm": 0.6098191738128662,
+      "learning_rate": 3.0524521870857793e-06,
+      "loss": 0.0859,
+      "step": 44750
+    },
+    {
+      "epoch": 4.237833743609165,
+      "grad_norm": 0.5259220600128174,
+      "learning_rate": 3.0486650255633402e-06,
+      "loss": 0.0962,
+      "step": 44760
+    },
+    {
+      "epoch": 4.238780533989774,
+      "grad_norm": 0.6392509937286377,
+      "learning_rate": 3.0448778640409015e-06,
+      "loss": 0.1002,
+      "step": 44770
+    },
+    {
+      "epoch": 4.239727324370384,
+      "grad_norm": 0.673291802406311,
+      "learning_rate": 3.041090702518463e-06,
+      "loss": 0.0896,
+      "step": 44780
+    },
+    {
+      "epoch": 4.240674114750994,
+      "grad_norm": 0.5085720419883728,
+      "learning_rate": 3.0373035409960238e-06,
+      "loss": 0.0902,
+      "step": 44790
+    },
+    {
+      "epoch": 4.241620905131604,
+      "grad_norm": 0.5305391550064087,
+      "learning_rate": 3.033516379473585e-06,
+      "loss": 0.0941,
+      "step": 44800
+    },
+    {
+      "epoch": 4.242567695512213,
+      "grad_norm": 0.4959773123264313,
+      "learning_rate": 3.0297292179511456e-06,
+      "loss": 0.0939,
+      "step": 44810
+    },
+    {
+      "epoch": 4.243514485892823,
+      "grad_norm": 0.6148700714111328,
+      "learning_rate": 3.025942056428707e-06,
+      "loss": 0.1018,
+      "step": 44820
+    },
+    {
+      "epoch": 4.244461276273433,
+      "grad_norm": 0.5363166928291321,
+      "learning_rate": 3.0221548949062678e-06,
+      "loss": 0.0916,
+      "step": 44830
+    },
+    {
+      "epoch": 4.245408066654043,
+      "grad_norm": 0.5104047656059265,
+      "learning_rate": 3.018367733383829e-06,
+      "loss": 0.0917,
+      "step": 44840
+    },
+    {
+      "epoch": 4.246354857034652,
+      "grad_norm": 0.5864166021347046,
+      "learning_rate": 3.01458057186139e-06,
+      "loss": 0.0875,
+      "step": 44850
+    },
+    {
+      "epoch": 4.247301647415262,
+      "grad_norm": 0.6654096245765686,
+      "learning_rate": 3.0107934103389513e-06,
+      "loss": 0.0902,
+      "step": 44860
+    },
+    {
+      "epoch": 4.248248437795872,
+      "grad_norm": 0.49257123470306396,
+      "learning_rate": 3.0070062488165126e-06,
+      "loss": 0.0956,
+      "step": 44870
+    },
+    {
+      "epoch": 4.2491952281764815,
+      "grad_norm": 0.4794057309627533,
+      "learning_rate": 3.003219087294073e-06,
+      "loss": 0.0885,
+      "step": 44880
+    },
+    {
+      "epoch": 4.250142018557091,
+      "grad_norm": 0.5510391592979431,
+      "learning_rate": 2.9994319257716344e-06,
+      "loss": 0.0933,
+      "step": 44890
+    },
+    {
+      "epoch": 4.251088808937701,
+      "grad_norm": 0.5095749497413635,
+      "learning_rate": 2.9956447642491953e-06,
+      "loss": 0.1042,
+      "step": 44900
+    },
+    {
+      "epoch": 4.252035599318311,
+      "grad_norm": 0.5804000496864319,
+      "learning_rate": 2.9918576027267566e-06,
+      "loss": 0.0976,
+      "step": 44910
+    },
+    {
+      "epoch": 4.2529823896989205,
+      "grad_norm": 0.47743797302246094,
+      "learning_rate": 2.9880704412043175e-06,
+      "loss": 0.0949,
+      "step": 44920
+    },
+    {
+      "epoch": 4.25392918007953,
+      "grad_norm": 0.5327832102775574,
+      "learning_rate": 2.984283279681879e-06,
+      "loss": 0.0912,
+      "step": 44930
+    },
+    {
+      "epoch": 4.25487597046014,
+      "grad_norm": 0.6161805391311646,
+      "learning_rate": 2.9804961181594398e-06,
+      "loss": 0.0891,
+      "step": 44940
+    },
+    {
+      "epoch": 4.25582276084075,
+      "grad_norm": 0.6115200519561768,
+      "learning_rate": 2.9767089566370006e-06,
+      "loss": 0.1009,
+      "step": 44950
+    },
+    {
+      "epoch": 4.2567695512213595,
+      "grad_norm": 0.5854682922363281,
+      "learning_rate": 2.9729217951145615e-06,
+      "loss": 0.0943,
+      "step": 44960
+    },
+    {
+      "epoch": 4.257716341601969,
+      "grad_norm": 0.4380771219730377,
+      "learning_rate": 2.969134633592123e-06,
+      "loss": 0.0829,
+      "step": 44970
+    },
+    {
+      "epoch": 4.258663131982579,
+      "grad_norm": 0.5447367429733276,
+      "learning_rate": 2.965347472069684e-06,
+      "loss": 0.0949,
+      "step": 44980
+    },
+    {
+      "epoch": 4.259609922363189,
+      "grad_norm": 0.4817304313182831,
+      "learning_rate": 2.961560310547245e-06,
+      "loss": 0.0893,
+      "step": 44990
+    },
+    {
+      "epoch": 4.260556712743798,
+      "grad_norm": 0.37457752227783203,
+      "learning_rate": 2.9577731490248064e-06,
+      "loss": 0.0844,
+      "step": 45000
+    },
+    {
+      "epoch": 4.261503503124408,
+      "grad_norm": 0.5563647747039795,
+      "learning_rate": 2.9539859875023673e-06,
+      "loss": 0.094,
+      "step": 45010
+    },
+    {
+      "epoch": 4.262450293505018,
+      "grad_norm": 0.5248158574104309,
+      "learning_rate": 2.950198825979928e-06,
+      "loss": 0.0954,
+      "step": 45020
+    },
+    {
+      "epoch": 4.263397083885628,
+      "grad_norm": 0.5105319023132324,
+      "learning_rate": 2.946411664457489e-06,
+      "loss": 0.0929,
+      "step": 45030
+    },
+    {
+      "epoch": 4.264343874266237,
+      "grad_norm": 0.4475686252117157,
+      "learning_rate": 2.9426245029350504e-06,
+      "loss": 0.0855,
+      "step": 45040
+    },
+    {
+      "epoch": 4.265290664646847,
+      "grad_norm": 0.46911075711250305,
+      "learning_rate": 2.9388373414126113e-06,
+      "loss": 0.0874,
+      "step": 45050
+    },
+    {
+      "epoch": 4.266237455027457,
+      "grad_norm": 0.579980731010437,
+      "learning_rate": 2.9350501798901726e-06,
+      "loss": 0.0953,
+      "step": 45060
+    },
+    {
+      "epoch": 4.267184245408067,
+      "grad_norm": 0.5501284599304199,
+      "learning_rate": 2.931263018367734e-06,
+      "loss": 0.0919,
+      "step": 45070
+    },
+    {
+      "epoch": 4.268131035788676,
+      "grad_norm": 0.44050830602645874,
+      "learning_rate": 2.927475856845295e-06,
+      "loss": 0.0929,
+      "step": 45080
+    },
+    {
+      "epoch": 4.269077826169286,
+      "grad_norm": 0.6507890820503235,
+      "learning_rate": 2.9236886953228553e-06,
+      "loss": 0.0945,
+      "step": 45090
+    },
+    {
+      "epoch": 4.270024616549896,
+      "grad_norm": 0.5835983753204346,
+      "learning_rate": 2.9199015338004166e-06,
+      "loss": 0.0931,
+      "step": 45100
+    },
+    {
+      "epoch": 4.270971406930506,
+      "grad_norm": 0.6553801894187927,
+      "learning_rate": 2.916114372277978e-06,
+      "loss": 0.0948,
+      "step": 45110
+    },
+    {
+      "epoch": 4.271918197311115,
+      "grad_norm": 0.4947971701622009,
+      "learning_rate": 2.912327210755539e-06,
+      "loss": 0.0886,
+      "step": 45120
+    },
+    {
+      "epoch": 4.272864987691725,
+      "grad_norm": 0.6035920977592468,
+      "learning_rate": 2.9085400492331e-06,
+      "loss": 0.0933,
+      "step": 45130
+    },
+    {
+      "epoch": 4.273811778072335,
+      "grad_norm": 0.46540600061416626,
+      "learning_rate": 2.904752887710661e-06,
+      "loss": 0.0925,
+      "step": 45140
+    },
+    {
+      "epoch": 4.2747585684529446,
+      "grad_norm": 0.5728728771209717,
+      "learning_rate": 2.9009657261882224e-06,
+      "loss": 0.0954,
+      "step": 45150
+    },
+    {
+      "epoch": 4.275705358833554,
+      "grad_norm": 0.5419313311576843,
+      "learning_rate": 2.897178564665783e-06,
+      "loss": 0.0843,
+      "step": 45160
+    },
+    {
+      "epoch": 4.276652149214164,
+      "grad_norm": 0.41763198375701904,
+      "learning_rate": 2.893391403143344e-06,
+      "loss": 0.0896,
+      "step": 45170
+    },
+    {
+      "epoch": 4.277598939594774,
+      "grad_norm": 0.4943348467350006,
+      "learning_rate": 2.889604241620905e-06,
+      "loss": 0.092,
+      "step": 45180
+    },
+    {
+      "epoch": 4.2785457299753835,
+      "grad_norm": 0.5788520574569702,
+      "learning_rate": 2.8858170800984664e-06,
+      "loss": 0.0891,
+      "step": 45190
+    },
+    {
+      "epoch": 4.279492520355993,
+      "grad_norm": 0.4970496594905853,
+      "learning_rate": 2.8820299185760277e-06,
+      "loss": 0.0962,
+      "step": 45200
+    },
+    {
+      "epoch": 4.280439310736603,
+      "grad_norm": 0.6108376979827881,
+      "learning_rate": 2.8782427570535886e-06,
+      "loss": 0.0921,
+      "step": 45210
+    },
+    {
+      "epoch": 4.281386101117213,
+      "grad_norm": 0.55495685338974,
+      "learning_rate": 2.87445559553115e-06,
+      "loss": 0.0909,
+      "step": 45220
+    },
+    {
+      "epoch": 4.2823328914978225,
+      "grad_norm": 0.5994839668273926,
+      "learning_rate": 2.870668434008711e-06,
+      "loss": 0.0983,
+      "step": 45230
+    },
+    {
+      "epoch": 4.283279681878432,
+      "grad_norm": 0.5374669432640076,
+      "learning_rate": 2.8668812724862717e-06,
+      "loss": 0.0994,
+      "step": 45240
+    },
+    {
+      "epoch": 4.284226472259042,
+      "grad_norm": 0.5874958038330078,
+      "learning_rate": 2.8630941109638326e-06,
+      "loss": 0.0955,
+      "step": 45250
+    },
+    {
+      "epoch": 4.285173262639652,
+      "grad_norm": 0.6345770359039307,
+      "learning_rate": 2.859306949441394e-06,
+      "loss": 0.0882,
+      "step": 45260
+    },
+    {
+      "epoch": 4.2861200530202614,
+      "grad_norm": 0.5831238627433777,
+      "learning_rate": 2.855519787918955e-06,
+      "loss": 0.093,
+      "step": 45270
+    },
+    {
+      "epoch": 4.287066843400871,
+      "grad_norm": 0.6263869404792786,
+      "learning_rate": 2.851732626396516e-06,
+      "loss": 0.0918,
+      "step": 45280
+    },
+    {
+      "epoch": 4.288013633781481,
+      "grad_norm": 0.5837584733963013,
+      "learning_rate": 2.8479454648740775e-06,
+      "loss": 0.0931,
+      "step": 45290
+    },
+    {
+      "epoch": 4.288960424162091,
+      "grad_norm": 0.4665307104587555,
+      "learning_rate": 2.8441583033516384e-06,
+      "loss": 0.0907,
+      "step": 45300
+    },
+    {
+      "epoch": 4.2899072145427,
+      "grad_norm": 0.5143084526062012,
+      "learning_rate": 2.840371141829199e-06,
+      "loss": 0.1047,
+      "step": 45310
+    },
+    {
+      "epoch": 4.29085400492331,
+      "grad_norm": 0.6588802933692932,
+      "learning_rate": 2.83658398030676e-06,
+      "loss": 0.0964,
+      "step": 45320
+    },
+    {
+      "epoch": 4.29180079530392,
+      "grad_norm": 0.6557096242904663,
+      "learning_rate": 2.8327968187843215e-06,
+      "loss": 0.0914,
+      "step": 45330
+    },
+    {
+      "epoch": 4.29274758568453,
+      "grad_norm": 0.5459699630737305,
+      "learning_rate": 2.8290096572618824e-06,
+      "loss": 0.0905,
+      "step": 45340
+    },
+    {
+      "epoch": 4.293694376065139,
+      "grad_norm": 0.5911346673965454,
+      "learning_rate": 2.8252224957394437e-06,
+      "loss": 0.088,
+      "step": 45350
+    },
+    {
+      "epoch": 4.294641166445749,
+      "grad_norm": 0.5709453821182251,
+      "learning_rate": 2.8214353342170046e-06,
+      "loss": 0.103,
+      "step": 45360
+    },
+    {
+      "epoch": 4.295587956826359,
+      "grad_norm": 0.7081350684165955,
+      "learning_rate": 2.817648172694566e-06,
+      "loss": 0.0754,
+      "step": 45370
+    },
+    {
+      "epoch": 4.296534747206969,
+      "grad_norm": 0.529180645942688,
+      "learning_rate": 2.8138610111721264e-06,
+      "loss": 0.0876,
+      "step": 45380
+    },
+    {
+      "epoch": 4.297481537587578,
+      "grad_norm": 0.5527861714363098,
+      "learning_rate": 2.8100738496496877e-06,
+      "loss": 0.085,
+      "step": 45390
+    },
+    {
+      "epoch": 4.298428327968188,
+      "grad_norm": 0.5242286920547485,
+      "learning_rate": 2.8062866881272486e-06,
+      "loss": 0.0858,
+      "step": 45400
+    },
+    {
+      "epoch": 4.299375118348798,
+      "grad_norm": 0.5668191313743591,
+      "learning_rate": 2.80249952660481e-06,
+      "loss": 0.0865,
+      "step": 45410
+    },
+    {
+      "epoch": 4.300321908729408,
+      "grad_norm": 0.3957083821296692,
+      "learning_rate": 2.7987123650823713e-06,
+      "loss": 0.0847,
+      "step": 45420
+    },
+    {
+      "epoch": 4.301268699110017,
+      "grad_norm": 0.458646297454834,
+      "learning_rate": 2.794925203559932e-06,
+      "loss": 0.0982,
+      "step": 45430
+    },
+    {
+      "epoch": 4.302215489490627,
+      "grad_norm": 0.45908018946647644,
+      "learning_rate": 2.7911380420374935e-06,
+      "loss": 0.0922,
+      "step": 45440
+    },
+    {
+      "epoch": 4.303162279871237,
+      "grad_norm": 0.6364571452140808,
+      "learning_rate": 2.787350880515054e-06,
+      "loss": 0.0889,
+      "step": 45450
+    },
+    {
+      "epoch": 4.3041090702518465,
+      "grad_norm": 0.5598451495170593,
+      "learning_rate": 2.7835637189926153e-06,
+      "loss": 0.0781,
+      "step": 45460
+    },
+    {
+      "epoch": 4.305055860632456,
+      "grad_norm": 0.44022366404533386,
+      "learning_rate": 2.779776557470176e-06,
+      "loss": 0.0911,
+      "step": 45470
+    },
+    {
+      "epoch": 4.306002651013066,
+      "grad_norm": 0.6158979535102844,
+      "learning_rate": 2.7759893959477375e-06,
+      "loss": 0.0982,
+      "step": 45480
+    },
+    {
+      "epoch": 4.306949441393676,
+      "grad_norm": 0.6025349497795105,
+      "learning_rate": 2.7722022344252984e-06,
+      "loss": 0.1083,
+      "step": 45490
+    },
+    {
+      "epoch": 4.3078962317742855,
+      "grad_norm": 0.5192539691925049,
+      "learning_rate": 2.7684150729028597e-06,
+      "loss": 0.0909,
+      "step": 45500
+    },
+    {
+      "epoch": 4.308843022154895,
+      "grad_norm": 0.592406153678894,
+      "learning_rate": 2.764627911380421e-06,
+      "loss": 0.0974,
+      "step": 45510
+    },
+    {
+      "epoch": 4.309789812535505,
+      "grad_norm": 0.3901938199996948,
+      "learning_rate": 2.7608407498579815e-06,
+      "loss": 0.0904,
+      "step": 45520
+    },
+    {
+      "epoch": 4.310736602916115,
+      "grad_norm": 0.5858525037765503,
+      "learning_rate": 2.7570535883355424e-06,
+      "loss": 0.0891,
+      "step": 45530
+    },
+    {
+      "epoch": 4.3116833932967245,
+      "grad_norm": 0.7031994462013245,
+      "learning_rate": 2.7532664268131037e-06,
+      "loss": 0.1008,
+      "step": 45540
+    },
+    {
+      "epoch": 4.312630183677334,
+      "grad_norm": 0.4433521628379822,
+      "learning_rate": 2.749479265290665e-06,
+      "loss": 0.0892,
+      "step": 45550
+    },
+    {
+      "epoch": 4.313576974057944,
+      "grad_norm": 0.4655055105686188,
+      "learning_rate": 2.745692103768226e-06,
+      "loss": 0.0889,
+      "step": 45560
+    },
+    {
+      "epoch": 4.314523764438554,
+      "grad_norm": 0.5043991804122925,
+      "learning_rate": 2.7419049422457872e-06,
+      "loss": 0.0945,
+      "step": 45570
+    },
+    {
+      "epoch": 4.315470554819163,
+      "grad_norm": 0.4252864420413971,
+      "learning_rate": 2.738117780723348e-06,
+      "loss": 0.0956,
+      "step": 45580
+    },
+    {
+      "epoch": 4.316417345199773,
+      "grad_norm": 0.6047373414039612,
+      "learning_rate": 2.734330619200909e-06,
+      "loss": 0.0986,
+      "step": 45590
+    },
+    {
+      "epoch": 4.317364135580383,
+      "grad_norm": 0.4813744127750397,
+      "learning_rate": 2.73054345767847e-06,
+      "loss": 0.0969,
+      "step": 45600
+    },
+    {
+      "epoch": 4.318310925960993,
+      "grad_norm": 0.4027179777622223,
+      "learning_rate": 2.7267562961560313e-06,
+      "loss": 0.09,
+      "step": 45610
+    },
+    {
+      "epoch": 4.319257716341602,
+      "grad_norm": 0.682832658290863,
+      "learning_rate": 2.722969134633592e-06,
+      "loss": 0.0912,
+      "step": 45620
+    },
+    {
+      "epoch": 4.320204506722211,
+      "grad_norm": 0.6026838421821594,
+      "learning_rate": 2.7191819731111535e-06,
+      "loss": 0.0972,
+      "step": 45630
+    },
+    {
+      "epoch": 4.321151297102821,
+      "grad_norm": 0.5975409746170044,
+      "learning_rate": 2.715394811588715e-06,
+      "loss": 0.085,
+      "step": 45640
+    },
+    {
+      "epoch": 4.322098087483431,
+      "grad_norm": 0.5089371800422668,
+      "learning_rate": 2.7116076500662757e-06,
+      "loss": 0.0953,
+      "step": 45650
+    },
+    {
+      "epoch": 4.3230448778640405,
+      "grad_norm": 0.4977569580078125,
+      "learning_rate": 2.7078204885438366e-06,
+      "loss": 0.1012,
+      "step": 45660
+    },
+    {
+      "epoch": 4.32399166824465,
+      "grad_norm": 0.7111302614212036,
+      "learning_rate": 2.7040333270213975e-06,
+      "loss": 0.0931,
+      "step": 45670
+    },
+    {
+      "epoch": 4.32493845862526,
+      "grad_norm": 0.4907667934894562,
+      "learning_rate": 2.700246165498959e-06,
+      "loss": 0.0789,
+      "step": 45680
+    },
+    {
+      "epoch": 4.32588524900587,
+      "grad_norm": 0.5581902265548706,
+      "learning_rate": 2.6964590039765197e-06,
+      "loss": 0.0841,
+      "step": 45690
+    },
+    {
+      "epoch": 4.326832039386479,
+      "grad_norm": 0.5252041220664978,
+      "learning_rate": 2.692671842454081e-06,
+      "loss": 0.0931,
+      "step": 45700
+    },
+    {
+      "epoch": 4.327778829767089,
+      "grad_norm": 0.6080724000930786,
+      "learning_rate": 2.688884680931642e-06,
+      "loss": 0.0977,
+      "step": 45710
+    },
+    {
+      "epoch": 4.328725620147699,
+      "grad_norm": 0.4370320737361908,
+      "learning_rate": 2.6850975194092032e-06,
+      "loss": 0.09,
+      "step": 45720
+    },
+    {
+      "epoch": 4.329672410528309,
+      "grad_norm": 0.4596753418445587,
+      "learning_rate": 2.6813103578867637e-06,
+      "loss": 0.0828,
+      "step": 45730
+    },
+    {
+      "epoch": 4.330619200908918,
+      "grad_norm": 0.5326861143112183,
+      "learning_rate": 2.677523196364325e-06,
+      "loss": 0.0834,
+      "step": 45740
+    },
+    {
+      "epoch": 4.331565991289528,
+      "grad_norm": 0.5321633815765381,
+      "learning_rate": 2.6737360348418863e-06,
+      "loss": 0.0885,
+      "step": 45750
+    },
+    {
+      "epoch": 4.332512781670138,
+      "grad_norm": 0.5375781655311584,
+      "learning_rate": 2.6699488733194472e-06,
+      "loss": 0.0909,
+      "step": 45760
+    },
+    {
+      "epoch": 4.333459572050748,
+      "grad_norm": 0.5236536264419556,
+      "learning_rate": 2.6661617117970086e-06,
+      "loss": 0.0964,
+      "step": 45770
+    },
+    {
+      "epoch": 4.334406362431357,
+      "grad_norm": 0.4828406274318695,
+      "learning_rate": 2.6623745502745695e-06,
+      "loss": 0.0861,
+      "step": 45780
+    },
+    {
+      "epoch": 4.335353152811967,
+      "grad_norm": 0.4819585382938385,
+      "learning_rate": 2.6585873887521308e-06,
+      "loss": 0.087,
+      "step": 45790
+    },
+    {
+      "epoch": 4.336299943192577,
+      "grad_norm": 0.489366739988327,
+      "learning_rate": 2.6548002272296913e-06,
+      "loss": 0.0862,
+      "step": 45800
+    },
+    {
+      "epoch": 4.337246733573187,
+      "grad_norm": 0.4460524320602417,
+      "learning_rate": 2.6510130657072526e-06,
+      "loss": 0.0917,
+      "step": 45810
+    },
+    {
+      "epoch": 4.338193523953796,
+      "grad_norm": 0.562499463558197,
+      "learning_rate": 2.6472259041848135e-06,
+      "loss": 0.0886,
+      "step": 45820
+    },
+    {
+      "epoch": 4.339140314334406,
+      "grad_norm": 0.4999210834503174,
+      "learning_rate": 2.643438742662375e-06,
+      "loss": 0.0786,
+      "step": 45830
+    },
+    {
+      "epoch": 4.340087104715016,
+      "grad_norm": 0.5683452486991882,
+      "learning_rate": 2.639651581139936e-06,
+      "loss": 0.1064,
+      "step": 45840
+    },
+    {
+      "epoch": 4.3410338950956255,
+      "grad_norm": 0.6871407628059387,
+      "learning_rate": 2.635864419617497e-06,
+      "loss": 0.0964,
+      "step": 45850
+    },
+    {
+      "epoch": 4.341980685476235,
+      "grad_norm": 0.5813667178153992,
+      "learning_rate": 2.6320772580950583e-06,
+      "loss": 0.0922,
+      "step": 45860
+    },
+    {
+      "epoch": 4.342927475856845,
+      "grad_norm": 0.6370223760604858,
+      "learning_rate": 2.628290096572619e-06,
+      "loss": 0.0942,
+      "step": 45870
+    },
+    {
+      "epoch": 4.343874266237455,
+      "grad_norm": 0.4169613718986511,
+      "learning_rate": 2.62450293505018e-06,
+      "loss": 0.0964,
+      "step": 45880
+    },
+    {
+      "epoch": 4.3448210566180645,
+      "grad_norm": 0.5772116780281067,
+      "learning_rate": 2.620715773527741e-06,
+      "loss": 0.0938,
+      "step": 45890
+    },
+    {
+      "epoch": 4.345767846998674,
+      "grad_norm": 0.5945174098014832,
+      "learning_rate": 2.6169286120053023e-06,
+      "loss": 0.0913,
+      "step": 45900
+    },
+    {
+      "epoch": 4.346714637379284,
+      "grad_norm": 0.6152166724205017,
+      "learning_rate": 2.6131414504828632e-06,
+      "loss": 0.097,
+      "step": 45910
+    },
+    {
+      "epoch": 4.347661427759894,
+      "grad_norm": 0.6659379601478577,
+      "learning_rate": 2.6093542889604246e-06,
+      "loss": 0.0872,
+      "step": 45920
+    },
+    {
+      "epoch": 4.3486082181405035,
+      "grad_norm": 0.5365942120552063,
+      "learning_rate": 2.605567127437986e-06,
+      "loss": 0.0881,
+      "step": 45930
+    },
+    {
+      "epoch": 4.349555008521113,
+      "grad_norm": 0.5780148506164551,
+      "learning_rate": 2.6017799659155463e-06,
+      "loss": 0.0986,
+      "step": 45940
+    },
+    {
+      "epoch": 4.350501798901723,
+      "grad_norm": 0.530954122543335,
+      "learning_rate": 2.5979928043931072e-06,
+      "loss": 0.0837,
+      "step": 45950
+    },
+    {
+      "epoch": 4.351448589282333,
+      "grad_norm": 0.49948498606681824,
+      "learning_rate": 2.5942056428706686e-06,
+      "loss": 0.0903,
+      "step": 45960
+    },
+    {
+      "epoch": 4.352395379662942,
+      "grad_norm": 0.5024592876434326,
+      "learning_rate": 2.59041848134823e-06,
+      "loss": 0.0984,
+      "step": 45970
+    },
+    {
+      "epoch": 4.353342170043552,
+      "grad_norm": 0.6981245279312134,
+      "learning_rate": 2.5866313198257908e-06,
+      "loss": 0.0929,
+      "step": 45980
+    },
+    {
+      "epoch": 4.354288960424162,
+      "grad_norm": 0.5182442665100098,
+      "learning_rate": 2.582844158303352e-06,
+      "loss": 0.0947,
+      "step": 45990
+    },
+    {
+      "epoch": 4.355235750804772,
+      "grad_norm": 0.4673933684825897,
+      "learning_rate": 2.579056996780913e-06,
+      "loss": 0.0954,
+      "step": 46000
+    },
+    {
+      "epoch": 4.356182541185381,
+      "grad_norm": 0.5071808695793152,
+      "learning_rate": 2.575269835258474e-06,
+      "loss": 0.0925,
+      "step": 46010
+    },
+    {
+      "epoch": 4.357129331565991,
+      "grad_norm": 0.7042328715324402,
+      "learning_rate": 2.571482673736035e-06,
+      "loss": 0.0966,
+      "step": 46020
+    },
+    {
+      "epoch": 4.358076121946601,
+      "grad_norm": 0.4986877739429474,
+      "learning_rate": 2.567695512213596e-06,
+      "loss": 0.0865,
+      "step": 46030
+    },
+    {
+      "epoch": 4.359022912327211,
+      "grad_norm": 0.6330245137214661,
+      "learning_rate": 2.563908350691157e-06,
+      "loss": 0.0896,
+      "step": 46040
+    },
+    {
+      "epoch": 4.35996970270782,
+      "grad_norm": 0.4807395935058594,
+      "learning_rate": 2.5601211891687183e-06,
+      "loss": 0.0935,
+      "step": 46050
+    },
+    {
+      "epoch": 4.36091649308843,
+      "grad_norm": 0.5322906374931335,
+      "learning_rate": 2.5563340276462796e-06,
+      "loss": 0.0862,
+      "step": 46060
+    },
+    {
+      "epoch": 4.36186328346904,
+      "grad_norm": 0.6274631023406982,
+      "learning_rate": 2.5525468661238405e-06,
+      "loss": 0.0954,
+      "step": 46070
+    },
+    {
+      "epoch": 4.36281007384965,
+      "grad_norm": 0.5602927207946777,
+      "learning_rate": 2.548759704601401e-06,
+      "loss": 0.0784,
+      "step": 46080
+    },
+    {
+      "epoch": 4.363756864230259,
+      "grad_norm": 0.517584502696991,
+      "learning_rate": 2.5449725430789623e-06,
+      "loss": 0.083,
+      "step": 46090
+    },
+    {
+      "epoch": 4.364703654610869,
+      "grad_norm": 0.6797549724578857,
+      "learning_rate": 2.5411853815565237e-06,
+      "loss": 0.1039,
+      "step": 46100
+    },
+    {
+      "epoch": 4.365650444991479,
+      "grad_norm": 0.5814264416694641,
+      "learning_rate": 2.5373982200340846e-06,
+      "loss": 0.0909,
+      "step": 46110
+    },
+    {
+      "epoch": 4.3665972353720885,
+      "grad_norm": 0.6943686604499817,
+      "learning_rate": 2.533611058511646e-06,
+      "loss": 0.0994,
+      "step": 46120
+    },
+    {
+      "epoch": 4.367544025752698,
+      "grad_norm": 0.554736316204071,
+      "learning_rate": 2.5298238969892068e-06,
+      "loss": 0.0923,
+      "step": 46130
+    },
+    {
+      "epoch": 4.368490816133308,
+      "grad_norm": 0.5204230546951294,
+      "learning_rate": 2.526036735466768e-06,
+      "loss": 0.0977,
+      "step": 46140
+    },
+    {
+      "epoch": 4.369437606513918,
+      "grad_norm": 0.6369757056236267,
+      "learning_rate": 2.5222495739443286e-06,
+      "loss": 0.0994,
+      "step": 46150
+    },
+    {
+      "epoch": 4.3703843968945275,
+      "grad_norm": 0.6230180263519287,
+      "learning_rate": 2.51846241242189e-06,
+      "loss": 0.0898,
+      "step": 46160
+    },
+    {
+      "epoch": 4.371331187275137,
+      "grad_norm": 0.5078501105308533,
+      "learning_rate": 2.5146752508994508e-06,
+      "loss": 0.0942,
+      "step": 46170
+    },
+    {
+      "epoch": 4.372277977655747,
+      "grad_norm": 0.5081248879432678,
+      "learning_rate": 2.510888089377012e-06,
+      "loss": 0.0845,
+      "step": 46180
+    },
+    {
+      "epoch": 4.373224768036357,
+      "grad_norm": 0.5332440733909607,
+      "learning_rate": 2.5071009278545734e-06,
+      "loss": 0.0962,
+      "step": 46190
+    },
+    {
+      "epoch": 4.3741715584169665,
+      "grad_norm": 0.44974285364151,
+      "learning_rate": 2.5033137663321343e-06,
+      "loss": 0.0843,
+      "step": 46200
+    },
+    {
+      "epoch": 4.375118348797576,
+      "grad_norm": 0.500975489616394,
+      "learning_rate": 2.499526604809695e-06,
+      "loss": 0.0938,
+      "step": 46210
+    },
+    {
+      "epoch": 4.376065139178186,
+      "grad_norm": 0.5823422074317932,
+      "learning_rate": 2.4957394432872565e-06,
+      "loss": 0.085,
+      "step": 46220
+    },
+    {
+      "epoch": 4.377011929558796,
+      "grad_norm": 0.4945879280567169,
+      "learning_rate": 2.4919522817648174e-06,
+      "loss": 0.0902,
+      "step": 46230
+    },
+    {
+      "epoch": 4.377958719939405,
+      "grad_norm": 0.4944179356098175,
+      "learning_rate": 2.4881651202423783e-06,
+      "loss": 0.1018,
+      "step": 46240
+    },
+    {
+      "epoch": 4.378905510320015,
+      "grad_norm": 0.5736690163612366,
+      "learning_rate": 2.4843779587199396e-06,
+      "loss": 0.0926,
+      "step": 46250
+    },
+    {
+      "epoch": 4.379852300700625,
+      "grad_norm": 0.5215355753898621,
+      "learning_rate": 2.4805907971975005e-06,
+      "loss": 0.09,
+      "step": 46260
+    },
+    {
+      "epoch": 4.380799091081235,
+      "grad_norm": 0.5324762463569641,
+      "learning_rate": 2.476803635675062e-06,
+      "loss": 0.093,
+      "step": 46270
+    },
+    {
+      "epoch": 4.381745881461844,
+      "grad_norm": 0.4614587128162384,
+      "learning_rate": 2.4730164741526228e-06,
+      "loss": 0.0891,
+      "step": 46280
+    },
+    {
+      "epoch": 4.382692671842454,
+      "grad_norm": 0.4408273994922638,
+      "learning_rate": 2.469229312630184e-06,
+      "loss": 0.0968,
+      "step": 46290
+    },
+    {
+      "epoch": 4.383639462223064,
+      "grad_norm": 0.4188309907913208,
+      "learning_rate": 2.465442151107745e-06,
+      "loss": 0.0792,
+      "step": 46300
+    },
+    {
+      "epoch": 4.384586252603674,
+      "grad_norm": 0.5207238793373108,
+      "learning_rate": 2.461654989585306e-06,
+      "loss": 0.092,
+      "step": 46310
+    },
+    {
+      "epoch": 4.385533042984283,
+      "grad_norm": 0.5580665469169617,
+      "learning_rate": 2.457867828062867e-06,
+      "loss": 0.089,
+      "step": 46320
+    },
+    {
+      "epoch": 4.386479833364893,
+      "grad_norm": 0.6222299933433533,
+      "learning_rate": 2.454080666540428e-06,
+      "loss": 0.0925,
+      "step": 46330
+    },
+    {
+      "epoch": 4.387426623745503,
+      "grad_norm": 0.5015923976898193,
+      "learning_rate": 2.4502935050179894e-06,
+      "loss": 0.0992,
+      "step": 46340
+    },
+    {
+      "epoch": 4.388373414126113,
+      "grad_norm": 0.5996383428573608,
+      "learning_rate": 2.4465063434955503e-06,
+      "loss": 0.0803,
+      "step": 46350
+    },
+    {
+      "epoch": 4.389320204506722,
+      "grad_norm": 0.5574952960014343,
+      "learning_rate": 2.442719181973111e-06,
+      "loss": 0.0826,
+      "step": 46360
+    },
+    {
+      "epoch": 4.390266994887332,
+      "grad_norm": 0.6148197054862976,
+      "learning_rate": 2.4389320204506725e-06,
+      "loss": 0.0994,
+      "step": 46370
+    },
+    {
+      "epoch": 4.391213785267942,
+      "grad_norm": 0.5188667178153992,
+      "learning_rate": 2.4351448589282334e-06,
+      "loss": 0.0922,
+      "step": 46380
+    },
+    {
+      "epoch": 4.392160575648552,
+      "grad_norm": 0.5669800043106079,
+      "learning_rate": 2.4313576974057947e-06,
+      "loss": 0.0964,
+      "step": 46390
+    },
+    {
+      "epoch": 4.393107366029161,
+      "grad_norm": 0.5456835031509399,
+      "learning_rate": 2.4275705358833556e-06,
+      "loss": 0.0906,
+      "step": 46400
+    },
+    {
+      "epoch": 4.394054156409771,
+      "grad_norm": 0.4881961941719055,
+      "learning_rate": 2.423783374360917e-06,
+      "loss": 0.0896,
+      "step": 46410
+    },
+    {
+      "epoch": 4.395000946790381,
+      "grad_norm": 0.42108121514320374,
+      "learning_rate": 2.419996212838478e-06,
+      "loss": 0.1075,
+      "step": 46420
+    },
+    {
+      "epoch": 4.3959477371709905,
+      "grad_norm": 0.5552435517311096,
+      "learning_rate": 2.4162090513160387e-06,
+      "loss": 0.096,
+      "step": 46430
+    },
+    {
+      "epoch": 4.3968945275516,
+      "grad_norm": 0.6532780528068542,
+      "learning_rate": 2.4124218897936e-06,
+      "loss": 0.0881,
+      "step": 46440
+    },
+    {
+      "epoch": 4.39784131793221,
+      "grad_norm": 0.4636865258216858,
+      "learning_rate": 2.408634728271161e-06,
+      "loss": 0.0916,
+      "step": 46450
+    },
+    {
+      "epoch": 4.39878810831282,
+      "grad_norm": 0.4746432304382324,
+      "learning_rate": 2.404847566748722e-06,
+      "loss": 0.0911,
+      "step": 46460
+    },
+    {
+      "epoch": 4.3997348986934295,
+      "grad_norm": 0.6184666752815247,
+      "learning_rate": 2.401060405226283e-06,
+      "loss": 0.1002,
+      "step": 46470
+    },
+    {
+      "epoch": 4.400681689074039,
+      "grad_norm": 0.6605782508850098,
+      "learning_rate": 2.3972732437038445e-06,
+      "loss": 0.0955,
+      "step": 46480
+    },
+    {
+      "epoch": 4.401628479454649,
+      "grad_norm": 0.5042291283607483,
+      "learning_rate": 2.393486082181405e-06,
+      "loss": 0.0885,
+      "step": 46490
+    },
+    {
+      "epoch": 4.402575269835259,
+      "grad_norm": 0.5106028318405151,
+      "learning_rate": 2.3896989206589663e-06,
+      "loss": 0.0945,
+      "step": 46500
+    },
+    {
+      "epoch": 4.4035220602158684,
+      "grad_norm": 0.5014446973800659,
+      "learning_rate": 2.3859117591365276e-06,
+      "loss": 0.0889,
+      "step": 46510
+    },
+    {
+      "epoch": 4.404468850596478,
+      "grad_norm": 0.5890821814537048,
+      "learning_rate": 2.3821245976140885e-06,
+      "loss": 0.0865,
+      "step": 46520
+    },
+    {
+      "epoch": 4.405415640977088,
+      "grad_norm": 0.416118860244751,
+      "learning_rate": 2.3783374360916494e-06,
+      "loss": 0.0877,
+      "step": 46530
+    },
+    {
+      "epoch": 4.406362431357698,
+      "grad_norm": 0.7998496890068054,
+      "learning_rate": 2.3745502745692107e-06,
+      "loss": 0.0817,
+      "step": 46540
+    },
+    {
+      "epoch": 4.407309221738307,
+      "grad_norm": 0.48212817311286926,
+      "learning_rate": 2.3707631130467716e-06,
+      "loss": 0.0901,
+      "step": 46550
+    },
+    {
+      "epoch": 4.408256012118917,
+      "grad_norm": 0.44966939091682434,
+      "learning_rate": 2.3669759515243325e-06,
+      "loss": 0.0983,
+      "step": 46560
+    },
+    {
+      "epoch": 4.409202802499527,
+      "grad_norm": 0.4153594672679901,
+      "learning_rate": 2.363188790001894e-06,
+      "loss": 0.0952,
+      "step": 46570
+    },
+    {
+      "epoch": 4.410149592880137,
+      "grad_norm": 0.42512208223342896,
+      "learning_rate": 2.3594016284794547e-06,
+      "loss": 0.0979,
+      "step": 46580
+    },
+    {
+      "epoch": 4.411096383260746,
+      "grad_norm": 0.49382078647613525,
+      "learning_rate": 2.3556144669570156e-06,
+      "loss": 0.0817,
+      "step": 46590
+    },
+    {
+      "epoch": 4.412043173641356,
+      "grad_norm": 0.48640701174736023,
+      "learning_rate": 2.351827305434577e-06,
+      "loss": 0.0823,
+      "step": 46600
+    },
+    {
+      "epoch": 4.412989964021966,
+      "grad_norm": 0.48631972074508667,
+      "learning_rate": 2.3480401439121383e-06,
+      "loss": 0.09,
+      "step": 46610
+    },
+    {
+      "epoch": 4.413936754402576,
+      "grad_norm": 0.6041695475578308,
+      "learning_rate": 2.344252982389699e-06,
+      "loss": 0.0924,
+      "step": 46620
+    },
+    {
+      "epoch": 4.414883544783185,
+      "grad_norm": 0.41758155822753906,
+      "learning_rate": 2.34046582086726e-06,
+      "loss": 0.082,
+      "step": 46630
+    },
+    {
+      "epoch": 4.415830335163795,
+      "grad_norm": 0.47783786058425903,
+      "learning_rate": 2.3366786593448214e-06,
+      "loss": 0.0934,
+      "step": 46640
+    },
+    {
+      "epoch": 4.416777125544405,
+      "grad_norm": 0.5115928053855896,
+      "learning_rate": 2.3328914978223823e-06,
+      "loss": 0.0831,
+      "step": 46650
+    },
+    {
+      "epoch": 4.417723915925015,
+      "grad_norm": 0.5559064745903015,
+      "learning_rate": 2.329104336299943e-06,
+      "loss": 0.0944,
+      "step": 46660
+    },
+    {
+      "epoch": 4.418670706305624,
+      "grad_norm": 0.6243152618408203,
+      "learning_rate": 2.3253171747775045e-06,
+      "loss": 0.0967,
+      "step": 46670
+    },
+    {
+      "epoch": 4.419617496686234,
+      "grad_norm": 0.567179799079895,
+      "learning_rate": 2.3215300132550654e-06,
+      "loss": 0.0927,
+      "step": 46680
+    },
+    {
+      "epoch": 4.420564287066844,
+      "grad_norm": 0.6107397079467773,
+      "learning_rate": 2.3177428517326267e-06,
+      "loss": 0.0885,
+      "step": 46690
+    },
+    {
+      "epoch": 4.4215110774474535,
+      "grad_norm": 0.7441450953483582,
+      "learning_rate": 2.3139556902101876e-06,
+      "loss": 0.089,
+      "step": 46700
+    },
+    {
+      "epoch": 4.422457867828063,
+      "grad_norm": 0.49230247735977173,
+      "learning_rate": 2.3101685286877485e-06,
+      "loss": 0.0904,
+      "step": 46710
+    },
+    {
+      "epoch": 4.423404658208673,
+      "grad_norm": 0.5044174194335938,
+      "learning_rate": 2.30638136716531e-06,
+      "loss": 0.099,
+      "step": 46720
+    },
+    {
+      "epoch": 4.424351448589283,
+      "grad_norm": 0.5213074684143066,
+      "learning_rate": 2.3025942056428707e-06,
+      "loss": 0.1036,
+      "step": 46730
+    },
+    {
+      "epoch": 4.4252982389698925,
+      "grad_norm": 0.5848998427391052,
+      "learning_rate": 2.298807044120432e-06,
+      "loss": 0.0934,
+      "step": 46740
+    },
+    {
+      "epoch": 4.426245029350502,
+      "grad_norm": 0.5705507397651672,
+      "learning_rate": 2.295019882597993e-06,
+      "loss": 0.0944,
+      "step": 46750
+    },
+    {
+      "epoch": 4.427191819731112,
+      "grad_norm": 0.5704928040504456,
+      "learning_rate": 2.2912327210755543e-06,
+      "loss": 0.0906,
+      "step": 46760
+    },
+    {
+      "epoch": 4.428138610111722,
+      "grad_norm": 0.46121203899383545,
+      "learning_rate": 2.287445559553115e-06,
+      "loss": 0.0931,
+      "step": 46770
+    },
+    {
+      "epoch": 4.4290854004923315,
+      "grad_norm": 0.5511819124221802,
+      "learning_rate": 2.283658398030676e-06,
+      "loss": 0.0842,
+      "step": 46780
+    },
+    {
+      "epoch": 4.430032190872941,
+      "grad_norm": 0.6521610021591187,
+      "learning_rate": 2.2798712365082374e-06,
+      "loss": 0.0952,
+      "step": 46790
+    },
+    {
+      "epoch": 4.43097898125355,
+      "grad_norm": 0.48299136757850647,
+      "learning_rate": 2.2760840749857983e-06,
+      "loss": 0.0887,
+      "step": 46800
+    },
+    {
+      "epoch": 4.43192577163416,
+      "grad_norm": 0.5433517694473267,
+      "learning_rate": 2.272296913463359e-06,
+      "loss": 0.0947,
+      "step": 46810
+    },
+    {
+      "epoch": 4.4328725620147695,
+      "grad_norm": 0.4225456118583679,
+      "learning_rate": 2.2685097519409205e-06,
+      "loss": 0.0901,
+      "step": 46820
+    },
+    {
+      "epoch": 4.433819352395379,
+      "grad_norm": 0.5814269781112671,
+      "learning_rate": 2.264722590418482e-06,
+      "loss": 0.0831,
+      "step": 46830
+    },
+    {
+      "epoch": 4.434766142775989,
+      "grad_norm": 0.42077985405921936,
+      "learning_rate": 2.2609354288960427e-06,
+      "loss": 0.0984,
+      "step": 46840
+    },
+    {
+      "epoch": 4.435712933156599,
+      "grad_norm": 0.4770452380180359,
+      "learning_rate": 2.2571482673736036e-06,
+      "loss": 0.0832,
+      "step": 46850
+    },
+    {
+      "epoch": 4.4366597235372085,
+      "grad_norm": 0.47910043597221375,
+      "learning_rate": 2.253361105851165e-06,
+      "loss": 0.0877,
+      "step": 46860
+    },
+    {
+      "epoch": 4.437606513917818,
+      "grad_norm": 0.5875367522239685,
+      "learning_rate": 2.249573944328726e-06,
+      "loss": 0.0954,
+      "step": 46870
+    },
+    {
+      "epoch": 4.438553304298428,
+      "grad_norm": 0.4312451183795929,
+      "learning_rate": 2.2457867828062867e-06,
+      "loss": 0.0939,
+      "step": 46880
+    },
+    {
+      "epoch": 4.439500094679038,
+      "grad_norm": 0.5338612794876099,
+      "learning_rate": 2.241999621283848e-06,
+      "loss": 0.0901,
+      "step": 46890
+    },
+    {
+      "epoch": 4.4404468850596475,
+      "grad_norm": 0.516419529914856,
+      "learning_rate": 2.238212459761409e-06,
+      "loss": 0.0938,
+      "step": 46900
+    },
+    {
+      "epoch": 4.441393675440257,
+      "grad_norm": 0.7233615517616272,
+      "learning_rate": 2.23442529823897e-06,
+      "loss": 0.0881,
+      "step": 46910
+    },
+    {
+      "epoch": 4.442340465820867,
+      "grad_norm": 0.6186655163764954,
+      "learning_rate": 2.230638136716531e-06,
+      "loss": 0.0892,
+      "step": 46920
+    },
+    {
+      "epoch": 4.443287256201477,
+      "grad_norm": 0.5014260411262512,
+      "learning_rate": 2.2268509751940925e-06,
+      "loss": 0.0906,
+      "step": 46930
+    },
+    {
+      "epoch": 4.444234046582086,
+      "grad_norm": 0.5914987325668335,
+      "learning_rate": 2.2230638136716534e-06,
+      "loss": 0.0945,
+      "step": 46940
+    },
+    {
+      "epoch": 4.445180836962696,
+      "grad_norm": 0.5450988411903381,
+      "learning_rate": 2.2192766521492143e-06,
+      "loss": 0.0895,
+      "step": 46950
+    },
+    {
+      "epoch": 4.446127627343306,
+      "grad_norm": 0.5281708240509033,
+      "learning_rate": 2.2154894906267756e-06,
+      "loss": 0.0864,
+      "step": 46960
+    },
+    {
+      "epoch": 4.447074417723916,
+      "grad_norm": 0.5472949743270874,
+      "learning_rate": 2.2117023291043365e-06,
+      "loss": 0.0948,
+      "step": 46970
+    },
+    {
+      "epoch": 4.448021208104525,
+      "grad_norm": 0.6187710165977478,
+      "learning_rate": 2.2079151675818974e-06,
+      "loss": 0.0931,
+      "step": 46980
+    },
+    {
+      "epoch": 4.448967998485135,
+      "grad_norm": 0.4200582802295685,
+      "learning_rate": 2.2041280060594587e-06,
+      "loss": 0.0873,
+      "step": 46990
+    },
+    {
+      "epoch": 4.449914788865745,
+      "grad_norm": 0.4873574674129486,
+      "learning_rate": 2.2003408445370196e-06,
+      "loss": 0.0942,
+      "step": 47000
+    },
+    {
+      "epoch": 4.450861579246355,
+      "grad_norm": 0.4789665639400482,
+      "learning_rate": 2.196553683014581e-06,
+      "loss": 0.0916,
+      "step": 47010
+    },
+    {
+      "epoch": 4.451808369626964,
+      "grad_norm": 0.41863638162612915,
+      "learning_rate": 2.192766521492142e-06,
+      "loss": 0.0944,
+      "step": 47020
+    },
+    {
+      "epoch": 4.452755160007574,
+      "grad_norm": 0.6053740382194519,
+      "learning_rate": 2.1889793599697027e-06,
+      "loss": 0.0953,
+      "step": 47030
+    },
+    {
+      "epoch": 4.453701950388184,
+      "grad_norm": 0.613536536693573,
+      "learning_rate": 2.185192198447264e-06,
+      "loss": 0.0982,
+      "step": 47040
+    },
+    {
+      "epoch": 4.454648740768794,
+      "grad_norm": 0.5732889771461487,
+      "learning_rate": 2.181405036924825e-06,
+      "loss": 0.0998,
+      "step": 47050
+    },
+    {
+      "epoch": 4.455595531149403,
+      "grad_norm": 0.5707571506500244,
+      "learning_rate": 2.1776178754023862e-06,
+      "loss": 0.0998,
+      "step": 47060
+    },
+    {
+      "epoch": 4.456542321530013,
+      "grad_norm": 0.5420945882797241,
+      "learning_rate": 2.173830713879947e-06,
+      "loss": 0.1015,
+      "step": 47070
+    },
+    {
+      "epoch": 4.457489111910623,
+      "grad_norm": 0.49488115310668945,
+      "learning_rate": 2.1700435523575085e-06,
+      "loss": 0.0854,
+      "step": 47080
+    },
+    {
+      "epoch": 4.4584359022912325,
+      "grad_norm": 0.6199977993965149,
+      "learning_rate": 2.1662563908350694e-06,
+      "loss": 0.0994,
+      "step": 47090
+    },
+    {
+      "epoch": 4.459382692671842,
+      "grad_norm": 0.6206436157226562,
+      "learning_rate": 2.1624692293126302e-06,
+      "loss": 0.0893,
+      "step": 47100
+    },
+    {
+      "epoch": 4.460329483052452,
+      "grad_norm": 0.5212389826774597,
+      "learning_rate": 2.1586820677901916e-06,
+      "loss": 0.0995,
+      "step": 47110
+    },
+    {
+      "epoch": 4.461276273433062,
+      "grad_norm": 0.4194228947162628,
+      "learning_rate": 2.1548949062677525e-06,
+      "loss": 0.0898,
+      "step": 47120
+    },
+    {
+      "epoch": 4.4622230638136715,
+      "grad_norm": 0.6620252728462219,
+      "learning_rate": 2.1511077447453134e-06,
+      "loss": 0.0957,
+      "step": 47130
+    },
+    {
+      "epoch": 4.463169854194281,
+      "grad_norm": 0.6501525640487671,
+      "learning_rate": 2.1473205832228747e-06,
+      "loss": 0.0942,
+      "step": 47140
+    },
+    {
+      "epoch": 4.464116644574891,
+      "grad_norm": 0.46277403831481934,
+      "learning_rate": 2.143533421700436e-06,
+      "loss": 0.0901,
+      "step": 47150
+    },
+    {
+      "epoch": 4.465063434955501,
+      "grad_norm": 0.5377644896507263,
+      "learning_rate": 2.139746260177997e-06,
+      "loss": 0.0952,
+      "step": 47160
+    },
+    {
+      "epoch": 4.4660102253361105,
+      "grad_norm": 0.4474472403526306,
+      "learning_rate": 2.135959098655558e-06,
+      "loss": 0.085,
+      "step": 47170
+    },
+    {
+      "epoch": 4.46695701571672,
+      "grad_norm": 0.7432825565338135,
+      "learning_rate": 2.132171937133119e-06,
+      "loss": 0.0975,
+      "step": 47180
+    },
+    {
+      "epoch": 4.46790380609733,
+      "grad_norm": 0.6087955832481384,
+      "learning_rate": 2.12838477561068e-06,
+      "loss": 0.0924,
+      "step": 47190
+    },
+    {
+      "epoch": 4.46885059647794,
+      "grad_norm": 0.43486493825912476,
+      "learning_rate": 2.124597614088241e-06,
+      "loss": 0.0937,
+      "step": 47200
+    },
+    {
+      "epoch": 4.469797386858549,
+      "grad_norm": 0.597850501537323,
+      "learning_rate": 2.1208104525658022e-06,
+      "loss": 0.0919,
+      "step": 47210
+    },
+    {
+      "epoch": 4.470744177239159,
+      "grad_norm": 0.5512962937355042,
+      "learning_rate": 2.117023291043363e-06,
+      "loss": 0.0919,
+      "step": 47220
+    },
+    {
+      "epoch": 4.471690967619769,
+      "grad_norm": 0.6809868216514587,
+      "learning_rate": 2.113236129520924e-06,
+      "loss": 0.0883,
+      "step": 47230
+    },
+    {
+      "epoch": 4.472637758000379,
+      "grad_norm": 0.5824252963066101,
+      "learning_rate": 2.1094489679984853e-06,
+      "loss": 0.0948,
+      "step": 47240
+    },
+    {
+      "epoch": 4.473584548380988,
+      "grad_norm": 0.5019168853759766,
+      "learning_rate": 2.1056618064760467e-06,
+      "loss": 0.0888,
+      "step": 47250
+    },
+    {
+      "epoch": 4.474531338761598,
+      "grad_norm": 0.7237482666969299,
+      "learning_rate": 2.101874644953607e-06,
+      "loss": 0.0959,
+      "step": 47260
+    },
+    {
+      "epoch": 4.475478129142208,
+      "grad_norm": 0.5009791851043701,
+      "learning_rate": 2.0980874834311685e-06,
+      "loss": 0.0878,
+      "step": 47270
+    },
+    {
+      "epoch": 4.476424919522818,
+      "grad_norm": 0.6152024269104004,
+      "learning_rate": 2.0943003219087298e-06,
+      "loss": 0.1019,
+      "step": 47280
+    },
+    {
+      "epoch": 4.477371709903427,
+      "grad_norm": 0.589474618434906,
+      "learning_rate": 2.0905131603862907e-06,
+      "loss": 0.089,
+      "step": 47290
+    },
+    {
+      "epoch": 4.478318500284037,
+      "grad_norm": 0.5134758353233337,
+      "learning_rate": 2.0867259988638516e-06,
+      "loss": 0.0772,
+      "step": 47300
+    },
+    {
+      "epoch": 4.479265290664647,
+      "grad_norm": 0.5028401613235474,
+      "learning_rate": 2.082938837341413e-06,
+      "loss": 0.0933,
+      "step": 47310
+    },
+    {
+      "epoch": 4.480212081045257,
+      "grad_norm": 0.5507858395576477,
+      "learning_rate": 2.0791516758189738e-06,
+      "loss": 0.0952,
+      "step": 47320
+    },
+    {
+      "epoch": 4.481158871425866,
+      "grad_norm": 0.5241835713386536,
+      "learning_rate": 2.0753645142965347e-06,
+      "loss": 0.095,
+      "step": 47330
+    },
+    {
+      "epoch": 4.482105661806476,
+      "grad_norm": 0.4093499183654785,
+      "learning_rate": 2.071577352774096e-06,
+      "loss": 0.0999,
+      "step": 47340
+    },
+    {
+      "epoch": 4.483052452187086,
+      "grad_norm": 0.4314521253108978,
+      "learning_rate": 2.067790191251657e-06,
+      "loss": 0.0847,
+      "step": 47350
+    },
+    {
+      "epoch": 4.4839992425676956,
+      "grad_norm": 0.5057846903800964,
+      "learning_rate": 2.0640030297292182e-06,
+      "loss": 0.0932,
+      "step": 47360
+    },
+    {
+      "epoch": 4.484946032948305,
+      "grad_norm": 0.49404215812683105,
+      "learning_rate": 2.060215868206779e-06,
+      "loss": 0.0854,
+      "step": 47370
+    },
+    {
+      "epoch": 4.485892823328915,
+      "grad_norm": 0.46423012018203735,
+      "learning_rate": 2.0564287066843404e-06,
+      "loss": 0.0762,
+      "step": 47380
+    },
+    {
+      "epoch": 4.486839613709525,
+      "grad_norm": 0.568574070930481,
+      "learning_rate": 2.0526415451619013e-06,
+      "loss": 0.0926,
+      "step": 47390
+    },
+    {
+      "epoch": 4.4877864040901345,
+      "grad_norm": 0.45246657729148865,
+      "learning_rate": 2.0488543836394622e-06,
+      "loss": 0.0875,
+      "step": 47400
+    },
+    {
+      "epoch": 4.488733194470744,
+      "grad_norm": 0.6380802392959595,
+      "learning_rate": 2.0450672221170235e-06,
+      "loss": 0.0916,
+      "step": 47410
+    },
+    {
+      "epoch": 4.489679984851354,
+      "grad_norm": 0.6181575059890747,
+      "learning_rate": 2.0412800605945844e-06,
+      "loss": 0.1019,
+      "step": 47420
+    },
+    {
+      "epoch": 4.490626775231964,
+      "grad_norm": 0.47934210300445557,
+      "learning_rate": 2.0374928990721458e-06,
+      "loss": 0.0916,
+      "step": 47430
+    },
+    {
+      "epoch": 4.4915735656125735,
+      "grad_norm": 0.3730686902999878,
+      "learning_rate": 2.0337057375497067e-06,
+      "loss": 0.0926,
+      "step": 47440
+    },
+    {
+      "epoch": 4.492520355993183,
+      "grad_norm": 0.524203896522522,
+      "learning_rate": 2.0299185760272676e-06,
+      "loss": 0.0851,
+      "step": 47450
+    },
+    {
+      "epoch": 4.493467146373793,
+      "grad_norm": 0.602016806602478,
+      "learning_rate": 2.026131414504829e-06,
+      "loss": 0.1,
+      "step": 47460
+    },
+    {
+      "epoch": 4.494413936754403,
+      "grad_norm": 0.5126527547836304,
+      "learning_rate": 2.0223442529823898e-06,
+      "loss": 0.0947,
+      "step": 47470
+    },
+    {
+      "epoch": 4.4953607271350124,
+      "grad_norm": 0.5978550314903259,
+      "learning_rate": 2.0185570914599507e-06,
+      "loss": 0.0912,
+      "step": 47480
+    },
+    {
+      "epoch": 4.496307517515622,
+      "grad_norm": 0.6456809043884277,
+      "learning_rate": 2.014769929937512e-06,
+      "loss": 0.0906,
+      "step": 47490
+    },
+    {
+      "epoch": 4.497254307896232,
+      "grad_norm": 0.6243536472320557,
+      "learning_rate": 2.0109827684150733e-06,
+      "loss": 0.0892,
+      "step": 47500
+    },
+    {
+      "epoch": 4.498201098276842,
+      "grad_norm": 0.6239606738090515,
+      "learning_rate": 2.007195606892634e-06,
+      "loss": 0.1018,
+      "step": 47510
+    },
+    {
+      "epoch": 4.499147888657451,
+      "grad_norm": 0.5502533316612244,
+      "learning_rate": 2.003408445370195e-06,
+      "loss": 0.0876,
+      "step": 47520
+    },
+    {
+      "epoch": 4.500094679038061,
+      "grad_norm": 0.5233820676803589,
+      "learning_rate": 1.9996212838477564e-06,
+      "loss": 0.0894,
+      "step": 47530
+    },
+    {
+      "epoch": 4.501041469418671,
+      "grad_norm": 0.5940799713134766,
+      "learning_rate": 1.9958341223253173e-06,
+      "loss": 0.093,
+      "step": 47540
+    },
+    {
+      "epoch": 4.501988259799281,
+      "grad_norm": 0.4981343150138855,
+      "learning_rate": 1.9920469608028782e-06,
+      "loss": 0.0985,
+      "step": 47550
+    },
+    {
+      "epoch": 4.50293505017989,
+      "grad_norm": 0.5547967553138733,
+      "learning_rate": 1.9882597992804395e-06,
+      "loss": 0.0964,
+      "step": 47560
+    },
+    {
+      "epoch": 4.5038818405605,
+      "grad_norm": 0.5440186262130737,
+      "learning_rate": 1.9844726377580004e-06,
+      "loss": 0.0894,
+      "step": 47570
+    },
+    {
+      "epoch": 4.50482863094111,
+      "grad_norm": 0.4961601197719574,
+      "learning_rate": 1.9806854762355613e-06,
+      "loss": 0.0971,
+      "step": 47580
+    },
+    {
+      "epoch": 4.50577542132172,
+      "grad_norm": 0.533189594745636,
+      "learning_rate": 1.9768983147131226e-06,
+      "loss": 0.0958,
+      "step": 47590
+    },
+    {
+      "epoch": 4.506722211702329,
+      "grad_norm": 0.6105077862739563,
+      "learning_rate": 1.973111153190684e-06,
+      "loss": 0.1035,
+      "step": 47600
+    },
+    {
+      "epoch": 4.507669002082939,
+      "grad_norm": 0.48340150713920593,
+      "learning_rate": 1.969323991668245e-06,
+      "loss": 0.0964,
+      "step": 47610
+    },
+    {
+      "epoch": 4.508615792463549,
+      "grad_norm": 0.4797910451889038,
+      "learning_rate": 1.9655368301458058e-06,
+      "loss": 0.087,
+      "step": 47620
+    },
+    {
+      "epoch": 4.509562582844159,
+      "grad_norm": 0.4579121470451355,
+      "learning_rate": 1.961749668623367e-06,
+      "loss": 0.0753,
+      "step": 47630
+    },
+    {
+      "epoch": 4.510509373224768,
+      "grad_norm": 0.44602665305137634,
+      "learning_rate": 1.957962507100928e-06,
+      "loss": 0.0839,
+      "step": 47640
+    },
+    {
+      "epoch": 4.511456163605378,
+      "grad_norm": 0.5439406633377075,
+      "learning_rate": 1.954175345578489e-06,
+      "loss": 0.0939,
+      "step": 47650
+    },
+    {
+      "epoch": 4.512402953985988,
+      "grad_norm": 0.49075302481651306,
+      "learning_rate": 1.95038818405605e-06,
+      "loss": 0.0858,
+      "step": 47660
+    },
+    {
+      "epoch": 4.5133497443665975,
+      "grad_norm": 0.6823569536209106,
+      "learning_rate": 1.946601022533611e-06,
+      "loss": 0.0905,
+      "step": 47670
+    },
+    {
+      "epoch": 4.514296534747207,
+      "grad_norm": 0.6025480031967163,
+      "learning_rate": 1.9428138610111724e-06,
+      "loss": 0.0906,
+      "step": 47680
+    },
+    {
+      "epoch": 4.515243325127817,
+      "grad_norm": 0.5409242510795593,
+      "learning_rate": 1.9390266994887333e-06,
+      "loss": 0.0802,
+      "step": 47690
+    },
+    {
+      "epoch": 4.516190115508427,
+      "grad_norm": 0.7054596543312073,
+      "learning_rate": 1.9352395379662946e-06,
+      "loss": 0.0928,
+      "step": 47700
+    },
+    {
+      "epoch": 4.5171369058890365,
+      "grad_norm": 0.5168089270591736,
+      "learning_rate": 1.9314523764438555e-06,
+      "loss": 0.0991,
+      "step": 47710
+    },
+    {
+      "epoch": 4.518083696269646,
+      "grad_norm": 0.4419559836387634,
+      "learning_rate": 1.9276652149214164e-06,
+      "loss": 0.0989,
+      "step": 47720
+    },
+    {
+      "epoch": 4.519030486650256,
+      "grad_norm": 0.38130053877830505,
+      "learning_rate": 1.9238780533989777e-06,
+      "loss": 0.0909,
+      "step": 47730
+    },
+    {
+      "epoch": 4.519977277030866,
+      "grad_norm": 0.4823101758956909,
+      "learning_rate": 1.9200908918765386e-06,
+      "loss": 0.0903,
+      "step": 47740
+    },
+    {
+      "epoch": 4.5209240674114755,
+      "grad_norm": 0.47707557678222656,
+      "learning_rate": 1.9163037303541e-06,
+      "loss": 0.0872,
+      "step": 47750
+    },
+    {
+      "epoch": 4.521870857792085,
+      "grad_norm": 0.549338161945343,
+      "learning_rate": 1.912516568831661e-06,
+      "loss": 0.0893,
+      "step": 47760
+    },
+    {
+      "epoch": 4.522817648172695,
+      "grad_norm": 0.4239804148674011,
+      "learning_rate": 1.9087294073092218e-06,
+      "loss": 0.1028,
+      "step": 47770
+    },
+    {
+      "epoch": 4.523764438553305,
+      "grad_norm": 0.5911729335784912,
+      "learning_rate": 1.904942245786783e-06,
+      "loss": 0.0959,
+      "step": 47780
+    },
+    {
+      "epoch": 4.524711228933914,
+      "grad_norm": 0.635185956954956,
+      "learning_rate": 1.901155084264344e-06,
+      "loss": 0.0912,
+      "step": 47790
+    },
+    {
+      "epoch": 4.525658019314523,
+      "grad_norm": 0.43398451805114746,
+      "learning_rate": 1.897367922741905e-06,
+      "loss": 0.09,
+      "step": 47800
+    },
+    {
+      "epoch": 4.526604809695133,
+      "grad_norm": 0.4836176931858063,
+      "learning_rate": 1.8935807612194662e-06,
+      "loss": 0.0944,
+      "step": 47810
+    },
+    {
+      "epoch": 4.527551600075743,
+      "grad_norm": 0.5630194544792175,
+      "learning_rate": 1.8897935996970273e-06,
+      "loss": 0.0997,
+      "step": 47820
+    },
+    {
+      "epoch": 4.5284983904563525,
+      "grad_norm": 0.6335233449935913,
+      "learning_rate": 1.8860064381745882e-06,
+      "loss": 0.1002,
+      "step": 47830
+    },
+    {
+      "epoch": 4.529445180836962,
+      "grad_norm": 0.5927964448928833,
+      "learning_rate": 1.8822192766521493e-06,
+      "loss": 0.0922,
+      "step": 47840
+    },
+    {
+      "epoch": 4.530391971217572,
+      "grad_norm": 0.6638123393058777,
+      "learning_rate": 1.8784321151297106e-06,
+      "loss": 0.0957,
+      "step": 47850
+    },
+    {
+      "epoch": 4.531338761598182,
+      "grad_norm": 0.5143892765045166,
+      "learning_rate": 1.8746449536072713e-06,
+      "loss": 0.0997,
+      "step": 47860
+    },
+    {
+      "epoch": 4.5322855519787915,
+      "grad_norm": 0.6001013517379761,
+      "learning_rate": 1.8708577920848326e-06,
+      "loss": 0.0843,
+      "step": 47870
+    },
+    {
+      "epoch": 4.533232342359401,
+      "grad_norm": 0.5932102799415588,
+      "learning_rate": 1.8670706305623937e-06,
+      "loss": 0.0833,
+      "step": 47880
+    },
+    {
+      "epoch": 4.534179132740011,
+      "grad_norm": 0.5315927863121033,
+      "learning_rate": 1.8632834690399548e-06,
+      "loss": 0.0888,
+      "step": 47890
+    },
+    {
+      "epoch": 4.535125923120621,
+      "grad_norm": 0.47217097878456116,
+      "learning_rate": 1.8594963075175157e-06,
+      "loss": 0.0947,
+      "step": 47900
+    },
+    {
+      "epoch": 4.53607271350123,
+      "grad_norm": 0.554823100566864,
+      "learning_rate": 1.8557091459950768e-06,
+      "loss": 0.087,
+      "step": 47910
+    },
+    {
+      "epoch": 4.53701950388184,
+      "grad_norm": 0.4658668339252472,
+      "learning_rate": 1.851921984472638e-06,
+      "loss": 0.097,
+      "step": 47920
+    },
+    {
+      "epoch": 4.53796629426245,
+      "grad_norm": 0.6078159213066101,
+      "learning_rate": 1.8481348229501988e-06,
+      "loss": 0.0929,
+      "step": 47930
+    },
+    {
+      "epoch": 4.53891308464306,
+      "grad_norm": 0.42829376459121704,
+      "learning_rate": 1.84434766142776e-06,
+      "loss": 0.091,
+      "step": 47940
+    },
+    {
+      "epoch": 4.539859875023669,
+      "grad_norm": 0.5195884704589844,
+      "learning_rate": 1.840560499905321e-06,
+      "loss": 0.0881,
+      "step": 47950
+    },
+    {
+      "epoch": 4.540806665404279,
+      "grad_norm": 0.6427383422851562,
+      "learning_rate": 1.8367733383828824e-06,
+      "loss": 0.0967,
+      "step": 47960
+    },
+    {
+      "epoch": 4.541753455784889,
+      "grad_norm": 0.4192584753036499,
+      "learning_rate": 1.832986176860443e-06,
+      "loss": 0.0976,
+      "step": 47970
+    },
+    {
+      "epoch": 4.542700246165499,
+      "grad_norm": 0.5321178436279297,
+      "learning_rate": 1.8291990153380044e-06,
+      "loss": 0.0993,
+      "step": 47980
+    },
+    {
+      "epoch": 4.543647036546108,
+      "grad_norm": 0.5339673757553101,
+      "learning_rate": 1.8254118538155655e-06,
+      "loss": 0.0901,
+      "step": 47990
+    },
+    {
+      "epoch": 4.544593826926718,
+      "grad_norm": 0.5013030767440796,
+      "learning_rate": 1.8216246922931264e-06,
+      "loss": 0.0915,
+      "step": 48000
+    },
+    {
+      "epoch": 4.545540617307328,
+      "grad_norm": 0.5032514929771423,
+      "learning_rate": 1.8178375307706875e-06,
+      "loss": 0.0938,
+      "step": 48010
+    },
+    {
+      "epoch": 4.546487407687938,
+      "grad_norm": 0.48303863406181335,
+      "learning_rate": 1.8140503692482486e-06,
+      "loss": 0.0889,
+      "step": 48020
+    },
+    {
+      "epoch": 4.547434198068547,
+      "grad_norm": 0.5052141547203064,
+      "learning_rate": 1.8102632077258097e-06,
+      "loss": 0.0935,
+      "step": 48030
+    },
+    {
+      "epoch": 4.548380988449157,
+      "grad_norm": 0.5991350412368774,
+      "learning_rate": 1.8064760462033706e-06,
+      "loss": 0.0912,
+      "step": 48040
+    },
+    {
+      "epoch": 4.549327778829767,
+      "grad_norm": 0.5054553747177124,
+      "learning_rate": 1.8026888846809317e-06,
+      "loss": 0.0939,
+      "step": 48050
+    },
+    {
+      "epoch": 4.5502745692103765,
+      "grad_norm": 0.5401033759117126,
+      "learning_rate": 1.7989017231584928e-06,
+      "loss": 0.0977,
+      "step": 48060
+    },
+    {
+      "epoch": 4.551221359590986,
+      "grad_norm": 0.5015616416931152,
+      "learning_rate": 1.7951145616360537e-06,
+      "loss": 0.0882,
+      "step": 48070
+    },
+    {
+      "epoch": 4.552168149971596,
+      "grad_norm": 0.6362848877906799,
+      "learning_rate": 1.791327400113615e-06,
+      "loss": 0.095,
+      "step": 48080
+    },
+    {
+      "epoch": 4.553114940352206,
+      "grad_norm": 0.5170758962631226,
+      "learning_rate": 1.7875402385911762e-06,
+      "loss": 0.0863,
+      "step": 48090
+    },
+    {
+      "epoch": 4.5540617307328155,
+      "grad_norm": 0.5425009727478027,
+      "learning_rate": 1.7837530770687373e-06,
+      "loss": 0.0876,
+      "step": 48100
+    },
+    {
+      "epoch": 4.555008521113425,
+      "grad_norm": 0.5243383049964905,
+      "learning_rate": 1.7799659155462982e-06,
+      "loss": 0.0919,
+      "step": 48110
+    },
+    {
+      "epoch": 4.555955311494035,
+      "grad_norm": 0.8553285002708435,
+      "learning_rate": 1.7761787540238593e-06,
+      "loss": 0.0934,
+      "step": 48120
+    },
+    {
+      "epoch": 4.556902101874645,
+      "grad_norm": 0.548096239566803,
+      "learning_rate": 1.7723915925014204e-06,
+      "loss": 0.0834,
+      "step": 48130
+    },
+    {
+      "epoch": 4.5578488922552545,
+      "grad_norm": 0.7237502336502075,
+      "learning_rate": 1.7686044309789813e-06,
+      "loss": 0.0891,
+      "step": 48140
+    },
+    {
+      "epoch": 4.558795682635864,
+      "grad_norm": 0.5385095477104187,
+      "learning_rate": 1.7648172694565424e-06,
+      "loss": 0.0946,
+      "step": 48150
+    },
+    {
+      "epoch": 4.559742473016474,
+      "grad_norm": 0.577634871006012,
+      "learning_rate": 1.7610301079341035e-06,
+      "loss": 0.0991,
+      "step": 48160
+    },
+    {
+      "epoch": 4.560689263397084,
+      "grad_norm": 0.4976717531681061,
+      "learning_rate": 1.7572429464116648e-06,
+      "loss": 0.0842,
+      "step": 48170
+    },
+    {
+      "epoch": 4.561636053777693,
+      "grad_norm": 0.5380846261978149,
+      "learning_rate": 1.7534557848892255e-06,
+      "loss": 0.0935,
+      "step": 48180
+    },
+    {
+      "epoch": 4.562582844158303,
+      "grad_norm": 0.5093594789505005,
+      "learning_rate": 1.7496686233667868e-06,
+      "loss": 0.0862,
+      "step": 48190
+    },
+    {
+      "epoch": 4.563529634538913,
+      "grad_norm": 0.6310622096061707,
+      "learning_rate": 1.745881461844348e-06,
+      "loss": 0.1006,
+      "step": 48200
+    },
+    {
+      "epoch": 4.564476424919523,
+      "grad_norm": 0.47840484976768494,
+      "learning_rate": 1.7420943003219088e-06,
+      "loss": 0.0926,
+      "step": 48210
+    },
+    {
+      "epoch": 4.565423215300132,
+      "grad_norm": 0.5460345149040222,
+      "learning_rate": 1.73830713879947e-06,
+      "loss": 0.092,
+      "step": 48220
+    },
+    {
+      "epoch": 4.566370005680742,
+      "grad_norm": 0.46515849232673645,
+      "learning_rate": 1.734519977277031e-06,
+      "loss": 0.0905,
+      "step": 48230
+    },
+    {
+      "epoch": 4.567316796061352,
+      "grad_norm": 0.45555320382118225,
+      "learning_rate": 1.7307328157545921e-06,
+      "loss": 0.0841,
+      "step": 48240
+    },
+    {
+      "epoch": 4.568263586441962,
+      "grad_norm": 0.587450385093689,
+      "learning_rate": 1.726945654232153e-06,
+      "loss": 0.0927,
+      "step": 48250
+    },
+    {
+      "epoch": 4.569210376822571,
+      "grad_norm": 0.5282682776451111,
+      "learning_rate": 1.7231584927097142e-06,
+      "loss": 0.0789,
+      "step": 48260
+    },
+    {
+      "epoch": 4.570157167203181,
+      "grad_norm": 0.847499668598175,
+      "learning_rate": 1.7193713311872753e-06,
+      "loss": 0.0904,
+      "step": 48270
+    },
+    {
+      "epoch": 4.571103957583791,
+      "grad_norm": 0.5478895306587219,
+      "learning_rate": 1.7155841696648362e-06,
+      "loss": 0.0991,
+      "step": 48280
+    },
+    {
+      "epoch": 4.572050747964401,
+      "grad_norm": 0.49527424573898315,
+      "learning_rate": 1.7117970081423973e-06,
+      "loss": 0.089,
+      "step": 48290
+    },
+    {
+      "epoch": 4.57299753834501,
+      "grad_norm": 0.4628439247608185,
+      "learning_rate": 1.7080098466199586e-06,
+      "loss": 0.0934,
+      "step": 48300
+    },
+    {
+      "epoch": 4.57394432872562,
+      "grad_norm": 0.5064544677734375,
+      "learning_rate": 1.7042226850975197e-06,
+      "loss": 0.1002,
+      "step": 48310
+    },
+    {
+      "epoch": 4.57489111910623,
+      "grad_norm": 0.5350474715232849,
+      "learning_rate": 1.7004355235750806e-06,
+      "loss": 0.0985,
+      "step": 48320
+    },
+    {
+      "epoch": 4.5758379094868395,
+      "grad_norm": 0.45474421977996826,
+      "learning_rate": 1.6966483620526417e-06,
+      "loss": 0.0882,
+      "step": 48330
+    },
+    {
+      "epoch": 4.576784699867449,
+      "grad_norm": 0.480854332447052,
+      "learning_rate": 1.6928612005302028e-06,
+      "loss": 0.0882,
+      "step": 48340
+    },
+    {
+      "epoch": 4.577731490248059,
+      "grad_norm": 0.5241674780845642,
+      "learning_rate": 1.689074039007764e-06,
+      "loss": 0.0966,
+      "step": 48350
+    },
+    {
+      "epoch": 4.578678280628669,
+      "grad_norm": 0.5988481640815735,
+      "learning_rate": 1.6852868774853248e-06,
+      "loss": 0.0881,
+      "step": 48360
+    },
+    {
+      "epoch": 4.5796250710092785,
+      "grad_norm": 0.5592998266220093,
+      "learning_rate": 1.681499715962886e-06,
+      "loss": 0.084,
+      "step": 48370
+    },
+    {
+      "epoch": 4.580571861389888,
+      "grad_norm": 0.5882778167724609,
+      "learning_rate": 1.677712554440447e-06,
+      "loss": 0.0969,
+      "step": 48380
+    },
+    {
+      "epoch": 4.581518651770498,
+      "grad_norm": 0.5792142748832703,
+      "learning_rate": 1.673925392918008e-06,
+      "loss": 0.084,
+      "step": 48390
+    },
+    {
+      "epoch": 4.582465442151108,
+      "grad_norm": 0.586322009563446,
+      "learning_rate": 1.670138231395569e-06,
+      "loss": 0.0948,
+      "step": 48400
+    },
+    {
+      "epoch": 4.5834122325317175,
+      "grad_norm": 0.4606407880783081,
+      "learning_rate": 1.6663510698731304e-06,
+      "loss": 0.0815,
+      "step": 48410
+    },
+    {
+      "epoch": 4.584359022912327,
+      "grad_norm": 0.5143293142318726,
+      "learning_rate": 1.6625639083506915e-06,
+      "loss": 0.0904,
+      "step": 48420
+    },
+    {
+      "epoch": 4.585305813292937,
+      "grad_norm": 0.5697513818740845,
+      "learning_rate": 1.6587767468282524e-06,
+      "loss": 0.0917,
+      "step": 48430
+    },
+    {
+      "epoch": 4.586252603673547,
+      "grad_norm": 0.568105161190033,
+      "learning_rate": 1.6549895853058135e-06,
+      "loss": 0.0918,
+      "step": 48440
+    },
+    {
+      "epoch": 4.587199394054156,
+      "grad_norm": 0.5528663396835327,
+      "learning_rate": 1.6512024237833746e-06,
+      "loss": 0.094,
+      "step": 48450
+    },
+    {
+      "epoch": 4.588146184434766,
+      "grad_norm": 0.5852027535438538,
+      "learning_rate": 1.6474152622609355e-06,
+      "loss": 0.099,
+      "step": 48460
+    },
+    {
+      "epoch": 4.589092974815376,
+      "grad_norm": 0.7316393256187439,
+      "learning_rate": 1.6436281007384966e-06,
+      "loss": 0.099,
+      "step": 48470
+    },
+    {
+      "epoch": 4.590039765195986,
+      "grad_norm": 0.5696329474449158,
+      "learning_rate": 1.6398409392160577e-06,
+      "loss": 0.1,
+      "step": 48480
+    },
+    {
+      "epoch": 4.590986555576595,
+      "grad_norm": 0.4912358522415161,
+      "learning_rate": 1.6360537776936188e-06,
+      "loss": 0.0913,
+      "step": 48490
+    },
+    {
+      "epoch": 4.591933345957205,
+      "grad_norm": 0.6251103281974792,
+      "learning_rate": 1.6322666161711797e-06,
+      "loss": 0.0927,
+      "step": 48500
+    },
+    {
+      "epoch": 4.592880136337815,
+      "grad_norm": 0.6149888038635254,
+      "learning_rate": 1.628479454648741e-06,
+      "loss": 0.1018,
+      "step": 48510
+    },
+    {
+      "epoch": 4.593826926718425,
+      "grad_norm": 0.39401429891586304,
+      "learning_rate": 1.6246922931263021e-06,
+      "loss": 0.0951,
+      "step": 48520
+    },
+    {
+      "epoch": 4.594773717099034,
+      "grad_norm": 0.585252046585083,
+      "learning_rate": 1.620905131603863e-06,
+      "loss": 0.0879,
+      "step": 48530
+    },
+    {
+      "epoch": 4.595720507479644,
+      "grad_norm": 0.5147320628166199,
+      "learning_rate": 1.6171179700814241e-06,
+      "loss": 0.0849,
+      "step": 48540
+    },
+    {
+      "epoch": 4.596667297860254,
+      "grad_norm": 0.5566774010658264,
+      "learning_rate": 1.6133308085589852e-06,
+      "loss": 0.0954,
+      "step": 48550
+    },
+    {
+      "epoch": 4.597614088240864,
+      "grad_norm": 0.5427478551864624,
+      "learning_rate": 1.6095436470365463e-06,
+      "loss": 0.0856,
+      "step": 48560
+    },
+    {
+      "epoch": 4.598560878621473,
+      "grad_norm": 0.4353178143501282,
+      "learning_rate": 1.6057564855141072e-06,
+      "loss": 0.0823,
+      "step": 48570
+    },
+    {
+      "epoch": 4.599507669002083,
+      "grad_norm": 0.49626612663269043,
+      "learning_rate": 1.6019693239916683e-06,
+      "loss": 0.0825,
+      "step": 48580
+    },
+    {
+      "epoch": 4.600454459382693,
+      "grad_norm": 0.5396493077278137,
+      "learning_rate": 1.5981821624692295e-06,
+      "loss": 0.0931,
+      "step": 48590
+    },
+    {
+      "epoch": 4.601401249763303,
+      "grad_norm": 0.6535559892654419,
+      "learning_rate": 1.5943950009467904e-06,
+      "loss": 0.0897,
+      "step": 48600
+    },
+    {
+      "epoch": 4.602348040143912,
+      "grad_norm": 0.4790814518928528,
+      "learning_rate": 1.5906078394243515e-06,
+      "loss": 0.0877,
+      "step": 48610
+    },
+    {
+      "epoch": 4.603294830524522,
+      "grad_norm": 0.5527889132499695,
+      "learning_rate": 1.5868206779019128e-06,
+      "loss": 0.0879,
+      "step": 48620
+    },
+    {
+      "epoch": 4.604241620905132,
+      "grad_norm": 0.5145699977874756,
+      "learning_rate": 1.5830335163794739e-06,
+      "loss": 0.0951,
+      "step": 48630
+    },
+    {
+      "epoch": 4.6051884112857415,
+      "grad_norm": 0.5361632108688354,
+      "learning_rate": 1.5792463548570348e-06,
+      "loss": 0.0875,
+      "step": 48640
+    },
+    {
+      "epoch": 4.606135201666351,
+      "grad_norm": 0.4808364808559418,
+      "learning_rate": 1.5754591933345959e-06,
+      "loss": 0.0941,
+      "step": 48650
+    },
+    {
+      "epoch": 4.607081992046961,
+      "grad_norm": 0.46790459752082825,
+      "learning_rate": 1.571672031812157e-06,
+      "loss": 0.0998,
+      "step": 48660
+    },
+    {
+      "epoch": 4.608028782427571,
+      "grad_norm": 0.5993084907531738,
+      "learning_rate": 1.567884870289718e-06,
+      "loss": 0.0939,
+      "step": 48670
+    },
+    {
+      "epoch": 4.6089755728081805,
+      "grad_norm": 0.5489649772644043,
+      "learning_rate": 1.564097708767279e-06,
+      "loss": 0.0902,
+      "step": 48680
+    },
+    {
+      "epoch": 4.60992236318879,
+      "grad_norm": 0.4542575776576996,
+      "learning_rate": 1.5603105472448401e-06,
+      "loss": 0.0962,
+      "step": 48690
+    },
+    {
+      "epoch": 4.6108691535694,
+      "grad_norm": 0.5660474300384521,
+      "learning_rate": 1.5565233857224012e-06,
+      "loss": 0.1011,
+      "step": 48700
+    },
+    {
+      "epoch": 4.61181594395001,
+      "grad_norm": 0.36919817328453064,
+      "learning_rate": 1.5527362241999621e-06,
+      "loss": 0.0955,
+      "step": 48710
+    },
+    {
+      "epoch": 4.6127627343306195,
+      "grad_norm": 0.44812771677970886,
+      "learning_rate": 1.5489490626775232e-06,
+      "loss": 0.0887,
+      "step": 48720
+    },
+    {
+      "epoch": 4.613709524711229,
+      "grad_norm": 0.48133018612861633,
+      "learning_rate": 1.5451619011550845e-06,
+      "loss": 0.0901,
+      "step": 48730
+    },
+    {
+      "epoch": 4.614656315091839,
+      "grad_norm": 0.5825098156929016,
+      "learning_rate": 1.5413747396326452e-06,
+      "loss": 0.0916,
+      "step": 48740
+    },
+    {
+      "epoch": 4.615603105472449,
+      "grad_norm": 0.49545857310295105,
+      "learning_rate": 1.5375875781102066e-06,
+      "loss": 0.0859,
+      "step": 48750
+    },
+    {
+      "epoch": 4.616549895853058,
+      "grad_norm": 0.6294886469841003,
+      "learning_rate": 1.5338004165877677e-06,
+      "loss": 0.0948,
+      "step": 48760
+    },
+    {
+      "epoch": 4.617496686233668,
+      "grad_norm": 0.4956039488315582,
+      "learning_rate": 1.5300132550653288e-06,
+      "loss": 0.0992,
+      "step": 48770
+    },
+    {
+      "epoch": 4.618443476614278,
+      "grad_norm": 0.5199247598648071,
+      "learning_rate": 1.5262260935428897e-06,
+      "loss": 0.0933,
+      "step": 48780
+    },
+    {
+      "epoch": 4.619390266994888,
+      "grad_norm": 0.6329244375228882,
+      "learning_rate": 1.5224389320204508e-06,
+      "loss": 0.0943,
+      "step": 48790
+    },
+    {
+      "epoch": 4.620337057375497,
+      "grad_norm": 0.43671914935112,
+      "learning_rate": 1.5186517704980119e-06,
+      "loss": 0.0833,
+      "step": 48800
+    },
+    {
+      "epoch": 4.621283847756107,
+      "grad_norm": 0.7302520871162415,
+      "learning_rate": 1.5148646089755728e-06,
+      "loss": 0.0935,
+      "step": 48810
+    },
+    {
+      "epoch": 4.622230638136717,
+      "grad_norm": 0.5845633149147034,
+      "learning_rate": 1.5110774474531339e-06,
+      "loss": 0.0968,
+      "step": 48820
+    },
+    {
+      "epoch": 4.623177428517327,
+      "grad_norm": 0.4450472891330719,
+      "learning_rate": 1.507290285930695e-06,
+      "loss": 0.0898,
+      "step": 48830
+    },
+    {
+      "epoch": 4.624124218897936,
+      "grad_norm": 0.4358493685722351,
+      "learning_rate": 1.5035031244082563e-06,
+      "loss": 0.0884,
+      "step": 48840
+    },
+    {
+      "epoch": 4.625071009278546,
+      "grad_norm": 0.4955228865146637,
+      "learning_rate": 1.4997159628858172e-06,
+      "loss": 0.1026,
+      "step": 48850
+    },
+    {
+      "epoch": 4.626017799659156,
+      "grad_norm": 0.5216303467750549,
+      "learning_rate": 1.4959288013633783e-06,
+      "loss": 0.0972,
+      "step": 48860
+    },
+    {
+      "epoch": 4.626964590039766,
+      "grad_norm": 0.5317696928977966,
+      "learning_rate": 1.4921416398409394e-06,
+      "loss": 0.0997,
+      "step": 48870
+    },
+    {
+      "epoch": 4.627911380420375,
+      "grad_norm": 0.6008380055427551,
+      "learning_rate": 1.4883544783185003e-06,
+      "loss": 0.0917,
+      "step": 48880
+    },
+    {
+      "epoch": 4.628858170800985,
+      "grad_norm": 0.46178102493286133,
+      "learning_rate": 1.4845673167960614e-06,
+      "loss": 0.0947,
+      "step": 48890
+    },
+    {
+      "epoch": 4.629804961181595,
+      "grad_norm": 0.46813449263572693,
+      "learning_rate": 1.4807801552736225e-06,
+      "loss": 0.0991,
+      "step": 48900
+    },
+    {
+      "epoch": 4.6307517515622045,
+      "grad_norm": 0.5440060496330261,
+      "learning_rate": 1.4769929937511836e-06,
+      "loss": 0.0876,
+      "step": 48910
+    },
+    {
+      "epoch": 4.631698541942814,
+      "grad_norm": 0.6925702691078186,
+      "learning_rate": 1.4732058322287445e-06,
+      "loss": 0.0983,
+      "step": 48920
+    },
+    {
+      "epoch": 4.632645332323424,
+      "grad_norm": 0.5300967693328857,
+      "learning_rate": 1.4694186707063057e-06,
+      "loss": 0.0946,
+      "step": 48930
+    },
+    {
+      "epoch": 4.633592122704034,
+      "grad_norm": 0.6564103364944458,
+      "learning_rate": 1.465631509183867e-06,
+      "loss": 0.1005,
+      "step": 48940
+    },
+    {
+      "epoch": 4.6345389130846435,
+      "grad_norm": 0.7219815254211426,
+      "learning_rate": 1.4618443476614277e-06,
+      "loss": 0.0886,
+      "step": 48950
+    },
+    {
+      "epoch": 4.635485703465253,
+      "grad_norm": 0.486798495054245,
+      "learning_rate": 1.458057186138989e-06,
+      "loss": 0.0898,
+      "step": 48960
+    },
+    {
+      "epoch": 4.636432493845863,
+      "grad_norm": 0.5655710697174072,
+      "learning_rate": 1.45427002461655e-06,
+      "loss": 0.0954,
+      "step": 48970
+    },
+    {
+      "epoch": 4.637379284226473,
+      "grad_norm": 0.5758100152015686,
+      "learning_rate": 1.4504828630941112e-06,
+      "loss": 0.0938,
+      "step": 48980
+    },
+    {
+      "epoch": 4.6383260746070825,
+      "grad_norm": 0.4616601765155792,
+      "learning_rate": 1.446695701571672e-06,
+      "loss": 0.0915,
+      "step": 48990
+    },
+    {
+      "epoch": 4.639272864987692,
+      "grad_norm": 0.6310985088348389,
+      "learning_rate": 1.4429085400492332e-06,
+      "loss": 0.0851,
+      "step": 49000
+    },
+    {
+      "epoch": 4.640219655368302,
+      "grad_norm": 0.46795547008514404,
+      "learning_rate": 1.4391213785267943e-06,
+      "loss": 0.0965,
+      "step": 49010
+    },
+    {
+      "epoch": 4.641166445748912,
+      "grad_norm": 0.546395480632782,
+      "learning_rate": 1.4353342170043554e-06,
+      "loss": 0.096,
+      "step": 49020
+    },
+    {
+      "epoch": 4.6421132361295205,
+      "grad_norm": 0.5430060029029846,
+      "learning_rate": 1.4315470554819163e-06,
+      "loss": 0.0937,
+      "step": 49030
+    },
+    {
+      "epoch": 4.64306002651013,
+      "grad_norm": 0.5058007836341858,
+      "learning_rate": 1.4277598939594774e-06,
+      "loss": 0.0872,
+      "step": 49040
+    },
+    {
+      "epoch": 4.64400681689074,
+      "grad_norm": 0.7336437702178955,
+      "learning_rate": 1.4239727324370387e-06,
+      "loss": 0.0842,
+      "step": 49050
+    },
+    {
+      "epoch": 4.64495360727135,
+      "grad_norm": 0.4892963469028473,
+      "learning_rate": 1.4201855709145994e-06,
+      "loss": 0.0875,
+      "step": 49060
+    },
+    {
+      "epoch": 4.6459003976519595,
+      "grad_norm": 0.4091499447822571,
+      "learning_rate": 1.4163984093921607e-06,
+      "loss": 0.1047,
+      "step": 49070
+    },
+    {
+      "epoch": 4.646847188032569,
+      "grad_norm": 0.6091591715812683,
+      "learning_rate": 1.4126112478697219e-06,
+      "loss": 0.097,
+      "step": 49080
+    },
+    {
+      "epoch": 4.647793978413179,
+      "grad_norm": 0.5314242839813232,
+      "learning_rate": 1.408824086347283e-06,
+      "loss": 0.0954,
+      "step": 49090
+    },
+    {
+      "epoch": 4.648740768793789,
+      "grad_norm": 0.569369912147522,
+      "learning_rate": 1.4050369248248439e-06,
+      "loss": 0.1066,
+      "step": 49100
+    },
+    {
+      "epoch": 4.6496875591743985,
+      "grad_norm": 0.5993393659591675,
+      "learning_rate": 1.401249763302405e-06,
+      "loss": 0.0892,
+      "step": 49110
+    },
+    {
+      "epoch": 4.650634349555008,
+      "grad_norm": 0.5521275401115417,
+      "learning_rate": 1.397462601779966e-06,
+      "loss": 0.0922,
+      "step": 49120
+    },
+    {
+      "epoch": 4.651581139935618,
+      "grad_norm": 0.3962774872779846,
+      "learning_rate": 1.393675440257527e-06,
+      "loss": 0.0928,
+      "step": 49130
+    },
+    {
+      "epoch": 4.652527930316228,
+      "grad_norm": 0.5043448805809021,
+      "learning_rate": 1.389888278735088e-06,
+      "loss": 0.1001,
+      "step": 49140
+    },
+    {
+      "epoch": 4.653474720696837,
+      "grad_norm": 0.5763162970542908,
+      "learning_rate": 1.3861011172126492e-06,
+      "loss": 0.0874,
+      "step": 49150
+    },
+    {
+      "epoch": 4.654421511077447,
+      "grad_norm": 0.46237945556640625,
+      "learning_rate": 1.3823139556902105e-06,
+      "loss": 0.0857,
+      "step": 49160
+    },
+    {
+      "epoch": 4.655368301458057,
+      "grad_norm": 0.5338969826698303,
+      "learning_rate": 1.3785267941677712e-06,
+      "loss": 0.0896,
+      "step": 49170
+    },
+    {
+      "epoch": 4.656315091838667,
+      "grad_norm": 0.4410754144191742,
+      "learning_rate": 1.3747396326453325e-06,
+      "loss": 0.0927,
+      "step": 49180
+    },
+    {
+      "epoch": 4.657261882219276,
+      "grad_norm": 0.5380188822746277,
+      "learning_rate": 1.3709524711228936e-06,
+      "loss": 0.0926,
+      "step": 49190
+    },
+    {
+      "epoch": 4.658208672599886,
+      "grad_norm": 0.5101576447486877,
+      "learning_rate": 1.3671653096004545e-06,
+      "loss": 0.0946,
+      "step": 49200
+    },
+    {
+      "epoch": 4.659155462980496,
+      "grad_norm": 0.49067002534866333,
+      "learning_rate": 1.3633781480780156e-06,
+      "loss": 0.0882,
+      "step": 49210
+    },
+    {
+      "epoch": 4.660102253361106,
+      "grad_norm": 0.5032888054847717,
+      "learning_rate": 1.3595909865555767e-06,
+      "loss": 0.0909,
+      "step": 49220
+    },
+    {
+      "epoch": 4.661049043741715,
+      "grad_norm": 0.4651202857494354,
+      "learning_rate": 1.3558038250331378e-06,
+      "loss": 0.0969,
+      "step": 49230
+    },
+    {
+      "epoch": 4.661995834122325,
+      "grad_norm": 0.42336854338645935,
+      "learning_rate": 1.3520166635106987e-06,
+      "loss": 0.0834,
+      "step": 49240
+    },
+    {
+      "epoch": 4.662942624502935,
+      "grad_norm": 0.5473812222480774,
+      "learning_rate": 1.3482295019882598e-06,
+      "loss": 0.0901,
+      "step": 49250
+    },
+    {
+      "epoch": 4.663889414883545,
+      "grad_norm": 0.5984389781951904,
+      "learning_rate": 1.344442340465821e-06,
+      "loss": 0.0965,
+      "step": 49260
+    },
+    {
+      "epoch": 4.664836205264154,
+      "grad_norm": 0.5616798400878906,
+      "learning_rate": 1.3406551789433819e-06,
+      "loss": 0.0972,
+      "step": 49270
+    },
+    {
+      "epoch": 4.665782995644764,
+      "grad_norm": 0.5246201753616333,
+      "learning_rate": 1.3368680174209432e-06,
+      "loss": 0.0831,
+      "step": 49280
+    },
+    {
+      "epoch": 4.666729786025374,
+      "grad_norm": 0.5149614214897156,
+      "learning_rate": 1.3330808558985043e-06,
+      "loss": 0.0932,
+      "step": 49290
+    },
+    {
+      "epoch": 4.6676765764059835,
+      "grad_norm": 0.5594366192817688,
+      "learning_rate": 1.3292936943760654e-06,
+      "loss": 0.0884,
+      "step": 49300
+    },
+    {
+      "epoch": 4.668623366786593,
+      "grad_norm": 0.4831557869911194,
+      "learning_rate": 1.3255065328536263e-06,
+      "loss": 0.0873,
+      "step": 49310
+    },
+    {
+      "epoch": 4.669570157167203,
+      "grad_norm": 0.47912707924842834,
+      "learning_rate": 1.3217193713311874e-06,
+      "loss": 0.0916,
+      "step": 49320
+    },
+    {
+      "epoch": 4.670516947547813,
+      "grad_norm": 0.525671124458313,
+      "learning_rate": 1.3179322098087485e-06,
+      "loss": 0.0933,
+      "step": 49330
+    },
+    {
+      "epoch": 4.6714637379284225,
+      "grad_norm": 0.4452461898326874,
+      "learning_rate": 1.3141450482863094e-06,
+      "loss": 0.09,
+      "step": 49340
+    },
+    {
+      "epoch": 4.672410528309032,
+      "grad_norm": 0.5709280371665955,
+      "learning_rate": 1.3103578867638705e-06,
+      "loss": 0.0957,
+      "step": 49350
+    },
+    {
+      "epoch": 4.673357318689642,
+      "grad_norm": 0.48085644841194153,
+      "learning_rate": 1.3065707252414316e-06,
+      "loss": 0.0907,
+      "step": 49360
+    },
+    {
+      "epoch": 4.674304109070252,
+      "grad_norm": 0.5009046196937561,
+      "learning_rate": 1.302783563718993e-06,
+      "loss": 0.094,
+      "step": 49370
+    },
+    {
+      "epoch": 4.6752508994508615,
+      "grad_norm": 0.5080695152282715,
+      "learning_rate": 1.2989964021965536e-06,
+      "loss": 0.0946,
+      "step": 49380
+    },
+    {
+      "epoch": 4.676197689831471,
+      "grad_norm": 0.5521093010902405,
+      "learning_rate": 1.295209240674115e-06,
+      "loss": 0.0906,
+      "step": 49390
+    },
+    {
+      "epoch": 4.677144480212081,
+      "grad_norm": 0.5247788429260254,
+      "learning_rate": 1.291422079151676e-06,
+      "loss": 0.0859,
+      "step": 49400
+    },
+    {
+      "epoch": 4.678091270592691,
+      "grad_norm": 0.6452503204345703,
+      "learning_rate": 1.287634917629237e-06,
+      "loss": 0.1062,
+      "step": 49410
+    },
+    {
+      "epoch": 4.6790380609733,
+      "grad_norm": 0.5267934799194336,
+      "learning_rate": 1.283847756106798e-06,
+      "loss": 0.0984,
+      "step": 49420
+    },
+    {
+      "epoch": 4.67998485135391,
+      "grad_norm": 0.5547208786010742,
+      "learning_rate": 1.2800605945843592e-06,
+      "loss": 0.0953,
+      "step": 49430
+    },
+    {
+      "epoch": 4.68093164173452,
+      "grad_norm": 0.6340035796165466,
+      "learning_rate": 1.2762734330619203e-06,
+      "loss": 0.088,
+      "step": 49440
+    },
+    {
+      "epoch": 4.68187843211513,
+      "grad_norm": 0.4992550313472748,
+      "learning_rate": 1.2724862715394812e-06,
+      "loss": 0.0956,
+      "step": 49450
+    },
+    {
+      "epoch": 4.682825222495739,
+      "grad_norm": 0.5360420346260071,
+      "learning_rate": 1.2686991100170423e-06,
+      "loss": 0.0925,
+      "step": 49460
+    },
+    {
+      "epoch": 4.683772012876349,
+      "grad_norm": 0.5174246430397034,
+      "learning_rate": 1.2649119484946034e-06,
+      "loss": 0.0936,
+      "step": 49470
+    },
+    {
+      "epoch": 4.684718803256959,
+      "grad_norm": 0.514578104019165,
+      "learning_rate": 1.2611247869721643e-06,
+      "loss": 0.0926,
+      "step": 49480
+    },
+    {
+      "epoch": 4.685665593637569,
+      "grad_norm": 0.41404736042022705,
+      "learning_rate": 1.2573376254497254e-06,
+      "loss": 0.0923,
+      "step": 49490
+    },
+    {
+      "epoch": 4.686612384018178,
+      "grad_norm": 0.648402214050293,
+      "learning_rate": 1.2535504639272867e-06,
+      "loss": 0.0917,
+      "step": 49500
+    },
+    {
+      "epoch": 4.687559174398788,
+      "grad_norm": 0.45198550820350647,
+      "learning_rate": 1.2497633024048476e-06,
+      "loss": 0.0917,
+      "step": 49510
+    },
+    {
+      "epoch": 4.688505964779398,
+      "grad_norm": 0.5977778434753418,
+      "learning_rate": 1.2459761408824087e-06,
+      "loss": 0.0857,
+      "step": 49520
+    },
+    {
+      "epoch": 4.689452755160008,
+      "grad_norm": 0.5270425081253052,
+      "learning_rate": 1.2421889793599698e-06,
+      "loss": 0.0884,
+      "step": 49530
+    },
+    {
+      "epoch": 4.690399545540617,
+      "grad_norm": 0.5529866218566895,
+      "learning_rate": 1.238401817837531e-06,
+      "loss": 0.0892,
+      "step": 49540
+    },
+    {
+      "epoch": 4.691346335921227,
+      "grad_norm": 0.5758625268936157,
+      "learning_rate": 1.234614656315092e-06,
+      "loss": 0.0875,
+      "step": 49550
+    },
+    {
+      "epoch": 4.692293126301837,
+      "grad_norm": 0.40848255157470703,
+      "learning_rate": 1.230827494792653e-06,
+      "loss": 0.0908,
+      "step": 49560
+    },
+    {
+      "epoch": 4.6932399166824466,
+      "grad_norm": 0.7464273571968079,
+      "learning_rate": 1.227040333270214e-06,
+      "loss": 0.0883,
+      "step": 49570
+    },
+    {
+      "epoch": 4.694186707063056,
+      "grad_norm": 0.5910485982894897,
+      "learning_rate": 1.2232531717477752e-06,
+      "loss": 0.1056,
+      "step": 49580
+    },
+    {
+      "epoch": 4.695133497443666,
+      "grad_norm": 0.6291195750236511,
+      "learning_rate": 1.2194660102253363e-06,
+      "loss": 0.0964,
+      "step": 49590
+    },
+    {
+      "epoch": 4.696080287824276,
+      "grad_norm": 0.5667415857315063,
+      "learning_rate": 1.2156788487028974e-06,
+      "loss": 0.0881,
+      "step": 49600
+    },
+    {
+      "epoch": 4.6970270782048855,
+      "grad_norm": 0.5887658596038818,
+      "learning_rate": 1.2118916871804585e-06,
+      "loss": 0.1005,
+      "step": 49610
+    },
+    {
+      "epoch": 4.697973868585495,
+      "grad_norm": 0.6646347045898438,
+      "learning_rate": 1.2081045256580194e-06,
+      "loss": 0.1049,
+      "step": 49620
+    },
+    {
+      "epoch": 4.698920658966105,
+      "grad_norm": 0.6309397220611572,
+      "learning_rate": 1.2043173641355805e-06,
+      "loss": 0.0966,
+      "step": 49630
+    },
+    {
+      "epoch": 4.699867449346715,
+      "grad_norm": 0.47189533710479736,
+      "learning_rate": 1.2005302026131416e-06,
+      "loss": 0.0959,
+      "step": 49640
+    },
+    {
+      "epoch": 4.7008142397273245,
+      "grad_norm": 0.550778329372406,
+      "learning_rate": 1.1967430410907025e-06,
+      "loss": 0.0953,
+      "step": 49650
+    },
+    {
+      "epoch": 4.701761030107934,
+      "grad_norm": 0.5566471219062805,
+      "learning_rate": 1.1929558795682638e-06,
+      "loss": 0.0952,
+      "step": 49660
+    },
+    {
+      "epoch": 4.702707820488544,
+      "grad_norm": 0.4658328592777252,
+      "learning_rate": 1.1891687180458247e-06,
+      "loss": 0.0936,
+      "step": 49670
+    },
+    {
+      "epoch": 4.703654610869154,
+      "grad_norm": 0.49258533120155334,
+      "learning_rate": 1.1853815565233858e-06,
+      "loss": 0.0836,
+      "step": 49680
+    },
+    {
+      "epoch": 4.7046014012497634,
+      "grad_norm": 0.4751582741737366,
+      "learning_rate": 1.181594395000947e-06,
+      "loss": 0.0952,
+      "step": 49690
+    },
+    {
+      "epoch": 4.705548191630373,
+      "grad_norm": 0.3998904824256897,
+      "learning_rate": 1.1778072334785078e-06,
+      "loss": 0.0842,
+      "step": 49700
+    },
+    {
+      "epoch": 4.706494982010983,
+      "grad_norm": 0.48084378242492676,
+      "learning_rate": 1.1740200719560691e-06,
+      "loss": 0.0868,
+      "step": 49710
+    },
+    {
+      "epoch": 4.707441772391593,
+      "grad_norm": 0.6053601503372192,
+      "learning_rate": 1.17023291043363e-06,
+      "loss": 0.093,
+      "step": 49720
+    },
+    {
+      "epoch": 4.708388562772202,
+      "grad_norm": 0.6152405142784119,
+      "learning_rate": 1.1664457489111911e-06,
+      "loss": 0.0939,
+      "step": 49730
+    },
+    {
+      "epoch": 4.709335353152812,
+      "grad_norm": 0.5622721910476685,
+      "learning_rate": 1.1626585873887522e-06,
+      "loss": 0.0994,
+      "step": 49740
+    },
+    {
+      "epoch": 4.710282143533422,
+      "grad_norm": 0.4358634650707245,
+      "learning_rate": 1.1588714258663134e-06,
+      "loss": 0.091,
+      "step": 49750
+    },
+    {
+      "epoch": 4.711228933914032,
+      "grad_norm": 0.6170639395713806,
+      "learning_rate": 1.1550842643438743e-06,
+      "loss": 0.0965,
+      "step": 49760
+    },
+    {
+      "epoch": 4.712175724294641,
+      "grad_norm": 0.49188241362571716,
+      "learning_rate": 1.1512971028214354e-06,
+      "loss": 0.0889,
+      "step": 49770
+    },
+    {
+      "epoch": 4.713122514675251,
+      "grad_norm": 0.5877458453178406,
+      "learning_rate": 1.1475099412989965e-06,
+      "loss": 0.0862,
+      "step": 49780
+    },
+    {
+      "epoch": 4.714069305055861,
+      "grad_norm": 0.6240220665931702,
+      "learning_rate": 1.1437227797765576e-06,
+      "loss": 0.1044,
+      "step": 49790
+    },
+    {
+      "epoch": 4.715016095436471,
+      "grad_norm": 0.5297105312347412,
+      "learning_rate": 1.1399356182541187e-06,
+      "loss": 0.0975,
+      "step": 49800
+    },
+    {
+      "epoch": 4.71596288581708,
+      "grad_norm": 0.5716924071311951,
+      "learning_rate": 1.1361484567316796e-06,
+      "loss": 0.0828,
+      "step": 49810
+    },
+    {
+      "epoch": 4.71690967619769,
+      "grad_norm": 0.516631543636322,
+      "learning_rate": 1.132361295209241e-06,
+      "loss": 0.1032,
+      "step": 49820
+    },
+    {
+      "epoch": 4.7178564665783,
+      "grad_norm": 0.5588654279708862,
+      "learning_rate": 1.1285741336868018e-06,
+      "loss": 0.0918,
+      "step": 49830
+    },
+    {
+      "epoch": 4.71880325695891,
+      "grad_norm": 0.6855382919311523,
+      "learning_rate": 1.124786972164363e-06,
+      "loss": 0.0926,
+      "step": 49840
+    },
+    {
+      "epoch": 4.719750047339519,
+      "grad_norm": 0.5521231293678284,
+      "learning_rate": 1.120999810641924e-06,
+      "loss": 0.0953,
+      "step": 49850
+    },
+    {
+      "epoch": 4.720696837720129,
+      "grad_norm": 0.4538123905658722,
+      "learning_rate": 1.117212649119485e-06,
+      "loss": 0.086,
+      "step": 49860
+    },
+    {
+      "epoch": 4.721643628100739,
+      "grad_norm": 0.6084619760513306,
+      "learning_rate": 1.1134254875970462e-06,
+      "loss": 0.0997,
+      "step": 49870
+    },
+    {
+      "epoch": 4.7225904184813485,
+      "grad_norm": 0.5663161277770996,
+      "learning_rate": 1.1096383260746071e-06,
+      "loss": 0.0975,
+      "step": 49880
+    },
+    {
+      "epoch": 4.723537208861958,
+      "grad_norm": 0.6240411996841431,
+      "learning_rate": 1.1058511645521682e-06,
+      "loss": 0.0902,
+      "step": 49890
+    },
+    {
+      "epoch": 4.724483999242568,
+      "grad_norm": 0.5772308111190796,
+      "learning_rate": 1.1020640030297293e-06,
+      "loss": 0.0963,
+      "step": 49900
+    },
+    {
+      "epoch": 4.725430789623178,
+      "grad_norm": 0.5128609538078308,
+      "learning_rate": 1.0982768415072905e-06,
+      "loss": 0.093,
+      "step": 49910
+    },
+    {
+      "epoch": 4.7263775800037875,
+      "grad_norm": 0.5880887508392334,
+      "learning_rate": 1.0944896799848514e-06,
+      "loss": 0.099,
+      "step": 49920
+    },
+    {
+      "epoch": 4.727324370384397,
+      "grad_norm": 0.5767456293106079,
+      "learning_rate": 1.0907025184624125e-06,
+      "loss": 0.0891,
+      "step": 49930
+    },
+    {
+      "epoch": 4.728271160765007,
+      "grad_norm": 0.49970582127571106,
+      "learning_rate": 1.0869153569399736e-06,
+      "loss": 0.1076,
+      "step": 49940
+    },
+    {
+      "epoch": 4.729217951145617,
+      "grad_norm": 0.5847047567367554,
+      "learning_rate": 1.0831281954175347e-06,
+      "loss": 0.0907,
+      "step": 49950
+    },
+    {
+      "epoch": 4.7301647415262265,
+      "grad_norm": 0.5303926467895508,
+      "learning_rate": 1.0793410338950958e-06,
+      "loss": 0.0815,
+      "step": 49960
+    },
+    {
+      "epoch": 4.731111531906836,
+      "grad_norm": 0.5159422159194946,
+      "learning_rate": 1.0755538723726567e-06,
+      "loss": 0.0913,
+      "step": 49970
+    },
+    {
+      "epoch": 4.732058322287446,
+      "grad_norm": 0.5303027033805847,
+      "learning_rate": 1.071766710850218e-06,
+      "loss": 0.1028,
+      "step": 49980
+    },
+    {
+      "epoch": 4.733005112668056,
+      "grad_norm": 0.5290799140930176,
+      "learning_rate": 1.067979549327779e-06,
+      "loss": 0.0884,
+      "step": 49990
+    },
+    {
+      "epoch": 4.7339519030486645,
+      "grad_norm": 0.5512375831604004,
+      "learning_rate": 1.06419238780534e-06,
+      "loss": 0.091,
+      "step": 50000
+    },
+    {
+      "epoch": 4.734898693429274,
+      "grad_norm": 0.5875411629676819,
+      "learning_rate": 1.0604052262829011e-06,
+      "loss": 0.0945,
+      "step": 50010
+    },
+    {
+      "epoch": 4.735845483809884,
+      "grad_norm": 0.4429018199443817,
+      "learning_rate": 1.056618064760462e-06,
+      "loss": 0.0963,
+      "step": 50020
+    },
+    {
+      "epoch": 4.736792274190494,
+      "grad_norm": 0.46107998490333557,
+      "learning_rate": 1.0528309032380233e-06,
+      "loss": 0.0905,
+      "step": 50030
+    },
+    {
+      "epoch": 4.7377390645711035,
+      "grad_norm": 0.5040816068649292,
+      "learning_rate": 1.0490437417155842e-06,
+      "loss": 0.0918,
+      "step": 50040
+    },
+    {
+      "epoch": 4.738685854951713,
+      "grad_norm": 0.7133190631866455,
+      "learning_rate": 1.0452565801931453e-06,
+      "loss": 0.0888,
+      "step": 50050
+    },
+    {
+      "epoch": 4.739632645332323,
+      "grad_norm": 0.5230016112327576,
+      "learning_rate": 1.0414694186707064e-06,
+      "loss": 0.0957,
+      "step": 50060
+    },
+    {
+      "epoch": 4.740579435712933,
+      "grad_norm": 0.5962569117546082,
+      "learning_rate": 1.0376822571482673e-06,
+      "loss": 0.09,
+      "step": 50070
+    },
+    {
+      "epoch": 4.7415262260935425,
+      "grad_norm": 0.6349098086357117,
+      "learning_rate": 1.0338950956258284e-06,
+      "loss": 0.0887,
+      "step": 50080
+    },
+    {
+      "epoch": 4.742473016474152,
+      "grad_norm": 0.5445008277893066,
+      "learning_rate": 1.0301079341033896e-06,
+      "loss": 0.0943,
+      "step": 50090
+    },
+    {
+      "epoch": 4.743419806854762,
+      "grad_norm": 0.6481450200080872,
+      "learning_rate": 1.0263207725809507e-06,
+      "loss": 0.0951,
+      "step": 50100
+    },
+    {
+      "epoch": 4.744366597235372,
+      "grad_norm": 0.44210201501846313,
+      "learning_rate": 1.0225336110585118e-06,
+      "loss": 0.0906,
+      "step": 50110
+    },
+    {
+      "epoch": 4.745313387615981,
+      "grad_norm": 0.5137997269630432,
+      "learning_rate": 1.0187464495360729e-06,
+      "loss": 0.0924,
+      "step": 50120
+    },
+    {
+      "epoch": 4.746260177996591,
+      "grad_norm": 0.6586678624153137,
+      "learning_rate": 1.0149592880136338e-06,
+      "loss": 0.0893,
+      "step": 50130
+    },
+    {
+      "epoch": 4.747206968377201,
+      "grad_norm": 0.46986591815948486,
+      "learning_rate": 1.0111721264911949e-06,
+      "loss": 0.0896,
+      "step": 50140
+    },
+    {
+      "epoch": 4.748153758757811,
+      "grad_norm": 0.5114980936050415,
+      "learning_rate": 1.007384964968756e-06,
+      "loss": 0.0951,
+      "step": 50150
+    },
+    {
+      "epoch": 4.74910054913842,
+      "grad_norm": 0.6802167296409607,
+      "learning_rate": 1.003597803446317e-06,
+      "loss": 0.0996,
+      "step": 50160
+    },
+    {
+      "epoch": 4.75004733951903,
+      "grad_norm": 0.46343374252319336,
+      "learning_rate": 9.998106419238782e-07,
+      "loss": 0.0969,
+      "step": 50170
+    },
+    {
+      "epoch": 4.75099412989964,
+      "grad_norm": 0.5714529752731323,
+      "learning_rate": 9.960234804014391e-07,
+      "loss": 0.0931,
+      "step": 50180
+    },
+    {
+      "epoch": 4.75194092028025,
+      "grad_norm": 0.6029381155967712,
+      "learning_rate": 9.922363188790002e-07,
+      "loss": 0.0856,
+      "step": 50190
+    },
+    {
+      "epoch": 4.752887710660859,
+      "grad_norm": 0.5086356997489929,
+      "learning_rate": 9.884491573565613e-07,
+      "loss": 0.0866,
+      "step": 50200
+    },
+    {
+      "epoch": 4.753834501041469,
+      "grad_norm": 0.6164889931678772,
+      "learning_rate": 9.846619958341224e-07,
+      "loss": 0.0884,
+      "step": 50210
+    },
+    {
+      "epoch": 4.754781291422079,
+      "grad_norm": 0.5729707479476929,
+      "learning_rate": 9.808748343116835e-07,
+      "loss": 0.09,
+      "step": 50220
+    },
+    {
+      "epoch": 4.755728081802689,
+      "grad_norm": 0.3943701982498169,
+      "learning_rate": 9.770876727892444e-07,
+      "loss": 0.0936,
+      "step": 50230
+    },
+    {
+      "epoch": 4.756674872183298,
+      "grad_norm": 0.4985075891017914,
+      "learning_rate": 9.733005112668055e-07,
+      "loss": 0.0933,
+      "step": 50240
+    },
+    {
+      "epoch": 4.757621662563908,
+      "grad_norm": 0.4944251775741577,
+      "learning_rate": 9.695133497443667e-07,
+      "loss": 0.0931,
+      "step": 50250
+    },
+    {
+      "epoch": 4.758568452944518,
+      "grad_norm": 0.6448076963424683,
+      "learning_rate": 9.657261882219278e-07,
+      "loss": 0.095,
+      "step": 50260
+    },
+    {
+      "epoch": 4.7595152433251275,
+      "grad_norm": 0.4403682053089142,
+      "learning_rate": 9.619390266994889e-07,
+      "loss": 0.0919,
+      "step": 50270
+    },
+    {
+      "epoch": 4.760462033705737,
+      "grad_norm": 0.49853387475013733,
+      "learning_rate": 9.5815186517705e-07,
+      "loss": 0.0899,
+      "step": 50280
+    },
+    {
+      "epoch": 4.761408824086347,
+      "grad_norm": 0.5590994358062744,
+      "learning_rate": 9.543647036546109e-07,
+      "loss": 0.0904,
+      "step": 50290
+    },
+    {
+      "epoch": 4.762355614466957,
+      "grad_norm": 0.555829644203186,
+      "learning_rate": 9.50577542132172e-07,
+      "loss": 0.0903,
+      "step": 50300
+    },
+    {
+      "epoch": 4.7633024048475665,
+      "grad_norm": 0.5076011419296265,
+      "learning_rate": 9.467903806097331e-07,
+      "loss": 0.0881,
+      "step": 50310
+    },
+    {
+      "epoch": 4.764249195228176,
+      "grad_norm": 0.5209220051765442,
+      "learning_rate": 9.430032190872941e-07,
+      "loss": 0.0877,
+      "step": 50320
+    },
+    {
+      "epoch": 4.765195985608786,
+      "grad_norm": 0.591728687286377,
+      "learning_rate": 9.392160575648553e-07,
+      "loss": 0.0913,
+      "step": 50330
+    },
+    {
+      "epoch": 4.766142775989396,
+      "grad_norm": 0.6746892333030701,
+      "learning_rate": 9.354288960424163e-07,
+      "loss": 0.1078,
+      "step": 50340
+    },
+    {
+      "epoch": 4.7670895663700055,
+      "grad_norm": 0.5808528065681458,
+      "learning_rate": 9.316417345199774e-07,
+      "loss": 0.0884,
+      "step": 50350
+    },
+    {
+      "epoch": 4.768036356750615,
+      "grad_norm": 0.49978432059288025,
+      "learning_rate": 9.278545729975384e-07,
+      "loss": 0.096,
+      "step": 50360
+    },
+    {
+      "epoch": 4.768983147131225,
+      "grad_norm": 0.5775094628334045,
+      "learning_rate": 9.240674114750994e-07,
+      "loss": 0.083,
+      "step": 50370
+    },
+    {
+      "epoch": 4.769929937511835,
+      "grad_norm": 0.5577254891395569,
+      "learning_rate": 9.202802499526605e-07,
+      "loss": 0.0953,
+      "step": 50380
+    },
+    {
+      "epoch": 4.770876727892444,
+      "grad_norm": 0.45562031865119934,
+      "learning_rate": 9.164930884302215e-07,
+      "loss": 0.0793,
+      "step": 50390
+    },
+    {
+      "epoch": 4.771823518273054,
+      "grad_norm": 0.594912588596344,
+      "learning_rate": 9.127059269077827e-07,
+      "loss": 0.0974,
+      "step": 50400
+    },
+    {
+      "epoch": 4.772770308653664,
+      "grad_norm": 0.5201529860496521,
+      "learning_rate": 9.089187653853438e-07,
+      "loss": 0.082,
+      "step": 50410
+    },
+    {
+      "epoch": 4.773717099034274,
+      "grad_norm": 0.5733709335327148,
+      "learning_rate": 9.051316038629049e-07,
+      "loss": 0.085,
+      "step": 50420
+    },
+    {
+      "epoch": 4.774663889414883,
+      "grad_norm": 0.5458856225013733,
+      "learning_rate": 9.013444423404659e-07,
+      "loss": 0.0922,
+      "step": 50430
+    },
+    {
+      "epoch": 4.775610679795493,
+      "grad_norm": 0.6832992434501648,
+      "learning_rate": 8.975572808180269e-07,
+      "loss": 0.098,
+      "step": 50440
+    },
+    {
+      "epoch": 4.776557470176103,
+      "grad_norm": 0.5935986638069153,
+      "learning_rate": 8.937701192955881e-07,
+      "loss": 0.0934,
+      "step": 50450
+    },
+    {
+      "epoch": 4.777504260556713,
+      "grad_norm": 0.4659850001335144,
+      "learning_rate": 8.899829577731491e-07,
+      "loss": 0.0904,
+      "step": 50460
+    },
+    {
+      "epoch": 4.778451050937322,
+      "grad_norm": 0.5135877132415771,
+      "learning_rate": 8.861957962507102e-07,
+      "loss": 0.0974,
+      "step": 50470
+    },
+    {
+      "epoch": 4.779397841317932,
+      "grad_norm": 0.5025619268417358,
+      "learning_rate": 8.824086347282712e-07,
+      "loss": 0.0916,
+      "step": 50480
+    },
+    {
+      "epoch": 4.780344631698542,
+      "grad_norm": 0.5468993186950684,
+      "learning_rate": 8.786214732058324e-07,
+      "loss": 0.0944,
+      "step": 50490
+    },
+    {
+      "epoch": 4.781291422079152,
+      "grad_norm": 0.46831241250038147,
+      "learning_rate": 8.748343116833934e-07,
+      "loss": 0.0931,
+      "step": 50500
+    },
+    {
+      "epoch": 4.782238212459761,
+      "grad_norm": 0.5284664630889893,
+      "learning_rate": 8.710471501609544e-07,
+      "loss": 0.0966,
+      "step": 50510
+    },
+    {
+      "epoch": 4.783185002840371,
+      "grad_norm": 0.5607685446739197,
+      "learning_rate": 8.672599886385155e-07,
+      "loss": 0.1012,
+      "step": 50520
+    },
+    {
+      "epoch": 4.784131793220981,
+      "grad_norm": 0.6038400530815125,
+      "learning_rate": 8.634728271160765e-07,
+      "loss": 0.0884,
+      "step": 50530
+    },
+    {
+      "epoch": 4.7850785836015906,
+      "grad_norm": 0.5385847091674805,
+      "learning_rate": 8.596856655936376e-07,
+      "loss": 0.1006,
+      "step": 50540
+    },
+    {
+      "epoch": 4.7860253739822,
+      "grad_norm": 0.5663634538650513,
+      "learning_rate": 8.558985040711986e-07,
+      "loss": 0.0893,
+      "step": 50550
+    },
+    {
+      "epoch": 4.78697216436281,
+      "grad_norm": 0.5027875900268555,
+      "learning_rate": 8.521113425487598e-07,
+      "loss": 0.0979,
+      "step": 50560
+    },
+    {
+      "epoch": 4.78791895474342,
+      "grad_norm": 0.5991628170013428,
+      "learning_rate": 8.483241810263208e-07,
+      "loss": 0.0864,
+      "step": 50570
+    },
+    {
+      "epoch": 4.7888657451240295,
+      "grad_norm": 0.5896027684211731,
+      "learning_rate": 8.44537019503882e-07,
+      "loss": 0.0851,
+      "step": 50580
+    },
+    {
+      "epoch": 4.789812535504639,
+      "grad_norm": 0.5217905640602112,
+      "learning_rate": 8.40749857981443e-07,
+      "loss": 0.0859,
+      "step": 50590
+    },
+    {
+      "epoch": 4.790759325885249,
+      "grad_norm": 0.5823928117752075,
+      "learning_rate": 8.36962696459004e-07,
+      "loss": 0.1002,
+      "step": 50600
+    },
+    {
+      "epoch": 4.791706116265859,
+      "grad_norm": 0.49396812915802,
+      "learning_rate": 8.331755349365652e-07,
+      "loss": 0.0977,
+      "step": 50610
+    },
+    {
+      "epoch": 4.7926529066464685,
+      "grad_norm": 0.6325224041938782,
+      "learning_rate": 8.293883734141262e-07,
+      "loss": 0.0994,
+      "step": 50620
+    },
+    {
+      "epoch": 4.793599697027078,
+      "grad_norm": 0.3896678388118744,
+      "learning_rate": 8.256012118916873e-07,
+      "loss": 0.0823,
+      "step": 50630
+    },
+    {
+      "epoch": 4.794546487407688,
+      "grad_norm": 0.5307965278625488,
+      "learning_rate": 8.218140503692483e-07,
+      "loss": 0.0837,
+      "step": 50640
+    },
+    {
+      "epoch": 4.795493277788298,
+      "grad_norm": 0.47504711151123047,
+      "learning_rate": 8.180268888468094e-07,
+      "loss": 0.0868,
+      "step": 50650
+    },
+    {
+      "epoch": 4.7964400681689074,
+      "grad_norm": 0.432176411151886,
+      "learning_rate": 8.142397273243705e-07,
+      "loss": 0.0866,
+      "step": 50660
+    },
+    {
+      "epoch": 4.797386858549517,
+      "grad_norm": 0.5039948225021362,
+      "learning_rate": 8.104525658019315e-07,
+      "loss": 0.1057,
+      "step": 50670
+    },
+    {
+      "epoch": 4.798333648930127,
+      "grad_norm": 0.6106250882148743,
+      "learning_rate": 8.066654042794926e-07,
+      "loss": 0.0948,
+      "step": 50680
+    },
+    {
+      "epoch": 4.799280439310737,
+      "grad_norm": 0.6046525835990906,
+      "learning_rate": 8.028782427570536e-07,
+      "loss": 0.0845,
+      "step": 50690
+    },
+    {
+      "epoch": 4.800227229691346,
+      "grad_norm": 0.6091911792755127,
+      "learning_rate": 7.990910812346147e-07,
+      "loss": 0.0823,
+      "step": 50700
+    },
+    {
+      "epoch": 4.801174020071956,
+      "grad_norm": 0.4663586914539337,
+      "learning_rate": 7.953039197121757e-07,
+      "loss": 0.0902,
+      "step": 50710
+    },
+    {
+      "epoch": 4.802120810452566,
+      "grad_norm": 0.6265435814857483,
+      "learning_rate": 7.915167581897369e-07,
+      "loss": 0.1029,
+      "step": 50720
+    },
+    {
+      "epoch": 4.803067600833176,
+      "grad_norm": 0.5090877413749695,
+      "learning_rate": 7.877295966672979e-07,
+      "loss": 0.0911,
+      "step": 50730
+    },
+    {
+      "epoch": 4.804014391213785,
+      "grad_norm": 0.7234902381896973,
+      "learning_rate": 7.83942435144859e-07,
+      "loss": 0.0988,
+      "step": 50740
+    },
+    {
+      "epoch": 4.804961181594395,
+      "grad_norm": 0.46384572982788086,
+      "learning_rate": 7.801552736224201e-07,
+      "loss": 0.0962,
+      "step": 50750
+    },
+    {
+      "epoch": 4.805907971975005,
+      "grad_norm": 0.6380556225776672,
+      "learning_rate": 7.763681120999811e-07,
+      "loss": 0.0934,
+      "step": 50760
+    },
+    {
+      "epoch": 4.806854762355615,
+      "grad_norm": 0.6317399144172668,
+      "learning_rate": 7.725809505775423e-07,
+      "loss": 0.0901,
+      "step": 50770
+    },
+    {
+      "epoch": 4.807801552736224,
+      "grad_norm": 0.6209624409675598,
+      "learning_rate": 7.687937890551033e-07,
+      "loss": 0.0879,
+      "step": 50780
+    },
+    {
+      "epoch": 4.808748343116834,
+      "grad_norm": 0.45979273319244385,
+      "learning_rate": 7.650066275326644e-07,
+      "loss": 0.0969,
+      "step": 50790
+    },
+    {
+      "epoch": 4.809695133497444,
+      "grad_norm": 0.5645127892494202,
+      "learning_rate": 7.612194660102254e-07,
+      "loss": 0.0925,
+      "step": 50800
+    },
+    {
+      "epoch": 4.810641923878054,
+      "grad_norm": 0.5529855489730835,
+      "learning_rate": 7.574323044877864e-07,
+      "loss": 0.0911,
+      "step": 50810
+    },
+    {
+      "epoch": 4.811588714258663,
+      "grad_norm": 0.6504641175270081,
+      "learning_rate": 7.536451429653475e-07,
+      "loss": 0.0852,
+      "step": 50820
+    },
+    {
+      "epoch": 4.812535504639273,
+      "grad_norm": 0.6094434261322021,
+      "learning_rate": 7.498579814429086e-07,
+      "loss": 0.0921,
+      "step": 50830
+    },
+    {
+      "epoch": 4.813482295019883,
+      "grad_norm": 0.6175083518028259,
+      "learning_rate": 7.460708199204697e-07,
+      "loss": 0.0906,
+      "step": 50840
+    },
+    {
+      "epoch": 4.8144290854004925,
+      "grad_norm": 0.4623327851295471,
+      "learning_rate": 7.422836583980307e-07,
+      "loss": 0.0888,
+      "step": 50850
+    },
+    {
+      "epoch": 4.815375875781102,
+      "grad_norm": 0.47847726941108704,
+      "learning_rate": 7.384964968755918e-07,
+      "loss": 0.0926,
+      "step": 50860
+    },
+    {
+      "epoch": 4.816322666161712,
+      "grad_norm": 0.49766966700553894,
+      "learning_rate": 7.347093353531528e-07,
+      "loss": 0.0964,
+      "step": 50870
+    },
+    {
+      "epoch": 4.817269456542322,
+      "grad_norm": 0.6062334179878235,
+      "learning_rate": 7.309221738307138e-07,
+      "loss": 0.0953,
+      "step": 50880
+    },
+    {
+      "epoch": 4.8182162469229315,
+      "grad_norm": 0.5141642093658447,
+      "learning_rate": 7.27135012308275e-07,
+      "loss": 0.095,
+      "step": 50890
+    },
+    {
+      "epoch": 4.819163037303541,
+      "grad_norm": 0.5506232380867004,
+      "learning_rate": 7.23347850785836e-07,
+      "loss": 0.0992,
+      "step": 50900
+    },
+    {
+      "epoch": 4.820109827684151,
+      "grad_norm": 0.4917549788951874,
+      "learning_rate": 7.195606892633972e-07,
+      "loss": 0.0851,
+      "step": 50910
+    },
+    {
+      "epoch": 4.821056618064761,
+      "grad_norm": 0.5677406787872314,
+      "learning_rate": 7.157735277409582e-07,
+      "loss": 0.0859,
+      "step": 50920
+    },
+    {
+      "epoch": 4.8220034084453705,
+      "grad_norm": 0.45178139209747314,
+      "learning_rate": 7.119863662185194e-07,
+      "loss": 0.097,
+      "step": 50930
+    },
+    {
+      "epoch": 4.82295019882598,
+      "grad_norm": 0.5761244297027588,
+      "learning_rate": 7.081992046960804e-07,
+      "loss": 0.0885,
+      "step": 50940
+    },
+    {
+      "epoch": 4.82389698920659,
+      "grad_norm": 0.4566514790058136,
+      "learning_rate": 7.044120431736415e-07,
+      "loss": 0.1038,
+      "step": 50950
+    },
+    {
+      "epoch": 4.8248437795872,
+      "grad_norm": 0.4556839168071747,
+      "learning_rate": 7.006248816512025e-07,
+      "loss": 0.085,
+      "step": 50960
+    },
+    {
+      "epoch": 4.825790569967809,
+      "grad_norm": 0.45772460103034973,
+      "learning_rate": 6.968377201287635e-07,
+      "loss": 0.0882,
+      "step": 50970
+    },
+    {
+      "epoch": 4.826737360348419,
+      "grad_norm": 0.7039437890052795,
+      "learning_rate": 6.930505586063246e-07,
+      "loss": 0.09,
+      "step": 50980
+    },
+    {
+      "epoch": 4.827684150729029,
+      "grad_norm": 0.44567543268203735,
+      "learning_rate": 6.892633970838856e-07,
+      "loss": 0.0843,
+      "step": 50990
+    },
+    {
+      "epoch": 4.828630941109639,
+      "grad_norm": 0.6256598830223083,
+      "learning_rate": 6.854762355614468e-07,
+      "loss": 0.0886,
+      "step": 51000
+    },
+    {
+      "epoch": 4.829577731490248,
+      "grad_norm": 0.6110970377922058,
+      "learning_rate": 6.816890740390078e-07,
+      "loss": 0.0879,
+      "step": 51010
+    },
+    {
+      "epoch": 4.830524521870858,
+      "grad_norm": 0.594294548034668,
+      "learning_rate": 6.779019125165689e-07,
+      "loss": 0.0857,
+      "step": 51020
+    },
+    {
+      "epoch": 4.831471312251468,
+      "grad_norm": 0.5996149778366089,
+      "learning_rate": 6.741147509941299e-07,
+      "loss": 0.0977,
+      "step": 51030
+    },
+    {
+      "epoch": 4.832418102632078,
+      "grad_norm": 0.46598851680755615,
+      "learning_rate": 6.703275894716909e-07,
+      "loss": 0.0779,
+      "step": 51040
+    },
+    {
+      "epoch": 4.833364893012687,
+      "grad_norm": 0.5769855380058289,
+      "learning_rate": 6.665404279492521e-07,
+      "loss": 0.0891,
+      "step": 51050
+    },
+    {
+      "epoch": 4.834311683393297,
+      "grad_norm": 0.4835968017578125,
+      "learning_rate": 6.627532664268131e-07,
+      "loss": 0.1008,
+      "step": 51060
+    },
+    {
+      "epoch": 4.835258473773907,
+      "grad_norm": 0.5602681040763855,
+      "learning_rate": 6.589661049043743e-07,
+      "loss": 0.0915,
+      "step": 51070
+    },
+    {
+      "epoch": 4.836205264154517,
+      "grad_norm": 0.47379618883132935,
+      "learning_rate": 6.551789433819353e-07,
+      "loss": 0.0871,
+      "step": 51080
+    },
+    {
+      "epoch": 4.837152054535126,
+      "grad_norm": 0.6135598421096802,
+      "learning_rate": 6.513917818594965e-07,
+      "loss": 0.0923,
+      "step": 51090
+    },
+    {
+      "epoch": 4.838098844915736,
+      "grad_norm": 0.7021900415420532,
+      "learning_rate": 6.476046203370575e-07,
+      "loss": 0.0874,
+      "step": 51100
+    },
+    {
+      "epoch": 4.839045635296346,
+      "grad_norm": 0.561215341091156,
+      "learning_rate": 6.438174588146185e-07,
+      "loss": 0.0867,
+      "step": 51110
+    },
+    {
+      "epoch": 4.8399924256769555,
+      "grad_norm": 0.49121299386024475,
+      "learning_rate": 6.400302972921796e-07,
+      "loss": 0.0854,
+      "step": 51120
+    },
+    {
+      "epoch": 4.840939216057565,
+      "grad_norm": 0.5570336580276489,
+      "learning_rate": 6.362431357697406e-07,
+      "loss": 0.1007,
+      "step": 51130
+    },
+    {
+      "epoch": 4.841886006438175,
+      "grad_norm": 0.8793237805366516,
+      "learning_rate": 6.324559742473017e-07,
+      "loss": 0.0888,
+      "step": 51140
+    },
+    {
+      "epoch": 4.842832796818785,
+      "grad_norm": 0.47588610649108887,
+      "learning_rate": 6.286688127248627e-07,
+      "loss": 0.0858,
+      "step": 51150
+    },
+    {
+      "epoch": 4.8437795871993945,
+      "grad_norm": 0.44155871868133545,
+      "learning_rate": 6.248816512024238e-07,
+      "loss": 0.084,
+      "step": 51160
+    },
+    {
+      "epoch": 4.844726377580004,
+      "grad_norm": 0.5633000731468201,
+      "learning_rate": 6.210944896799849e-07,
+      "loss": 0.1027,
+      "step": 51170
+    },
+    {
+      "epoch": 4.845673167960614,
+      "grad_norm": 0.44512370228767395,
+      "learning_rate": 6.17307328157546e-07,
+      "loss": 0.0938,
+      "step": 51180
+    },
+    {
+      "epoch": 4.846619958341224,
+      "grad_norm": 0.5156192183494568,
+      "learning_rate": 6.13520166635107e-07,
+      "loss": 0.0919,
+      "step": 51190
+    },
+    {
+      "epoch": 4.8475667487218335,
+      "grad_norm": 0.6136853098869324,
+      "learning_rate": 6.097330051126681e-07,
+      "loss": 0.0896,
+      "step": 51200
+    },
+    {
+      "epoch": 4.848513539102443,
+      "grad_norm": 0.6080178618431091,
+      "learning_rate": 6.059458435902292e-07,
+      "loss": 0.0972,
+      "step": 51210
+    },
+    {
+      "epoch": 4.849460329483053,
+      "grad_norm": 0.5267219543457031,
+      "learning_rate": 6.021586820677902e-07,
+      "loss": 0.0873,
+      "step": 51220
+    },
+    {
+      "epoch": 4.850407119863662,
+      "grad_norm": 0.5620561838150024,
+      "learning_rate": 5.983715205453512e-07,
+      "loss": 0.0993,
+      "step": 51230
+    },
+    {
+      "epoch": 4.8513539102442715,
+      "grad_norm": 0.5927213430404663,
+      "learning_rate": 5.945843590229124e-07,
+      "loss": 0.0996,
+      "step": 51240
+    },
+    {
+      "epoch": 4.852300700624881,
+      "grad_norm": 0.5121906399726868,
+      "learning_rate": 5.907971975004735e-07,
+      "loss": 0.0854,
+      "step": 51250
+    },
+    {
+      "epoch": 4.853247491005491,
+      "grad_norm": 0.546197772026062,
+      "learning_rate": 5.870100359780346e-07,
+      "loss": 0.0863,
+      "step": 51260
+    },
+    {
+      "epoch": 4.854194281386101,
+      "grad_norm": 0.5386756658554077,
+      "learning_rate": 5.832228744555956e-07,
+      "loss": 0.0932,
+      "step": 51270
+    },
+    {
+      "epoch": 4.8551410717667105,
+      "grad_norm": 0.5520429611206055,
+      "learning_rate": 5.794357129331567e-07,
+      "loss": 0.098,
+      "step": 51280
+    },
+    {
+      "epoch": 4.85608786214732,
+      "grad_norm": 0.5074911713600159,
+      "learning_rate": 5.756485514107177e-07,
+      "loss": 0.0906,
+      "step": 51290
+    },
+    {
+      "epoch": 4.85703465252793,
+      "grad_norm": 0.6046466827392578,
+      "learning_rate": 5.718613898882788e-07,
+      "loss": 0.0918,
+      "step": 51300
+    },
+    {
+      "epoch": 4.85798144290854,
+      "grad_norm": 0.5096051096916199,
+      "learning_rate": 5.680742283658398e-07,
+      "loss": 0.0943,
+      "step": 51310
+    },
+    {
+      "epoch": 4.8589282332891495,
+      "grad_norm": 0.5715464353561401,
+      "learning_rate": 5.642870668434009e-07,
+      "loss": 0.093,
+      "step": 51320
+    },
+    {
+      "epoch": 4.859875023669759,
+      "grad_norm": 0.540324330329895,
+      "learning_rate": 5.60499905320962e-07,
+      "loss": 0.0941,
+      "step": 51330
+    },
+    {
+      "epoch": 4.860821814050369,
+      "grad_norm": 0.5595934391021729,
+      "learning_rate": 5.567127437985231e-07,
+      "loss": 0.0846,
+      "step": 51340
+    },
+    {
+      "epoch": 4.861768604430979,
+      "grad_norm": 0.5944637060165405,
+      "learning_rate": 5.529255822760841e-07,
+      "loss": 0.0862,
+      "step": 51350
+    },
+    {
+      "epoch": 4.862715394811588,
+      "grad_norm": 0.5972811579704285,
+      "learning_rate": 5.491384207536452e-07,
+      "loss": 0.0895,
+      "step": 51360
+    },
+    {
+      "epoch": 4.863662185192198,
+      "grad_norm": 0.6000321507453918,
+      "learning_rate": 5.453512592312062e-07,
+      "loss": 0.0938,
+      "step": 51370
+    },
+    {
+      "epoch": 4.864608975572808,
+      "grad_norm": 0.5122269988059998,
+      "learning_rate": 5.415640977087673e-07,
+      "loss": 0.0908,
+      "step": 51380
+    },
+    {
+      "epoch": 4.865555765953418,
+      "grad_norm": 0.534418523311615,
+      "learning_rate": 5.377769361863283e-07,
+      "loss": 0.0884,
+      "step": 51390
+    },
+    {
+      "epoch": 4.866502556334027,
+      "grad_norm": 0.5945922136306763,
+      "learning_rate": 5.339897746638894e-07,
+      "loss": 0.0935,
+      "step": 51400
+    },
+    {
+      "epoch": 4.867449346714637,
+      "grad_norm": 0.6264474391937256,
+      "learning_rate": 5.302026131414506e-07,
+      "loss": 0.0939,
+      "step": 51410
+    },
+    {
+      "epoch": 4.868396137095247,
+      "grad_norm": 0.5093408823013306,
+      "learning_rate": 5.264154516190117e-07,
+      "loss": 0.083,
+      "step": 51420
+    },
+    {
+      "epoch": 4.869342927475857,
+      "grad_norm": 0.48322850465774536,
+      "learning_rate": 5.226282900965727e-07,
+      "loss": 0.0948,
+      "step": 51430
+    },
+    {
+      "epoch": 4.870289717856466,
+      "grad_norm": 0.5509017109870911,
+      "learning_rate": 5.188411285741337e-07,
+      "loss": 0.0864,
+      "step": 51440
+    },
+    {
+      "epoch": 4.871236508237076,
+      "grad_norm": 0.6064262986183167,
+      "learning_rate": 5.150539670516948e-07,
+      "loss": 0.0866,
+      "step": 51450
+    },
+    {
+      "epoch": 4.872183298617686,
+      "grad_norm": 0.5785761475563049,
+      "learning_rate": 5.112668055292559e-07,
+      "loss": 0.0941,
+      "step": 51460
+    },
+    {
+      "epoch": 4.873130088998296,
+      "grad_norm": 0.6652457118034363,
+      "learning_rate": 5.074796440068169e-07,
+      "loss": 0.0828,
+      "step": 51470
+    },
+    {
+      "epoch": 4.874076879378905,
+      "grad_norm": 0.5285108089447021,
+      "learning_rate": 5.03692482484378e-07,
+      "loss": 0.0979,
+      "step": 51480
+    },
+    {
+      "epoch": 4.875023669759515,
+      "grad_norm": 0.4987425208091736,
+      "learning_rate": 4.999053209619391e-07,
+      "loss": 0.1033,
+      "step": 51490
+    },
+    {
+      "epoch": 4.875970460140125,
+      "grad_norm": 0.6044142842292786,
+      "learning_rate": 4.961181594395001e-07,
+      "loss": 0.0953,
+      "step": 51500
+    },
+    {
+      "epoch": 4.8769172505207345,
+      "grad_norm": 0.5120283961296082,
+      "learning_rate": 4.923309979170612e-07,
+      "loss": 0.0954,
+      "step": 51510
+    },
+    {
+      "epoch": 4.877864040901344,
+      "grad_norm": 0.4522251784801483,
+      "learning_rate": 4.885438363946222e-07,
+      "loss": 0.0936,
+      "step": 51520
+    },
+    {
+      "epoch": 4.878810831281954,
+      "grad_norm": 0.5160796642303467,
+      "learning_rate": 4.847566748721833e-07,
+      "loss": 0.0902,
+      "step": 51530
+    },
+    {
+      "epoch": 4.879757621662564,
+      "grad_norm": 0.4830280840396881,
+      "learning_rate": 4.809695133497444e-07,
+      "loss": 0.0887,
+      "step": 51540
+    },
+    {
+      "epoch": 4.8807044120431735,
+      "grad_norm": 0.5954660177230835,
+      "learning_rate": 4.771823518273054e-07,
+      "loss": 0.0994,
+      "step": 51550
+    },
+    {
+      "epoch": 4.881651202423783,
+      "grad_norm": 0.5464210510253906,
+      "learning_rate": 4.7339519030486655e-07,
+      "loss": 0.0893,
+      "step": 51560
+    },
+    {
+      "epoch": 4.882597992804393,
+      "grad_norm": 0.7138330936431885,
+      "learning_rate": 4.6960802878242765e-07,
+      "loss": 0.0908,
+      "step": 51570
+    },
+    {
+      "epoch": 4.883544783185003,
+      "grad_norm": 0.5895309448242188,
+      "learning_rate": 4.658208672599887e-07,
+      "loss": 0.0942,
+      "step": 51580
+    },
+    {
+      "epoch": 4.8844915735656125,
+      "grad_norm": 0.6211789846420288,
+      "learning_rate": 4.620337057375497e-07,
+      "loss": 0.0981,
+      "step": 51590
+    },
+    {
+      "epoch": 4.885438363946222,
+      "grad_norm": 0.4954429864883423,
+      "learning_rate": 4.5824654421511077e-07,
+      "loss": 0.1031,
+      "step": 51600
+    },
+    {
+      "epoch": 4.886385154326832,
+      "grad_norm": 0.550929605960846,
+      "learning_rate": 4.544593826926719e-07,
+      "loss": 0.0878,
+      "step": 51610
+    },
+    {
+      "epoch": 4.887331944707442,
+      "grad_norm": 0.5120697617530823,
+      "learning_rate": 4.5067222117023293e-07,
+      "loss": 0.0861,
+      "step": 51620
+    },
+    {
+      "epoch": 4.888278735088051,
+      "grad_norm": 0.5993196368217468,
+      "learning_rate": 4.4688505964779404e-07,
+      "loss": 0.0898,
+      "step": 51630
+    },
+    {
+      "epoch": 4.889225525468661,
+      "grad_norm": 0.627377450466156,
+      "learning_rate": 4.430978981253551e-07,
+      "loss": 0.0916,
+      "step": 51640
+    },
+    {
+      "epoch": 4.890172315849271,
+      "grad_norm": 0.4231056272983551,
+      "learning_rate": 4.393107366029162e-07,
+      "loss": 0.0837,
+      "step": 51650
+    },
+    {
+      "epoch": 4.891119106229881,
+      "grad_norm": 0.4605785310268402,
+      "learning_rate": 4.355235750804772e-07,
+      "loss": 0.1033,
+      "step": 51660
+    },
+    {
+      "epoch": 4.89206589661049,
+      "grad_norm": 0.57819002866745,
+      "learning_rate": 4.3173641355803826e-07,
+      "loss": 0.0947,
+      "step": 51670
+    },
+    {
+      "epoch": 4.8930126869911,
+      "grad_norm": 0.48372822999954224,
+      "learning_rate": 4.279492520355993e-07,
+      "loss": 0.104,
+      "step": 51680
+    },
+    {
+      "epoch": 4.89395947737171,
+      "grad_norm": 0.5683338642120361,
+      "learning_rate": 4.241620905131604e-07,
+      "loss": 0.1037,
+      "step": 51690
+    },
+    {
+      "epoch": 4.89490626775232,
+      "grad_norm": 0.5659564733505249,
+      "learning_rate": 4.203749289907215e-07,
+      "loss": 0.0902,
+      "step": 51700
+    },
+    {
+      "epoch": 4.895853058132929,
+      "grad_norm": 0.5283416509628296,
+      "learning_rate": 4.165877674682826e-07,
+      "loss": 0.0962,
+      "step": 51710
+    },
+    {
+      "epoch": 4.896799848513539,
+      "grad_norm": 0.5911533236503601,
+      "learning_rate": 4.1280060594584364e-07,
+      "loss": 0.0885,
+      "step": 51720
+    },
+    {
+      "epoch": 4.897746638894149,
+      "grad_norm": 0.5418230891227722,
+      "learning_rate": 4.090134444234047e-07,
+      "loss": 0.092,
+      "step": 51730
+    },
+    {
+      "epoch": 4.898693429274759,
+      "grad_norm": 0.6199533939361572,
+      "learning_rate": 4.0522628290096575e-07,
+      "loss": 0.0958,
+      "step": 51740
+    },
+    {
+      "epoch": 4.899640219655368,
+      "grad_norm": 0.4867435097694397,
+      "learning_rate": 4.014391213785268e-07,
+      "loss": 0.1006,
+      "step": 51750
+    },
+    {
+      "epoch": 4.900587010035978,
+      "grad_norm": 0.6489893198013306,
+      "learning_rate": 3.9765195985608786e-07,
+      "loss": 0.0924,
+      "step": 51760
+    },
+    {
+      "epoch": 4.901533800416588,
+      "grad_norm": 0.5418457388877869,
+      "learning_rate": 3.9386479833364897e-07,
+      "loss": 0.0877,
+      "step": 51770
+    },
+    {
+      "epoch": 4.902480590797198,
+      "grad_norm": 0.4436706006526947,
+      "learning_rate": 3.9007763681121003e-07,
+      "loss": 0.0911,
+      "step": 51780
+    },
+    {
+      "epoch": 4.903427381177807,
+      "grad_norm": 0.6262994408607483,
+      "learning_rate": 3.8629047528877114e-07,
+      "loss": 0.0925,
+      "step": 51790
+    },
+    {
+      "epoch": 4.904374171558417,
+      "grad_norm": 0.5168532729148865,
+      "learning_rate": 3.825033137663322e-07,
+      "loss": 0.0854,
+      "step": 51800
+    },
+    {
+      "epoch": 4.905320961939027,
+      "grad_norm": 0.578430712223053,
+      "learning_rate": 3.787161522438932e-07,
+      "loss": 0.0937,
+      "step": 51810
+    },
+    {
+      "epoch": 4.9062677523196365,
+      "grad_norm": 0.49829116463661194,
+      "learning_rate": 3.749289907214543e-07,
+      "loss": 0.0913,
+      "step": 51820
+    },
+    {
+      "epoch": 4.907214542700246,
+      "grad_norm": 0.6826627254486084,
+      "learning_rate": 3.7114182919901536e-07,
+      "loss": 0.0982,
+      "step": 51830
+    },
+    {
+      "epoch": 4.908161333080856,
+      "grad_norm": 0.670319139957428,
+      "learning_rate": 3.673546676765764e-07,
+      "loss": 0.0905,
+      "step": 51840
+    },
+    {
+      "epoch": 4.909108123461466,
+      "grad_norm": 0.5203000903129578,
+      "learning_rate": 3.635675061541375e-07,
+      "loss": 0.0947,
+      "step": 51850
+    },
+    {
+      "epoch": 4.9100549138420755,
+      "grad_norm": 0.6277847290039062,
+      "learning_rate": 3.597803446316986e-07,
+      "loss": 0.09,
+      "step": 51860
+    },
+    {
+      "epoch": 4.911001704222685,
+      "grad_norm": 0.4402257204055786,
+      "learning_rate": 3.559931831092597e-07,
+      "loss": 0.0889,
+      "step": 51870
+    },
+    {
+      "epoch": 4.911948494603295,
+      "grad_norm": 0.5525423884391785,
+      "learning_rate": 3.5220602158682074e-07,
+      "loss": 0.0873,
+      "step": 51880
+    },
+    {
+      "epoch": 4.912895284983905,
+      "grad_norm": 0.5541327595710754,
+      "learning_rate": 3.4841886006438174e-07,
+      "loss": 0.0901,
+      "step": 51890
+    },
+    {
+      "epoch": 4.9138420753645145,
+      "grad_norm": 0.7090700268745422,
+      "learning_rate": 3.446316985419428e-07,
+      "loss": 0.099,
+      "step": 51900
+    },
+    {
+      "epoch": 4.914788865745124,
+      "grad_norm": 0.45772573351860046,
+      "learning_rate": 3.408445370195039e-07,
+      "loss": 0.0945,
+      "step": 51910
+    },
+    {
+      "epoch": 4.915735656125734,
+      "grad_norm": 0.5060674548149109,
+      "learning_rate": 3.3705737549706496e-07,
+      "loss": 0.0889,
+      "step": 51920
+    },
+    {
+      "epoch": 4.916682446506344,
+      "grad_norm": 0.5407578349113464,
+      "learning_rate": 3.3327021397462607e-07,
+      "loss": 0.0935,
+      "step": 51930
+    },
+    {
+      "epoch": 4.917629236886953,
+      "grad_norm": 0.5680246353149414,
+      "learning_rate": 3.294830524521871e-07,
+      "loss": 0.0903,
+      "step": 51940
+    },
+    {
+      "epoch": 4.918576027267563,
+      "grad_norm": 0.4451833963394165,
+      "learning_rate": 3.2569589092974823e-07,
+      "loss": 0.0966,
+      "step": 51950
+    },
+    {
+      "epoch": 4.919522817648173,
+      "grad_norm": 0.5037840008735657,
+      "learning_rate": 3.2190872940730924e-07,
+      "loss": 0.0854,
+      "step": 51960
+    },
+    {
+      "epoch": 4.920469608028783,
+      "grad_norm": 0.614662766456604,
+      "learning_rate": 3.181215678848703e-07,
+      "loss": 0.093,
+      "step": 51970
+    },
+    {
+      "epoch": 4.921416398409392,
+      "grad_norm": 0.5942671895027161,
+      "learning_rate": 3.1433440636243135e-07,
+      "loss": 0.0971,
+      "step": 51980
+    },
+    {
+      "epoch": 4.922363188790002,
+      "grad_norm": 0.5831498503684998,
+      "learning_rate": 3.1054724483999246e-07,
+      "loss": 0.0887,
+      "step": 51990
+    },
+    {
+      "epoch": 4.923309979170612,
+      "grad_norm": 0.5798229575157166,
+      "learning_rate": 3.067600833175535e-07,
+      "loss": 0.0842,
+      "step": 52000
+    },
+    {
+      "epoch": 4.924256769551222,
+      "grad_norm": 0.44426295161247253,
+      "learning_rate": 3.029729217951146e-07,
+      "loss": 0.0889,
+      "step": 52010
+    },
+    {
+      "epoch": 4.925203559931831,
+      "grad_norm": 0.7449190616607666,
+      "learning_rate": 2.991857602726756e-07,
+      "loss": 0.0945,
+      "step": 52020
+    },
+    {
+      "epoch": 4.926150350312441,
+      "grad_norm": 0.5424138307571411,
+      "learning_rate": 2.9539859875023673e-07,
+      "loss": 0.0899,
+      "step": 52030
+    },
+    {
+      "epoch": 4.927097140693051,
+      "grad_norm": 0.5666702389717102,
+      "learning_rate": 2.916114372277978e-07,
+      "loss": 0.0969,
+      "step": 52040
+    },
+    {
+      "epoch": 4.928043931073661,
+      "grad_norm": 0.5209141373634338,
+      "learning_rate": 2.8782427570535884e-07,
+      "loss": 0.1057,
+      "step": 52050
+    },
+    {
+      "epoch": 4.92899072145427,
+      "grad_norm": 0.5510291457176208,
+      "learning_rate": 2.840371141829199e-07,
+      "loss": 0.0918,
+      "step": 52060
+    },
+    {
+      "epoch": 4.92993751183488,
+      "grad_norm": 0.46501290798187256,
+      "learning_rate": 2.80249952660481e-07,
+      "loss": 0.0858,
+      "step": 52070
+    },
+    {
+      "epoch": 4.93088430221549,
+      "grad_norm": 0.4720863103866577,
+      "learning_rate": 2.7646279113804206e-07,
+      "loss": 0.0912,
+      "step": 52080
+    },
+    {
+      "epoch": 4.9318310925960995,
+      "grad_norm": 0.48449814319610596,
+      "learning_rate": 2.726756296156031e-07,
+      "loss": 0.0955,
+      "step": 52090
+    },
+    {
+      "epoch": 4.932777882976709,
+      "grad_norm": 0.4848043918609619,
+      "learning_rate": 2.6888846809316417e-07,
+      "loss": 0.0839,
+      "step": 52100
+    },
+    {
+      "epoch": 4.933724673357319,
+      "grad_norm": 0.5846676826477051,
+      "learning_rate": 2.651013065707253e-07,
+      "loss": 0.1018,
+      "step": 52110
+    },
+    {
+      "epoch": 4.934671463737929,
+      "grad_norm": 0.5330713391304016,
+      "learning_rate": 2.6131414504828633e-07,
+      "loss": 0.0993,
+      "step": 52120
+    },
+    {
+      "epoch": 4.9356182541185385,
+      "grad_norm": 0.6754573583602905,
+      "learning_rate": 2.575269835258474e-07,
+      "loss": 0.0799,
+      "step": 52130
+    },
+    {
+      "epoch": 4.936565044499148,
+      "grad_norm": 0.6761156916618347,
+      "learning_rate": 2.5373982200340844e-07,
+      "loss": 0.091,
+      "step": 52140
+    },
+    {
+      "epoch": 4.937511834879758,
+      "grad_norm": 0.5080304741859436,
+      "learning_rate": 2.4995266048096955e-07,
+      "loss": 0.0866,
+      "step": 52150
+    },
+    {
+      "epoch": 4.938458625260368,
+      "grad_norm": 0.6686791777610779,
+      "learning_rate": 2.461654989585306e-07,
+      "loss": 0.0964,
+      "step": 52160
+    },
+    {
+      "epoch": 4.9394054156409775,
+      "grad_norm": 0.7539730668067932,
+      "learning_rate": 2.4237833743609166e-07,
+      "loss": 0.1112,
+      "step": 52170
+    },
+    {
+      "epoch": 4.940352206021587,
+      "grad_norm": 0.5886653661727905,
+      "learning_rate": 2.385911759136527e-07,
+      "loss": 0.0957,
+      "step": 52180
+    },
+    {
+      "epoch": 4.941298996402197,
+      "grad_norm": 0.7578022480010986,
+      "learning_rate": 2.3480401439121383e-07,
+      "loss": 0.0875,
+      "step": 52190
+    },
+    {
+      "epoch": 4.942245786782806,
+      "grad_norm": 0.584598958492279,
+      "learning_rate": 2.3101685286877486e-07,
+      "loss": 0.0958,
+      "step": 52200
+    },
+    {
+      "epoch": 4.9431925771634155,
+      "grad_norm": 0.6678522229194641,
+      "learning_rate": 2.2722969134633594e-07,
+      "loss": 0.0882,
+      "step": 52210
+    },
+    {
+      "epoch": 4.944139367544025,
+      "grad_norm": 0.49344995617866516,
+      "learning_rate": 2.2344252982389702e-07,
+      "loss": 0.0962,
+      "step": 52220
+    },
+    {
+      "epoch": 4.945086157924635,
+      "grad_norm": 0.5028517842292786,
+      "learning_rate": 2.196553683014581e-07,
+      "loss": 0.081,
+      "step": 52230
+    },
+    {
+      "epoch": 4.946032948305245,
+      "grad_norm": 0.5575220584869385,
+      "learning_rate": 2.1586820677901913e-07,
+      "loss": 0.0837,
+      "step": 52240
+    },
+    {
+      "epoch": 4.9469797386858545,
+      "grad_norm": 0.733795166015625,
+      "learning_rate": 2.120810452565802e-07,
+      "loss": 0.1,
+      "step": 52250
+    },
+    {
+      "epoch": 4.947926529066464,
+      "grad_norm": 0.5502561926841736,
+      "learning_rate": 2.082938837341413e-07,
+      "loss": 0.0875,
+      "step": 52260
+    },
+    {
+      "epoch": 4.948873319447074,
+      "grad_norm": 0.6022515296936035,
+      "learning_rate": 2.0450672221170235e-07,
+      "loss": 0.0782,
+      "step": 52270
+    },
+    {
+      "epoch": 4.949820109827684,
+      "grad_norm": 0.6461271643638611,
+      "learning_rate": 2.007195606892634e-07,
+      "loss": 0.099,
+      "step": 52280
+    },
+    {
+      "epoch": 4.9507669002082935,
+      "grad_norm": 0.5273721814155579,
+      "learning_rate": 1.9693239916682449e-07,
+      "loss": 0.0911,
+      "step": 52290
+    },
+    {
+      "epoch": 4.951713690588903,
+      "grad_norm": 0.4946170151233673,
+      "learning_rate": 1.9314523764438557e-07,
+      "loss": 0.0851,
+      "step": 52300
+    },
+    {
+      "epoch": 4.952660480969513,
+      "grad_norm": 0.5599589347839355,
+      "learning_rate": 1.893580761219466e-07,
+      "loss": 0.0975,
+      "step": 52310
+    },
+    {
+      "epoch": 4.953607271350123,
+      "grad_norm": 0.565903902053833,
+      "learning_rate": 1.8557091459950768e-07,
+      "loss": 0.0834,
+      "step": 52320
+    },
+    {
+      "epoch": 4.954554061730732,
+      "grad_norm": 0.6564542055130005,
+      "learning_rate": 1.8178375307706876e-07,
+      "loss": 0.0998,
+      "step": 52330
+    },
+    {
+      "epoch": 4.955500852111342,
+      "grad_norm": 0.5040982961654663,
+      "learning_rate": 1.7799659155462984e-07,
+      "loss": 0.0962,
+      "step": 52340
+    },
+    {
+      "epoch": 4.956447642491952,
+      "grad_norm": 0.6264041662216187,
+      "learning_rate": 1.7420943003219087e-07,
+      "loss": 0.0967,
+      "step": 52350
+    },
+    {
+      "epoch": 4.957394432872562,
+      "grad_norm": 0.6681390404701233,
+      "learning_rate": 1.7042226850975195e-07,
+      "loss": 0.0926,
+      "step": 52360
+    },
+    {
+      "epoch": 4.958341223253171,
+      "grad_norm": 0.6040511727333069,
+      "learning_rate": 1.6663510698731304e-07,
+      "loss": 0.0899,
+      "step": 52370
+    },
+    {
+      "epoch": 4.959288013633781,
+      "grad_norm": 0.6360688209533691,
+      "learning_rate": 1.6284794546487412e-07,
+      "loss": 0.1003,
+      "step": 52380
+    },
+    {
+      "epoch": 4.960234804014391,
+      "grad_norm": 0.5660171508789062,
+      "learning_rate": 1.5906078394243515e-07,
+      "loss": 0.1008,
+      "step": 52390
+    },
+    {
+      "epoch": 4.961181594395001,
+      "grad_norm": 0.5813804864883423,
+      "learning_rate": 1.5527362241999623e-07,
+      "loss": 0.093,
+      "step": 52400
+    },
+    {
+      "epoch": 4.96212838477561,
+      "grad_norm": 0.5464773774147034,
+      "learning_rate": 1.514864608975573e-07,
+      "loss": 0.1,
+      "step": 52410
+    },
+    {
+      "epoch": 4.96307517515622,
+      "grad_norm": 0.6224668622016907,
+      "learning_rate": 1.4769929937511836e-07,
+      "loss": 0.0953,
+      "step": 52420
+    },
+    {
+      "epoch": 4.96402196553683,
+      "grad_norm": 0.5166348814964294,
+      "learning_rate": 1.4391213785267942e-07,
+      "loss": 0.0877,
+      "step": 52430
+    },
+    {
+      "epoch": 4.96496875591744,
+      "grad_norm": 0.5422220826148987,
+      "learning_rate": 1.401249763302405e-07,
+      "loss": 0.0908,
+      "step": 52440
+    },
+    {
+      "epoch": 4.965915546298049,
+      "grad_norm": 0.5344117879867554,
+      "learning_rate": 1.3633781480780156e-07,
+      "loss": 0.0997,
+      "step": 52450
+    },
+    {
+      "epoch": 4.966862336678659,
+      "grad_norm": 0.46766147017478943,
+      "learning_rate": 1.3255065328536264e-07,
+      "loss": 0.0754,
+      "step": 52460
+    },
+    {
+      "epoch": 4.967809127059269,
+      "grad_norm": 0.5525385141372681,
+      "learning_rate": 1.287634917629237e-07,
+      "loss": 0.09,
+      "step": 52470
+    },
+    {
+      "epoch": 4.9687559174398785,
+      "grad_norm": 0.5809475183486938,
+      "learning_rate": 1.2497633024048478e-07,
+      "loss": 0.0963,
+      "step": 52480
+    },
+    {
+      "epoch": 4.969702707820488,
+      "grad_norm": 0.6854071021080017,
+      "learning_rate": 1.2118916871804583e-07,
+      "loss": 0.1061,
+      "step": 52490
+    },
+    {
+      "epoch": 4.970649498201098,
+      "grad_norm": 0.5188122391700745,
+      "learning_rate": 1.1740200719560691e-07,
+      "loss": 0.0921,
+      "step": 52500
+    },
+    {
+      "epoch": 4.971596288581708,
+      "grad_norm": 0.5574137568473816,
+      "learning_rate": 1.1361484567316797e-07,
+      "loss": 0.0948,
+      "step": 52510
+    },
+    {
+      "epoch": 4.9725430789623175,
+      "grad_norm": 0.44761478900909424,
+      "learning_rate": 1.0982768415072905e-07,
+      "loss": 0.0905,
+      "step": 52520
+    },
+    {
+      "epoch": 4.973489869342927,
+      "grad_norm": 0.5985987782478333,
+      "learning_rate": 1.060405226282901e-07,
+      "loss": 0.0954,
+      "step": 52530
+    },
+    {
+      "epoch": 4.974436659723537,
+      "grad_norm": 0.4684286415576935,
+      "learning_rate": 1.0225336110585117e-07,
+      "loss": 0.0884,
+      "step": 52540
+    },
+    {
+      "epoch": 4.975383450104147,
+      "grad_norm": 0.5261790156364441,
+      "learning_rate": 9.846619958341224e-08,
+      "loss": 0.0846,
+      "step": 52550
+    },
+    {
+      "epoch": 4.9763302404847565,
+      "grad_norm": 0.4636407196521759,
+      "learning_rate": 9.46790380609733e-08,
+      "loss": 0.0934,
+      "step": 52560
+    },
+    {
+      "epoch": 4.977277030865366,
+      "grad_norm": 0.5825064182281494,
+      "learning_rate": 9.089187653853438e-08,
+      "loss": 0.0837,
+      "step": 52570
+    },
+    {
+      "epoch": 4.978223821245976,
+      "grad_norm": 0.538732647895813,
+      "learning_rate": 8.710471501609544e-08,
+      "loss": 0.095,
+      "step": 52580
+    },
+    {
+      "epoch": 4.979170611626586,
+      "grad_norm": 0.6379238963127136,
+      "learning_rate": 8.331755349365652e-08,
+      "loss": 0.0948,
+      "step": 52590
+    },
+    {
+      "epoch": 4.980117402007195,
+      "grad_norm": 0.5082975029945374,
+      "learning_rate": 7.953039197121757e-08,
+      "loss": 0.0952,
+      "step": 52600
+    },
+    {
+      "epoch": 4.981064192387805,
+      "grad_norm": 0.5919937491416931,
+      "learning_rate": 7.574323044877865e-08,
+      "loss": 0.0916,
+      "step": 52610
+    },
+    {
+      "epoch": 4.982010982768415,
+      "grad_norm": 0.654510498046875,
+      "learning_rate": 7.195606892633971e-08,
+      "loss": 0.0862,
+      "step": 52620
+    },
+    {
+      "epoch": 4.982957773149025,
+      "grad_norm": 0.6276813745498657,
+      "learning_rate": 6.816890740390078e-08,
+      "loss": 0.0938,
+      "step": 52630
+    },
+    {
+      "epoch": 4.983904563529634,
+      "grad_norm": 0.5056014060974121,
+      "learning_rate": 6.438174588146185e-08,
+      "loss": 0.0918,
+      "step": 52640
+    },
+    {
+      "epoch": 4.984851353910244,
+      "grad_norm": 0.49018388986587524,
+      "learning_rate": 6.059458435902292e-08,
+      "loss": 0.0882,
+      "step": 52650
+    },
+    {
+      "epoch": 4.985798144290854,
+      "grad_norm": 0.5894728899002075,
+      "learning_rate": 5.6807422836583984e-08,
+      "loss": 0.0977,
+      "step": 52660
+    },
+    {
+      "epoch": 4.986744934671464,
+      "grad_norm": 0.5480173230171204,
+      "learning_rate": 5.302026131414505e-08,
+      "loss": 0.0968,
+      "step": 52670
+    },
+    {
+      "epoch": 4.987691725052073,
+      "grad_norm": 0.576349139213562,
+      "learning_rate": 4.923309979170612e-08,
+      "loss": 0.0972,
+      "step": 52680
+    },
+    {
+      "epoch": 4.988638515432683,
+      "grad_norm": 0.4693711996078491,
+      "learning_rate": 4.544593826926719e-08,
+      "loss": 0.0839,
+      "step": 52690
+    },
+    {
+      "epoch": 4.989585305813293,
+      "grad_norm": 0.3827964961528778,
+      "learning_rate": 4.165877674682826e-08,
+      "loss": 0.0773,
+      "step": 52700
+    },
+    {
+      "epoch": 4.990532096193903,
+      "grad_norm": 0.5371719598770142,
+      "learning_rate": 3.787161522438933e-08,
+      "loss": 0.0852,
+      "step": 52710
+    },
+    {
+      "epoch": 4.991478886574512,
+      "grad_norm": 0.5461321473121643,
+      "learning_rate": 3.408445370195039e-08,
+      "loss": 0.0826,
+      "step": 52720
+    },
+    {
+      "epoch": 4.992425676955122,
+      "grad_norm": 0.5288814902305603,
+      "learning_rate": 3.029729217951146e-08,
+      "loss": 0.1059,
+      "step": 52730
+    },
+    {
+      "epoch": 4.993372467335732,
+      "grad_norm": 0.6059397459030151,
+      "learning_rate": 2.6510130657072527e-08,
+      "loss": 0.0953,
+      "step": 52740
+    },
+    {
+      "epoch": 4.9943192577163416,
+      "grad_norm": 0.6540989279747009,
+      "learning_rate": 2.2722969134633595e-08,
+      "loss": 0.0852,
+      "step": 52750
+    },
+    {
+      "epoch": 4.995266048096951,
+      "grad_norm": 0.5415211319923401,
+      "learning_rate": 1.8935807612194664e-08,
+      "loss": 0.0922,
+      "step": 52760
+    },
+    {
+      "epoch": 4.996212838477561,
+      "grad_norm": 0.715372622013092,
+      "learning_rate": 1.514864608975573e-08,
+      "loss": 0.0942,
+      "step": 52770
+    },
+    {
+      "epoch": 4.997159628858171,
+      "grad_norm": 0.5425447225570679,
+      "learning_rate": 1.1361484567316798e-08,
+      "loss": 0.0904,
+      "step": 52780
+    },
+    {
+      "epoch": 4.9981064192387805,
+      "grad_norm": 0.6264485716819763,
+      "learning_rate": 7.574323044877864e-09,
+      "loss": 0.089,
+      "step": 52790
+    },
+    {
+      "epoch": 4.99905320961939,
+      "grad_norm": 0.5500597953796387,
+      "learning_rate": 3.787161522438932e-09,
+      "loss": 0.1056,
+      "step": 52800
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 1.679782748222351,
+      "learning_rate": 0.0,
+      "loss": 0.0979,
+      "step": 52810
+    },
+    {
+      "epoch": 5.0,
+      "eval_f1_micro": 0.3872254461700786,
+      "eval_loss": 0.11608477681875229,
+      "eval_precision": 0.5393189964157706,
+      "eval_recall": 0.3020454864805187,
+      "eval_runtime": 367.8742,
+      "eval_samples_per_second": 114.835,
+      "eval_steps_per_second": 7.179,
+      "step": 52810
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 52810,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.55886064391168e+16,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}